A Guide to Kant’s Psychologism (Routledge Studies in Eighteenth-Century Philosophy) 9780367141110, 9780429030222, 0367141116

Citation preview

A Guide to Kant’s Psychologism

This book presents an interpretation of Kant’s Critique of Pure Reason as a priori psychologism. It groups Kant’s philosophy together with those of the British empiricists—Locke, Berkeley, and Hume—in a single line of psychologistic succession and offers a clear explanation of how Kant’s psychologism differs from psychology and idealism. The book reconciles Kant’s philosophy with subsequent developments in science and mathematics, including post-Fregean mathematical logic, non-Euclidean geometry, and both relativity and quantum theory. It also relates Kant’s psychologism to Wittgenstein’s later conception of language. Finally, the author reveals the ways in which Kant’s philosophy dovetails with contemporary scientific theorizing about the natural phenomenon of consciousness and its place in nature. This book will be of interest to Kant scholars and historians of philosophy working on the British empiricists. Wayne Waxman is the author of Kant’s Anatomy of the Intelligent Mind, Kant and the Empiricists, Hume’s Theory of Consciousness, and Kant’s Model of the Mind. He is retired and lives in New Zealand.

Routledge Studies in Eighteenth-Century Philosophy

Rousseau’s Ethics of Truth A Sublime Science of Simple Souls Jason Neidleman Kant and the Scottish Enlightenment Edited by Elizabeth Robinson and Chris W. Surprenant Kant and the Reorientation of Aesthetics Finding the World Joseph J. Tinguely Hume’s Science of Human Nature Scientific Realism, Reason, and Substantial Explanation David Landy Hume’s Moral Philosophy and Contemporary Psychology Edited by Philip A. Reed and Rico Vitz Kant and the Problem of Self-Knowledge Luca Forgione Kant on Intuition Western and Asian Perspectives on Transcendental Idealism Edited by Stephen R. Palmquist Hume on Art, Emotions, and Superstition A Critical Study of the Four Dissertations Amyas Merivale A Guide to Kant’s Psychologism via Locke, Berkeley, Hume, and Wittgenstein Wayne Waxman For more information about this series, please visit: www.routledge.com/ Routledge-Studies-in-Eighteenth-Century-Philosophy/book-series/SE0391

A Guide to Kant’s Psychologism via Locke, Berkeley, Hume, and Wittgenstein Wayne Waxman

First published 2019 by Routledge 52 Vanderbilt Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2019 Taylor & Francis The right of Wayne Waxman to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested ISBN: 978-0-367-14111-0 (hbk) ISBN: 978-0-429-03022-2 (ebk) Typeset in Sabon by Apex CoVantage, LLC

He mea tāpae tēnei pukapuka ki a Immanuel Kant, ā, nā tōna rapunga whakaaro i āhei atu ahau ki ngā momo whakaaro mutungakore me ngā whakamānawa mātauranga huhua noa, tae atu ki te iwi whakamīharo o Aotearoa mō te ngākau makuru, ngākau atawhai hoki i whakamanuhiri nei i tēnei manene nō tētahi whenua e noho ana i roto i te pōuri.

Contents

Abbreviations Used in Referencingviii

Introduction: Kant’s Copernican Revolution

  1 Breaking the Mold

1 18

PART I

The Path to Kant: Psychologism and Conventionalism33   2 Locke’s Oyster

39

  3 Berkeley’s Vision

65

  4 Hume’s Cement

93

  5 From Hume to Kant via Wittgenstein

129

PART II

Nature in Mind: Through Kant’s I’s147   6 The Kantian Cogito155   7 The Logical I

167

  8 The Aesthetic I

198

  9 The Objective I

229

10 The I of Nature

262



304

Conclusion: After Kant

Glossary of Kantian Terms319 Bibliography329 Index331

Abbreviations Used in Referencing

Kant. All references are to, or derived from, critical editions based on the Prussian Academy edition of the Gesammelte Schriften, begun in 1901, and generally cited as the Akademie Ausgabe, abbreviated “AA.” For Kant’s published works, I mostly use the standard abbreviation scheme: “CPR” for the Critique of Pure Reason, with “A” for the 1781 edition and “B” for the 1787 edition; “CPrR” for the Critique of Practical Reason, “CJ” for the Critique of Judgment, “ID” for the Inaugural Dissertation, “PFM” for Prolegomena to Any Future Metaphysics, and “MFPNS” for Metaphysical First Principles of Natural Science. Some works I refer to by name: “Discovery” for On a Discovery Whereby Any New Critique of Pure Reason Is to Be Made Superfluous by an Older One, “Anthropology” for Anthropology From a Pragmatic Viewpoint, and “Logic” for the work of that name published posthumously by Kant’s editor G. B. Jäsche. Translations are my own. Locke. The edition of the Essay Concerning Human Understanding I  employ is the 1975 Oxford Clarendon, edited by Peter H. Nidditch, abbreviated “ECHU.” References will conform to the following practice: “ECHU 2.3.6” refers to the second of the Essay’s four books, chapter 3, numbered paragraph six. I have occasionally taken the liberty of updating Locke’s syntax and punctuation. Berkeley. Nearly all references are to Philosophical Works, Including the Works on Vision, edited by Michael R. Ayers (London: J. M. Dent, 1975). References will conform to the following practices: “V” for the New Theory of Vision; “VV” for A New Theory of Vision Vindicated; “PHK Intr. §3” for the third section of the Introduction to A Treatise Concerning the Principles of Human Knowledge; and “Alciphron” followed by the dialogue and section numbers for Alciphron, or the Minute Philosopher in Focus, edited by David Berman (London: Routledge, 1993). I have occasionally taken the liberty of updating Berkeley’s syntax and punctuation. Hume. Citations of A Treatise of Human Nature are from the Oxford Clarendon 1978 edition by Peter H. Nidditch, revising the 1888 edition by L. A. Selby-Bigge, abbreviated “T” followed by book, section, and

Abbreviations Used in Referencing ix paragraph. Citations from An Enquiry Concerning Human Understanding are from the 1999 Oxford University Press edition by Tom L. Beauchamp, abbreviated “EHU IV/i ¶3,” for the third paragraph of the first part of the fourth section. I have occasionally taken the liberty of updating Hume’s syntax and punctuation. Descartes. All references to Descartes are from the translation by John Cottingham, Robert Stoothoff, Dugald Murdoch, and Anthony Kenny, published in 1984, by Cambridge University Press. References cite abbreviated names or parts of individual works, such as “First Meditation” for Meditations on First Philosophy, together with the relevant volume and page number of the standard Latin and French edition by Adam and Tannery, e.g. “AT VII 27.” ­ revious Other. Since I  shall have occasion to refer to some of my p books, I shall use the abbreviations “KEUU” for Kant and the E ­ mpiricists: Understanding Understanding (New York: Oxford University Press, 2005), “KAIM” for Kant’s Anatomy of the Intelligent Mind (New York: Oxford University Press, 2014), “HTC” for Hume’s Theory of Consciousness (Cambridge: Cambridge University Press, 1994), and “KMM” for Kant’s Model of the Mind: A New Interpretation of Transcendental Idealism (New York: Oxford University Press, 1991).

Introduction Kant’s Copernican Revolution

What did K. mean? Galileo read further: “The transcendental unity of apperception is that unity through which all the manifold given in an intuition is united in a concept of the object.  .  .  .” The transcendental unity of apperception? What was K. trying to say? Why was it so hard to fathom? Galileo turned, ready to ask, but K. was gone. Galileo turned to Alturi, who had been listening and seemed bemused. K. might be right, said Galileo, but why should it be said so obscurely? Murky thoughts, like murky waters, can serve two purposes only: to hide what lies beneath, which is our ignorance, or to make the shallow seem deep. For once, Alturi nodded. “K. is bad enough, but by no means the worst of them,” he laughed, and then he added: “Avoid philosophers. In the fog of thought, banality dresses like a mysterious mistress.” Perhaps, said Galileo, but so does truth itself. —Giulio Tononi, Phi: A Voyage From the Brain to the Soul

This is a book about consciousness as viewed by the man who perhaps even to this date remains its foremost philosophical theorist, Immanuel Kant (1724–1804). It is aimed not merely, or even mainly, at scholars of the history of philosophy, but at everyone, philosopher and non-­ philosopher alike, with a serious interest in consciousness and its place in nature. Accordingly, no knowledge of the history of philosophy or indeed philosophy of any kind is presupposed, only an interest in the topic that extends to queries that, given the present state of knowledge, may still be regarded as philosophical. Primary texts are cited only to infuse key ideas with the flavor of their historical sources. I shall therefore bypass the myriad scholarly disputes engulfing Kant’s hermetic writings to focus exclusively on theory, and more particularly the idea that I believe lies at the heart of his philosophy: a priori psychologism. I will introduce this notion in due course. First, there are a number of preliminary matters to cover. A glossary of Kantian terms can be found at the end of the book. Although Kant’s terminology is arcane and may be found off-putting,

2  Introduction I have opted to adhere to it for a simple reason: since virtually all philosophies subsequent to Kant are built on anti-psychologistic foundations, there is no language better suited to expounding a priori psychologism or more familiar than Kant’s own. I have also departed from the common expository practice of drawing on post-Kantian treatments of the topics considered in his philosophy to illuminate his own. While motivated by the laudable desire to make Kant’s ideas more accessible to contemporary readers, the practice is based on an assumption I regard as mistaken: that Kant was an opponent of psychologism who shared the view of virtually all recent and contemporary philosophers that psychologism is not just false but a fallacy. On my reading, by contrast, far from being an opponent, Kant emerges as psychologism’s foremost exemplar. If true, it means that drawing on post-Kantian philosophy to update Kant’s theory of consciousness and make it more comprehensible is all but certain to obscure and mislead rather than illuminate.1 There is no gainsaying it: interpreting Kant’s theory of consciousness as I  do complicates matters for readers. When not interpreted as a priori psychologism, the psychological portions of Kant’s philosophy, along with much else, straightaway become redundant or impracticable. This has provided many expositors the opportunity to marginalize or altogether dispense with the purportedly excess conceptual apparatus, together with the associated terminology, and thereby produce less dense, jargony presentations of the philosophy. When interpreted as I do, however, every component of Kant’s philosophy proves equally essential, psychological no less than non-psychological; and since the former are almost invariably the most difficult to understand, both in their own right and in how they fit together with the rest, a certain amount of detailed exegesis is inevitable. I can only hope that anyone committed to understanding Kant’s philosophy will deem grappling with a challenging presentation a price well worth paying to master the idea at its heart. Kant’s reputation has been severely damaged by interpretations that convict him of leaving hostages to fortune in a number of matters of direct relevance to science. One of my primary aims in this book is to debunk these common misconceptions. Starting from the premise that Kant’s philosophy is best interpreted as a priori psychologism, I  show how doing so permits (i) Kant’s theory of space to accommodate non-Euclidean geometry, (ii) his theory of time to accommodate relativity theory, (iii) his theory of nature to accommodate quantum theory, (iv) his theory of a priori intuition to accommodate higher mathematics, (v) his theory of logic to accommodate mathematical logic, and (vi) his theory of mind to accommodate evolutionary biology. Accordingly, instead of attempting to make Kant’s philosophy more accessible by integrating contemporary philosophy into its presentation, I propose to do so by updating it in the light of contemporary science. This book is therefore addressed not only to philosophers but to linguists, logicians,

Introduction 3 scientists, and mathematicians whose inquiries have brought them to grips with fundamental philosophical questions relating to their disciplines. I hope too that my decision to incorporate Wittgenstein’s conventionalism into the discussions of Kant’s response to Humean skepticism and his theory of non-linguistic propositional thought will further broaden the book’s appeal. My reasons for doing so will be made clear in the synopsis of Chapter 5 in the introduction to Part I. Although aimed at a wide readership, the book is of particular importance for Kant scholars. True, it takes the scholarly case for reading Kant’s philosophy as a priori psychologism for granted in order to focus exclusively on the idea that everywhere informs it. But a book with this brief is needed precisely because scholarly desiderata can obscure or even run at cross purposes to philosophical understanding, especially in the case of an idea as unfamiliar as a priori psychologism is not only to philosophers generally but to Kant scholars in particular. So, in order to equip themselves with the philosophical understanding needed to properly evaluate my scholarly work, I strongly urge Kant scholars to study the present book first. Synopses of Chapters 2–5 are provided at the start of Part I and synopses of Chapters 6–10 at the start of Part II. The foregoing remarks should suffice for the same purpose vis à vis the remainder of the introduction and Chapter 1.

Retrospective This book is the successor to a work on self and understanding in Kant and British empiricism that I  published in two volumes, Kant and the Empiricists: Understanding Understanding [KEUU] (2005) and Kant’s Anatomy of the Intelligent Mind [KAIM] (2014). So, to explain its nature and purpose, I first need to say something about its predecessor. The philosophy expounded in Kant’s Critique of Pure Reason (1781, rev. 1787) is enormously difficult to understand and articulate, so it is not surprising that interpretations vary widely, not only from era to era, country to country, and school to school but even within schools. Nevertheless, there is a basic consensus that informs nearly all of them and has as a result fixed the image of Kant in the minds of philosophers and non-­ philosophers alike. It is perhaps best encapsulated by a 2005 editorial in The Guardian newspaper that declared Kant “the undefeated heavyweight philosophy champion of the world” because he had the “great insight . . . to remove psychology from epistemology.”2 Its removal became necessary because the preeminent place assigned by Locke to psychology in the analysis of thought and knowledge culminated in Hume’s accounts of causality and the identity of bodies and the self purely in terms of customary association. Since that analysis threatened to completely psychologize human understanding of nature, sparing only logic and mathematics, the

4  Introduction consensus presents Kant as having made it his mission to overturn Hume’s psychologization of epistemology by arguing as follows: 1. Cause and effect, space and time, and certain other concepts are essential constituents of objective understanding. 2. To explicate them in this capacity properly, however, requires that these concepts be recognized as a priori. 3. Psychology is incapable of explicating a priori concepts; and such representations as psychology is capable of explicating are unable to fill the role required of essential constituents of objective understanding. 4. Therefore, psychology needs to be removed from epistemology once and for all. There is a comparable negative consensus as to the shortcomings of Kant’s execution of his removal project. First, being the child of his age, Kant’s rejection of psychology left him no option but to fall back on a version of platonistic innatism not all that different from Rationalisms like those of G. W. Leibniz (1646–1716) and Christian Wolff (1679–1754). Second, by formulating the problem he devoted the Critique of Pure Reason to solving in terms of the untenable distinction between synthetic and analytic judgments, he effectively obliged well-meaning interpreters to look for ways to re-frame his project so that it might continue to be taken seriously. Third, by not only casting the Aristotelian logic of his time as a completed science to which nothing could be added or subtracted but also using it as the central organizing principle of his philosophy, Kant left it fatally exposed when the advent of mathematical logic and modern linguistic analysis cast Aristotelian logic into desuetude. Fourth, Kant’s analysis of Euclidean geometry as a necessary and universal ingredient of the representation of space was demolished by the advent of nonEuclidean geometry in the nineteenth century and its utilization in relativity theory in the twentieth. And finally, fifth, Kant left further hostages to fortune when he incorporated so much of the Newtonian conception of nature into his own that his philosophy of nature cannot be sustained in the face of the revolutions wrought by relativity and quantum theory. The consensus as to its extensive deficiencies has encouraged many to attempt to salvage Kant’s removal project by subjecting the presentation of his philosophy in the Critique of Pure Reason to extreme makeovers. Jettisoned are the remnants of early modern psychologism, including a priori sensible forms, a priori and empirical syntheses of imagination, and faculties of a priori understanding and a priori self-consciousness. With these also go their various applications, including everything that tends to psychologize understanding and knowledge of space, time, objectivity, and above all nature itself (with the added benefit of neutralizing the impulse that led Kant to embrace idealism). And with so little of Kant’s original exposition remaining, there is no longer anything to

Introduction 5 prevent interpreters from substituting more favored methods of propositional analysis and philosophical organization in place of Kant’s obsolete ones. Indeed, so far as doing so is deemed to serve Kant’s core aims better than anything possible in his era, the more extreme makeovers are often the most favored since they make Kantianism maximally comprehensible and persuasive to philosophers today. Yet, there are many who cannot help wondering, “why bother?” For them, Kantian makeovers, if less daunting and unfamiliar, tend to be just as philosophically shaky as Kant’s original. To the extent this is so, what is the point of departing from a text that, whatever its flaws, everyone acknowledges to be one of the seminal works in the history of philosophy? Moreover, makeovers can all too easily put one in mind of someone tasked with assembling a piece of novel machinery who excuses his failure to use most of the components and assemble an end-product that performs as advertised by claiming to have divined its designers’ true intentions better than the designers themselves—the proof being all those unnecessary pieces. Here makeover artists can do either of two things: admit defeat and set up in another line of business or lay the blame at Kant’s door. The latter course would be justified if the components of his philosophy were irreparably defective, could not be fitted together to yield a viable end-product, or the idea of the end-product itself were inherently flawed, a kind of philosophical perpetual motion machine. Yet, contending that the errors are all Kant’s can easily come off as special pleading. Shouldn’t our default position instead be that the blame is ours, not Kant’s, if the philosophy we present as his turns out to fall short, possibly far short, of the “Copernican revolution” in philosophy he strove for and claimed to have achieved? Surely, any interpretation of the Critique that discards components willy-nilly, organizes them anachronistically, and substitutes more modest aims in place of Kant’s revolutionary one cannot fairly blame its philosophical shortcomings on Kant. The interpretation developed in KEUU and KAIM is directly opposed to the consensus view. Far from removing psychology from epistemology, the thesis I  argue for is that Kant extended it further than even Hume contemplated doing, going so far as to encompass logic and mathematics within its scope. I call it a priori psychologism because instead of explicating the representations fundamental to logic, mathematics, and nature in terms of empirical sensibility, imagination, and self-consciousness, Kant developed a psychology of a priori sensibility, imagination, and self-consciousness with which to do so. The reason I believe Kant’s commitment to a priori psychologism is almost always overlooked is that the very idea of psychologism cuts so deeply against the contemporary grain that it is seldom even descried as part of the conceptual landscape, much less systematically and rigorously explored as a way of making sense of Kant. Having no predisposition either way, I  arrived at my interpretation because a priori psychologism seemed to me a far better textual,

6  Introduction contextual, and historical fit than the consensus view. And when all is said and done, what could be more revolutionary than a philosophy that inverts the order of explication favored by nearly all philosophers from René Descartes (1596–1650) to the present, and instead of attempting to explicate conscious representation logically, mathematically, scientifically, or metaphysically, does precisely the reverse by explicating all of these in terms of it? No doubt there are many justifications one could give to explain why most philosophers today would regard a priori psychologism as beyond the pale of acceptable theorizing. But what I set out to show in KEUU and KAIM is that a comprehensive, textually and contextually well-grounded interpretation of the transcendental philosophy of the Critique of Pure Reason as a priori psychologism reveals that these justifications do not include anything from the negative side of the consensus interpretation of Kant, all of which can be chalked up to misunderstandings or failures to understand. Not the least of these concerns Kant’s relation to David Hume (1711–76), for what is almost never appreciated is that Kant not only took the problem addressed in the Critique from Hume but appropriated Hume’s psychologistic method to solve it as well.3 This means that without a thorough understanding of Hume’s psychologistic approach to notions such as cause and effect, one cannot hope to comprehend what Kant was doing in the Critique of Pure Reason or recognize that the refutation of Hume presented in it is not, as almost invariably assumed, a rejection of psychologism per se but merely of its restriction to empirical psychology. To this end, KEUU focuses on showing how to read Hume as a bird of the same psychologistic feather as Kant, differing only insofar as the psychology Hume operated with was limited to empirical sensibility, imagination, and self-consciousness, whereas Kant extended his to include a priori iterations of these faculties. Since this cuts against the grain of contemporary Hume interpretation nearly as radically as my reading of Kant diverges from contemporary Kant interpretation,4 KEUU subjects Hume to the same exhaustively detailed textual and historical analysis to which KAIM subjects Kant, even to the point of including two 100+-page treatments of Hume’s principal proto-psychologistic antecedents, John Locke (1632–1704) and George Berkeley (1685–1753). Preceded by a four-­ chapter general introduction to the entire two-volume work that presents preliminary justifications for revising standard history of philosophy practice by grouping Kant together with the empiricists in a single line of psychologistic succession, the result is a volume as long and detailed as KAIM. The drawbacks of presenting Kant’s a priori psychologism in lengthy, densely written tomes are not difficult to divine. Most obviously, the basic lineaments of the theory tend to be obscured by the amount of scholarly spadework needed to establish its interpretive bona fides. The problem is compounded by the fact that text and philosophy are sometimes at cross purposes at crucial junctures, the most notorious being

Introduction 7 the occasion where Kant provides an example of how what he termed “pure concepts of the understanding” need to be understood and then leaves it to the reader to do the same for the other eleven.5 The exercise turns out to be far more difficult than Kant seems to have imagined, thereby confronting his interpreter with a dilemma: Kant’s clear intention is that one go beyond the text to fill in the philosophical gaps, whereas doing so violates the scholarly desideratum that requires one to confine one’s reading of texts to what’s actually in them. Most interpreters opt to leave the gaps unfilled, but thereby run a risk of fatally compromising the philosophy they are purporting to expound—a risk compounded with each successive application of the pure concepts of the understanding in what Kant termed “transcendental synthesis,” “transcendental schematism,” and finally “principles of pure understanding.” Since, in my view, nothing can be more counterproductive in the history of philosophy than to privilege the text at the expense of the philosophy, KAIM takes the opposite tack. Using all the available textual and contextual clues, it extends Kant’s mode of explication to the other pure concepts and then weaves the results into the analyses of transcendental synthesis (esp. in relation to mathematics), transcendental schematism (esp. in relation to physical space and time and nature as a system of objects subject everywhere to law), and the principles of pure understanding (esp. in relation to empirical psychology and the transformation of subjective experience into objective). Apart from the increased risk of saddling Kant with views that were not his, the downside of this procedure is that the account of pure concepts of the understanding in KAIM ends up being more detailed and complex than the Critique itself. Even at a purely philosophical level, a history of philosophy approach to Kant and his British empiricist precursors can obscure nearly as much as it reveals. For example, KAIM presents Kant’s analysis of the intelligent mind to be, at its core, a very simple structure, with all the components of his system either converging upon or diverging from its “supreme principle,” the representation ‘I think’ that he adapted from Descartes’s cogito ergo sum—“I think therefore I am.” But insofar as the book’s organization does not conform to the structure it establishes but instead adheres to history of philosophy canons by following the structure of the Critique itself, that simplicity is irremeably obscured. The only sure remedy for these and other drawbacks of two 600-page specialist monographs is an entirely new presentation that takes the scholarly bona fides for granted and focuses exclusively on Kant’s psychologism itself, even to the point of mirroring its structure. This is that presentation. It starts with the grounding of Kant’s a priori psychologism on his non-metaphysical, non-epistemological, purely logical reinterpretation of the Cartesian cogito and then follows the lines converging upon and diverging from it. I focus first on the latter, for the logically universal character of the ‘I think’ cannot be preserved without supplementation

8  Introduction by innate logical forms of judgment, which in turn open the way to purely psychologistic explications of logically structured propositional thought generally and pure concepts of the understanding in particular. With the ‘I think’ recognized as first and foremost a logical I, it then becomes clear what pre-logical consciousness (i.e. sensibility) has to be in order for the representation ‘I think’ itself to be possible, so that the logical I is revealed to be at the same time an aesthetic I. Putting these together with pure concepts via transcendental synthesis, there next emerges the objective I and finally, once transcendental schematism and the principles of pure understanding are factored in, the I of nature. Along the way, I  consider Kant’s accounts of the representations that instantiate the different phases of the I in beings with minds constituted like ours, including spatial and temporal appearances devoid of logical structure, their transformation into concepts and propositions, the transformation of these into objective space, time, and nature itself via pure concepts of the understanding, and finally their combination with empirical synthesis to yield objective experience, ordinary and scientific alike, via principles of pure understanding. The presentation of Kant’s a priori psychologism accordingly alternates between these two levels of generality, so that neither the inherent simplicity of Kantian egology nor its richness in content and consequences is ever lost from view. And Kant’s empiricist forebears? Having operated in a time and place steeped in psychological philosophy, Kant tended to focus his efforts on those aspects of psychologism where he had the most to contribute and left it to readers sharing this background to fill in the rest. However, as philosophy has turned more and more away from psychology in the intervening centuries, the very idea of psychologism now tends to be reflexively equated with a fallacy. So, bearing in mind that Kant’s psychologism is, at its core, a Humean solution to a Humean problem, I have prefaced my consideration of Kant with four chapters on the British empiricists, the purpose of which is to provide the most direct, perspicuous route to the context in which Kant himself philosophized. The result is a book barely one-quarter the size of its predecessor that is accessible to anyone intrigued by the idea of a non-fallacious psychologistic approach to philosophical questions regarding self, understanding, space, time, and nature. Caveat lector: “simpler,” “more accessible,” and (relatively) “brief” do not necessarily mean “easier.” The interpretation presented here is both detailed and comprehensive: no component of Kant’s psychologism is omitted, all are firmly mortared into their place within the whole, and Kantian terminology is almost everywhere utilized.6

Psychologism Many deny that there is anything very revolutionary about interpreting Kant’s theory of mind as a priori psychologism. They are mistaken.

Introduction 9 Nearly everyone who holds this does so, in my experience, because they fail to properly distinguish ‘psychologism’ from ‘psychology.’ It is obvious to all that Kant concerned himself extensively with psychology, and many scholars have made the case that at least some of his contributions to the discipline remain important even today. But psychology is not psychologism. Psychology, in the sense most relevant to Kant, is psychological philosophy of the kind exemplified by early modern “theory of ideas.” It concerns itself with the origin of notions familiar from language as conscious representations in the mind, and this is how scholars today understand ‘psychology’ in relation to Kant. But if the notion at issue is non-psychological, and particularly if it is deemed a priori and/ or ­objective—as commonly is the case with logical, mathematical, scientific, and metaphysical notions—then, while a psychological account of its origin in us may be interesting and even have important epistemic implications, it can tell us nothing whatsoever about the notion itself—its meaning, the scope of its application, or its origin—and, for these reasons, does not yet count as psychologism. Even those who recognize that Kant held that all notions of space, time, and nature are ineluctably bound up with consciousness typically have in mind not psychologism but idealism, and almost invariably a species of idealism that few seem able to distinguish from Berkeley’s antimaterialism. The problem is that the materialism question belongs not to psychology, or a fortiori to psychologism, but rather to philosophy of science and metaphysics. Even when Kant’s psychologism is not confounded with his idealism, hardly any descry its fundamental principle, order of progress, and structural unity. The preeminent exception is Kant’s most important direct successor, G. W. F. Hegel (1770–1831), who recognized that principle as the Cartesian ‘I think,’ saw that and how Kant used it to psychologistically explicate the logical universality of concepts, and appreciated that everything else in Kant’s psychologism flows inexorably from this purely psycho-logical insight. Unfortunately, Hegel was too busy with his own philosophical system to bequeath any treatises on Kant’s, so that with the shift to the now prevailing view of Kant as desiring to remove psychology from epistemology, the need for such treatises has never been greater.7 So what is psychologism? The term originally referred to a group of nineteenth-century German philosophers with roots in early modern British empiricism. It was then extended to include anyone who introduces psychological ingredients into the definition of non-psychological notions, whereupon it was branded a fallacy, and it is as such that it is still mainly known today. In fact, psychologism, as I  understand it, emerged from Locke’s psychological alternative to the method of definition whereby Descartes and other Rationalists sought to establish the objective bona fides of philosophically important terms by conferring the same rigorous clarity and distinctness on them that mathematicians’

10  Introduction definitions give to theirs.8 The evolution of Locke’s psychological alternative to the method of definition into psychologism started with Berkeley but was not properly grounded or systematically developed until Hume and Kant. In a nutshell, it is this: instead of proceeding by defining notions in terms of other notions without regard to whether there is, or even can be, any corresponding conscious representation, psychologism employs psychology with the express aim of seeking out such representations. If the search reveals that a notion owes none of the contents indispensable to its meaning to consciousness, then its scope of application is nowhere limited by it. But if, on the contrary, it can be shown to be beholden to consciousness for any of the ingredients essential to its meaning—­ingredients at least implicit, but often explicit, in definitions— then its scope of application is limited accordingly. This means that any application of the notion to contexts that exclude consciousness is ipso facto nonsensical in exactly the same way notions such as joy, pain, and daydreaming lose all meaning if applied to something non-conscious. What escapes many critics is that psychologism does not require that consciousness be included in definitions of ostensibly non-psychological notions to which it contributes essential ingredients of meaning. Definitions are crafted to serve particular linguistic purposes, e.g.  to make a term suitable for scientific or mathematical use. Accordingly, with the exception of psychological concepts themselves, incorporating psychological content into the definitions of non-psychological concepts is far more apt to be confusing and counterproductive than helpful. It is only when a non-psychologically definable notion is applied in consciousnessindependent contexts AND psychologistic explication reveals that essential ingredients of its content can be derived from no source other than consciousness that psychologism intervenes to warn against transgressing the bounds of meaning. Once psychologism is properly understood, its relation to idealism can be as well. It is not a metaphysical position asserting that the realities corresponding to our ordinary, scientific, and mathematical notions of space, time, nature, and much else besides are spiritual rather than material in nature. In fact, it has nothing to say on this or any other metaphysical matter, being no more than a method of explicating meaning that happens to be purely psychological. What it does say is that whatever consciousness-independent reality may be, no notion whose meaning and scope of application are ineluctably bound up with consciousness can be used to represent it. Thus, instead of asserting anything about what such reality is, it confines itself to instructing us about what it is not and can never intelligibly be supposed to be. Hume, for example, utilized psychologistic explication to show that all our notions of the cement of the universe—space, time, causality, identity at and through time, substantiality, etc., i.e. everything commonly supposed to be objective save logic and mathematics—depend on

Introduction 11 consciousness for essential elements of their content, and so can no more be conceived to retain their meaning in the absence of consciousness than a molecule can meaningfully be supposed to be stoically undergoing agonies of pain. Kant’s a priori psychologism is an even more extreme case: by going so far as to bring logic and mathematics within its scope in addition to space, time, etc., it leaves us totally bereft of the means to consciously represent consciousness-independent reality, which consequently vanishes into the oblivion he designated by such terms as ‘thing in itself,’ ‘transcendental object,’ and ‘noumenon in the negative sense.’ In other words, while leaving definitions crafted for scientific and other legitimate uses inviolate, Kant’s psychologism purports to render physical and metaphysical accounts of consciousness-independent reality complete nonstarters by demonstrating that we lack the representational wherewithal even to ask the questions such accounts are meant to address.9 One final remark on psychologism before proceeding. Normally, one approaches a work of philosophy by deploying one’s best understanding of the points at issue while striving to suspend one’s convictions regarding them as far possible so as to retain an open mind. By contrast, a book on psychologism requires one to suspend not only one’s convictions but the understanding that informs them. This is because psychologism’s primary purpose is to identify any and all psychological ingredients essential to the meaning of notions that, in language, are entirely free of any tincture of the psychological, at least when strictly defined (e.g. for scientific purposes). There therefore can be no expectation that the psychological contents adduced as essential to familiar, ostensibly non-psychological notions will themselves be familiar. Indeed, if, as with Kant, the psychologism is a priori, how could they be familiar if experience is in principle incapable of acquainting us with them?10 Suspending understanding as well as belief is certainly a hard ask, demanding an extreme degree of open-mindedness. But it seems to me well worth the effort if the meanings disclosed by psychologistic explication really do, as Kant claimed, amount to a revolution in philosophy no less profound than the revolutions wrought by Euclidean geometry and Newtonian physics in those disciplines.

Prospective I share the urge to present early modern philosophers in a manner that makes their thought not only accessible to non-specialists but relevant to contemporary philosophical concerns as well, and, in Kant’s case especially, essential to take into account even in the philosophizing engaged in by non-philosophers, particularly scientists. It is only insofar as this means starting from the premise that Kant set out to remove psychology from epistemology that I regard such efforts as misbegotten. My thesis that, far from removing it, he extended its scope to include not only

12  Introduction objective understanding and knowledge of nature but logic and mathematics as well renders Kantianism irreconcilable with most of today’s leading schools of philosophy, especially analytic philosophy, the one long regnant among anglophones. But matters are quite different where contemporary science is concerned, particularly as it relates to the mindbrain. Naturalistic theories like Kant’s and those developed by his British empiricist forebears were intentionally crafted to leave open a place for future science on which philosophy can never impinge. Where conscious mind is concerned, this meant budgeting for the possibility that science might one day be able to explain it fully, thereby proving it once and for all to be as wholly natural a phenomenon as lactation. I at least find nothing in their philosophies, particularly their purely psychological phases, that is incompatible with any of the major scientific advances they could not foresee, including evolutionary theory, ethology, neuroscience, and systems chemistry, as well as relativity, quantum theory, and the sciences they underpin. I shall therefore take the liberty of interleaving new scientific data and conceptual developments into my presentation of early modern psychologisms with a view to showing how easily and extensively they dovetail with scientific approaches to human and animal conscious minds and their place in nature at large.11 What this means in practice is the incorporation of several current assumptions.12 Obviously, the one that looms largest is evolution. Human minds are products of brains that give phenotypic expression to a genotype molded over the course of untold generations of Homo sapiens, preceded by untold generations of earlier Homo species, Australopithecus, and so on back to the last common ancestor(s) of all conscious species, extinct as well as extant. This is the phylogeny that has produced each of us, body and soul: Homo sapiens phylogeny atop Homo phylogeny atop Australopithecus phylogeny and so on, through all clades back to earth’s original conscious creatures. The second assumption is that our phylogeny, like that of every other creature, is the product of sheer serendipity. Our phenotypes are assembled via genotypes that have in each generation been subject to the lottery of natural selection, genetic drift, mutation, epigenetics, and various other genome-altering factors, as well as circumstances such as asteroids striking the earth, the oxygen content of the atmosphere plummeting or soaring because some distantly related organism went extinct when it did, a predator that was on course to drive one of our ancestor species extinct that fell victim to a newly mutated virus that might not have mutated when or how it did, etc. The third assumption is that some of the roads serendipitously not taken by evolution—or that have been or will be taken in extraterrestrial biospheres, or that futuristic genetic (or AI or other yet to be conceived modes of) engineering might produce—could have eventuated in brains, or organs of consciousness analogous to brains, that differ from ours

Introduction 13 in one or more faculties of consciousness but are otherwise exactly like ours. For example, just as the blind differ from the rest of us only in lacking sight, creatures otherwise like us might have evolved to respond to electromagnetic radiation not with sensations of color and light but sensations of a kind completely unknown in humans or any other terrestrial creature. Or just as those who have a defective hippocampus are deficient in memory relative to the rest of us, we may be lacking psychological capacities that might have evolved in creatures otherwise like us had evolution taken a relevantly different course. Or just as cutting edge mathematics is so far beyond the reach of most of humanity that trying to teach it to the mathematically challenged is as futile as trying to teach it to chimpanzees, creatures might have evolved whose cognitive reach was so immeasurably superior to ours that trying to teach what they know even to a Gauss or Einstein would have been just as futile. And while the current state of science prevents us from probing the full mental potential of evolution, genetic engineering, or anything comparable, both philosophers and scientists still need to budget for the possibility that it may be vastly greater than we are at present capable of appreciating. A further assumption is perhaps more philosophical than scientific, but still seems to me to bind contemporary mind science as much as it does philosophy of mind: consciousness is an expression purely of bodily chemistry. This assumption may one day be proven false, and some are convinced that it already has been. But as most cognoscenti accept it, I believe that our default position must be that consciousness is ­chemistry—nothing more, nothing less. At its heart is the notion of a neural correlate of consciousness. Although neural correlates can be characterized in various ways, my purposes are best served by doing so as follows: for any consciousness C at time t, given everything going on in the brain at t, there is a subset N, the existence of which is indistinguishable from C’s own existence. So construed, one must be careful never to confuse N with the neural cause of C since much that is involved in the causation of C may have no correlate in consciousness at all. Moreover, precisely because C and N need to be understood as one and the same existence, the causes and effects of N, outside as well as inside the brain, are ipso facto causes and effects of C and vice versa (on the principle that only what is distinct in existence can be causally distinguished or causally related). How then are C and N even distinct at all? The answer follows from two final assumptions: nothing can be known of N directly from C, and nothing can be known of C directly from N. However perfect one’s introspective acquaintance with the contents of C may be, there is quite simply zero neurological information about N or its causes secreted in it, and so none to be discovered there.13 Hence, no one wholly ignorant of the insides of skulls can form the least notion from C what (if anything) is in them, what (if anything) their insides do, or whether their insides have

14  Introduction anything at all to do with the particular sensations and other conscious representations one is having or with being conscious generally. Conversely, however perfect one’s scientific acquaintance with N might be, there is zero non-neurological psychological information secreted there, and so none to be discovered. For example, neuroscientists of an alien species who lacked sensations of color and light could no more divine from the study of the brains of sighted terrestrial creatures like ourselves what, e.g.,  a sunset looks like than could someone blind from birth. Instead of considering N and C in isolation, one must therefore rely on a mixture of theory, observation, experiment, and subjective reports to discover that N is correlated to any state of consciousness at all, much less to C in particular, or that C is correlated to anything non-conscious, much less neuronal, and in particular to N. Thus, not only are C and N distinct, they are fully distinct realities, one conscious and the other neural. A purely physical existence that is at the same time irreducible to physical reality: that is the mystery of consciousness as I shall understand it here. It is a mystery with multiple convolutions: • Is the information consciousness provides about, e.g.,  the visible world the same, or even remotely like, the physical world as it mindindependently is—even if only at the macroscopic scale of the objects that stimulate the human optical system? Isn’t it possible that visual, tactual, and all other sensory information, as well as time- and spaceconsciousness, are phylogenetic accidents whose place might have been taken, had evolution followed a different course, by conscious representations that share no features whatsoever in common with any of the visual, etc., representations we have? • No one doubts that our brains evolved the way they did because they enhanced the survival prospects of our ancestors. But who is to say that whatever survival value conscious representations had for our ancestors and continue to have for us today coincides with having objective, phylogeny-independent truth? After all, might, in the form of enhanced chances of survival, neither implies nor entails cognitive right.14 Nor does it seem that we could ever be certain that whatever evidence we propose to settle the matter is not itself phylogenetically relative, and therefore unsuitable for the purpose. • Since no one has ever had access to the world or to themselves, directly or via scientific means, when they were in dreamless sleep or coma, access essentially involves consciousness. So, if consciousness turns out to be in every respect a phylogenetically determinate natural phenomenon, how could we ever step outside of it to determine if the information accessed in it—neurological information included— is not always ineluctably phylogenetically relative, and so never truly unconditionally objective? • If we can never step outside our phylogenetically determined consciousness, doesn’t that mean that we can never know whether

Introduction 15 anything in our representations of mind-independent reality really is mind-independent? Indeed, if we can never step outside it, how can we even know anything we are conscious of in ourselves is not just as phylogenetically relative? Indeed, what is “the self” objectively? These are just some of the multitudinous facets of the mystery of consciousness that will be explored in this book. Here it suffices to emphasize that it concerns consciousness as a natural phenomenon as proper to sharks as teeth, to cats as whiskers, and to humans as bipedalism. Since it is the same phenomenon for us that it was for Kant and his empiricist forerunners, and since they too seem to have been prepared to concede its complete dependence on the body—and more particularly the brain— should science ever succeed in proving it, there thus seems to be nothing to prevent one from subjecting their psychological philosophies to the same scientific assumptions that prevail today. Of course, interleaving Darwin et  al. into the presentation of early modern psychological philosophy is not history of philosophy in the strict sense. Inevitably, a certain amount of revision is required, as when an architect is tasked to refashion a great work of the past for the present. But just as I. M. Pei’s Louvre revitalized the original without effacing its original conception and purpose, an updating of early modern psychological philosophy can likewise remain true to its originals in the very process of transforming them for the present—if done right. This is not architecture, and I am not Pei. But for me, “doing it right” means having such complete command of the original sources and their historical context as to be able, with high confidence, to narrow the range of possibilities down so far that to claim the updated presentation for one’s own would be as much philosophical plagiarism as publishing another’s words as one’s own is the literary sort. If successful, my updating of Kant will result in a presentation of his philosophy that is as genuinely revolutionary today as when he originally compared his démarche to Copernicus. The same cannot be said of updatings premised on the consensus interpretation since removing psychology from epistemology represents not a revolution but a restoration of the pre-Humean non-psychologistic ancien régime, whereupon nothing remains to distinguish Kant’s “critical philosophy” from pre-Humean “dogmatic philosophy” except the excision of metaphysically grounded theological commitments of the kind one finds in Descartes, Leibniz, and Berkeley. Updatings of Kant premised on the consensus interpretation cannot change this. They may provide a means of explaining conscious representation logically by utilizing post-Fregean mathematical logic; grammatically by applying philosophy of language and/or Chomskyan linguistics; mathematically by adapting something like Giulio Tononi’s integrated information theory of consciousness; or scientifically by drawing on anything from a purely neurophysiological account of consciousness of the sort pioneered by Gerald Edelman to Roger Penrose’s quantum

16  Introduction theory of consciousness. But what they can never do is yield a genuinely Copernican scale revolution of the kind that results when representations fundamental to logic, language, mathematics, and science are explicated through consciousness. Since that is precisely what Kant’s psychologism does, arguably without succumbing to psychologistic fallacy, it is thus the only way of reading him I know that is truly capable of translating his revolutionary project into the context of our time.

Notes 1. For this reason, the book’s discussions of post-Kantian philosophy center on what I regard as the most formidable challenges to a priori psychologism: Wittgenstein’s conventionalism above all, but also Hegelian logic, analytic philosophy as exemplified by Strawsonian connectionism, and the Heidegger of Being and Time. 2. “No Contest,” June 11, 2005. 3. The problem, as formulated by Kant, is “how are synthetic a priori judgments possible?” The key text regarding Kant’s answer—a priori psychologism—is PFM 258–9: see analyses in KEUU chapter 2-B and KAIM chapter 2-D & -E (though the passage itself along with a first pass at its analysis can be found in the chapter’s introductory pages). 4. As with Kant, Hume scholars tend to emphasize the epistemological dimension at the expense of the psychological, which, far from being deemed Hume’s chief legacy, tends to be either ignored or disparaged. 5. The key text is CPR B128–9. See discussion in KAIM ch 11-A, especially the concluding paragraph on p. 297. In the case of pure concepts and other pure representations, “pure” simply means purely a priori, i.e. including no admixture of the empirical whatsoever in their content. Kant’s conception of the purity of conscious representations will be explained more fully in part II and is included in the glossary at the end of the book. 6. I refer those with scholarly objections to my attribution of this or that psychologistic thesis to Kant or his empiricist precursors to my scholarly writings. The principal underpinnings of the chapters in this book are as follows: Chapter 1: KEUU chapters 1 & 4 and KAIM chapters 1–2, 12, & conclusion sect. A Chapter 2: KEUU chapters 5–6 Chapter 3: KEUU chapter 14 Chapter 4: KEUU part III, Hume’s Theory of Consciousness [HTC] Chapter 5: KEUU chapters 2–4 & 18-A, KAIM chapter 2, and HTC chapter 3-C Chapter 6: KAIM chapters 5 & 9 and conclusion sect. B Chapter 7: KAIM chapters 9–11 Chapter 8: KEUU chapter 3, KAIM chapters 3–4, 7, 12, & 17, Kant’s Model of the Mind Chapter 9: KAIM chapters 5–6, 15 Chapter 10: KAIM chapters 3–4, 8, 13–14, and 16–18 I will provide further references as occasion warrants. 7. Hegel did provide two extended summations of Kant—the one in Faith and Knowledge remains, in my view, state of the art—and included numerous remarks on Kant in his major works, above all the Science of Logic, which will be discussed in this book’s conclusion. One contemporary scholar who avoids the ‘p’-word but gets the essentials of Kant’s psychologism right is

Introduction 17 Béatrice Longuenesse (I am the proud dedicatee of the original French version of her principal monograph on Kant). Otherwise, interpreters who relate my interpretation to those of Patricia Kitcher, Andrew Brook, Lorne Falkenstein, Scott Edgar, and similarly minded scholars seem to me to do so only because they overlook or disregard the fundamental differences between psychologism and psychology, even sometimes using the terms as if they were interchangeable. To repeat: psychology as such (“theory of ideas”) is concerned only with whether and how a representation comes to be in us (empirically, innately, etc.), not how the representation itself comes to be, and so is fully compatible with the prevailing anti-psychologistic consensus that leaves Kant’s philosophy barely, if it all, distinguishable from the kind of Rationalist platonism one finds in Leibniz and more recent variants such as P. F. Strawson’s connectionist interpretation (also discussed in the conclusion). The a priori psychologism I attribute to Kant, by contrast, is as fundamentally incompatible with platonism of every kind as it is with purely empirical psychologisms like Hume’s. 8. This view can be traced back at least to the Fourth and Fifth of Descartes’s Meditations on First Philosophy and continues to inform much analytically oriented philosophy today. See chapter 1 note §11. 9. I would trace psychologism’s reputation as a fallacy to the justifiably hostile reaction to any attempt to incorporate the psychological into definitions of non-psychological notions. Hume and Kant, by contrast, were perfectly happy to seal off definitions of notions that, in language, are free of any tincture of psychological content from the implications of their psychologistic explication and limit the role of the latter to preventing non-psychological concepts that owe essential elements of their content to consciousness from being employed in consciousness-excluding contexts. 10. For an example, see the consideration of space in the third paragraph of the introduction to part II. 11. For this reason, most of my references to secondary literature will be from non-specialist works authored by scientists. Readers interested in my comments on history of philosophy secondary literature should consult my history of philosophy monographs and papers, particularly those referenced in note §6 earlier in the introduction. 12. The formulations provided here are purely introductory and should be taken as provisional. Each assumption will be revisited, refined, and elaborated in chapter 2 and beyond. 13. Are neuroscientists exceptions? Science is never a purely psychological matter of the contents of consciousness. Most obviously, science is impossible without language, language depends on convention, and convention is impossible in the absence of considerable social infrastructure (culture) and common human lifeways. Since this is just to say that science can never be reduced to the contents of the individual isolated psyche but is instead irreducibly sociological, the consciousness of scientists, in and of themselves, can tell them no more about the neural correlates of consciousness than anyone else’s consciousness tells them. 14. Scientific evidence that, evolutionarily speaking, right may be less mighty than wrong, both in scientific theory and in fact, has been adduced by the psychologist Donald D. Hoffman. See “Conscious realism and the mind-body problem,” Mind & Matter, Vol. 6 (1), pp. 87–121, and his April 21, 2016, interview with Amanda Gefter in Quantum Magazine, “The evolutionary argument against reality.”

1 Breaking the Mold

I argued in the introduction that Kant’s claim that his philosophy effects as revolutionary a transformation of its discipline as the revolution wrought by Copernicus in astronomy cannot be sustained unless it is interpreted as a priori psychologism. The thesis of the present chapter is that situating Kant’s a priori psychologism in relation to the conceptual landscape in which contemporary philosophers and scientists of mind operate requires that landscape’s expansion to include a place for a physicalism that nevertheless treats consciousness as essential to intelligence. I trace its omission to the well-nigh universal assumption that the questions of consciousness and intelligence are entirely separate, so that the answer to one neither implies nor entails an answer to the other—an assumption rooted, it seems to me, in a predilection to conceive consciousness in purely subjective terms of qualia, feeling, emotion, volitional autonomy, and similar such features that contrast sharply with objective cognitive intelligence of the kind epitomized by logic, computation, and applied mathematics. What I shall try to show here, and in greater detail in Part II, is that Kant’s a priori psychologism necessitates a reconceptualization of consciousness and its subjectivity so radical and transformative as to render logic and mathematics emblematic expressions of consciousness and the quintessentially subjective.

A Priori Psychologism in Contemporary Context This book was written not to challenge the reigning consensus regarding the mind but as a guide through a radically divergent conceptual landscape that theorists of mind are otherwise unlikely to confront. The landscape familiar to scientists and philosophers of mind is best approached through physicalism. Physicalism, as I understand it, is the view that everything going on in the mind, consciously or not, is wholly dependent on what is happening in the brain. Optimistic scientifically, it implies that neuroscience is in principle able to fully explain the mind. Antiphysicalism, by contrast, supposes that not everything happening in the mind is dependent on what is happening in the brain. It can take any of three

Breaking the Mold 19 forms, according to whether the mental that is posited to exist independently of the physical (1) constitutes a distinct, autonomous portion of reality (dualism), (2) itself underlies the physical (idealism), or (3) has to be explained at a third, more fundamental level of reality that likewise underlies physical reality (neutral monism). Since antiphysicalist views imply that the mind is in principle not fully explicable by science, they are rightly regarded by scientists and the scientifically minded as a last resort. Accordingly, the burden of proof falls on the antiphysicalist. Why would one feel justified, much less compelled, to abandon science in favor of antiphysicalism? The answer one most commonly encounters is science’s supposed neglect of the specifically subjective dimension of consciousness, variously described as lived experience, qualia, the inward, the private, the ineffable, etc. Hume, for example, explained remembering as a special “action of the mind in the meditation, that certain je-ne-sais-quoi of which ’tis impossible to give any definition or description, but which everyone sufficiently understands.”1 Each person knows this dimension of his or her various experiences purely from his or her own case, and knows it without any need for definition or ostension. Indeed, subjective experience does not lend itself to these: the qualia of savoring a fine wine, being startled by a bang, or trying to walk a straight line at the behest of the police can neither be named nor pointed to as the agents, actions, properties, and relations involved can so easily be. We can refer to them only because their privateness and utter ineffability single them out from the rest of experience, as indicated by formulas like “that certain je-ne-sais-quoi” or “what it’s like to. . . .” Our understanding of qualia is wholly subjective, a matter purely of inward experience, the private mental realm of each self or subject as such, exclusive of every other. Even a mind-reader could not gain access to the qualia internal to another’s mental life since the action of reading it would be the locus of the mind-reader’s qualia, not the qualia experienced by the mind being read. For, on this conception, the subjective is in principle unobservable by another. As this includes scientific observation and experiment, there is and can be no scientific evidence for the existence of qualia except via the testimony of conscious subjects. Yet, because to the subjects reporting it that evidence is so strong as to be undeniable, the exclusion of independent scientific evidence of qualia does not detract one iota from subjects’ certitude regarding their reality. We thus seem obliged to admit that reality includes more than language and ostension can signal, conceptual analysis (including philosophy) can reveal, or science can investigate without depending on subjective testimony—with the consequence that antiphysicalism must be true. Physicalists typically respond either by denying subjectivity altogether or by denying that it is inexplicable scientifically. Denialists of the first stripe argue that notions of subjectivity are relics of folk-psychological thinking, the result of being misled by language (“a shadow cast by

20  Breaking the Mold grammar”), or some other variety of subreption. Perhaps the most potent portrayal of this kind targets the temporal dimension of subjectivity. Our subjectivity exists always in the temporal present and consists in experiencing time’s change—the present passing into and forever lost to the past, former futures becoming present, and then themselves vanishing into the past. New perceptions perpetually replace their predecessors to be displaced in turn once their moment in the subjective sun of consciousness has passed. Thoughts form successive perceptions into connected trains; melodies, rhythms, and speech have distinctive temporal profiles; emotions flare up suddenly and slowly subside; etc. Relativity theory, however, tells us that all this is impossible. Everything that exists exists in a mathematically defined four-dimensional spacetime in which the temporal dimension is just as fixed and permanent as the spatial dimensions. So, while there is no denying the reality of the subjective illusion that time flows, it has no objective reality, no truth, where time itself is concerned, and so cannot be accorded a separate, independent ontological status alongside the physical. And since the other varieties of subjectivity mentioned so far are inextricably bound up with subjective temporality, they must be no less illusory. Even setting aside the illusory aspect of subjectivity, physicalists deny that the mind can be in any way scientifically inexplicable. They target the claim that there can never come a day when scientists have sufficient knowledge of the brain to fully explain subjectivity.2 Physicalists attribute doubts in this regard to ignorance. Since we have no clue today what in the brain is responsible for the indescribable “je-ne-sais-quoi” qualia of subjective experience, much less how and why that particular neural system evolved, we cannot today even begin to guess the principles of its operation. But how can one presume to say the same of future scientists, much less claim now that it is impossible ever to know? Ignorance alone is not a basis for doubting, much less denying, the capacity of a science able to fully explain the brain to fully explain subjectivity as well. Thus, the consensus among physicalists is that the doubts of antiphysicalists have little if any validity. The physicalist/antiphysicalist divide just described is intended to capture the full range of positions one finds among philosophers and scientists of mind today. But what about Kant? Many would rank him with the antis if only because he labeled his philosophy ‘idealism.’ Yet Kant was very careful to distinguish his transcendental idealism from the empirical sort that privileges the subjective mental inner over the objective physical outer, including both the problematic idealism he ascribed to dualists like Descartes and the dogmatic idealism he attributed to anti-materialist monists like Berkeley. Because the dualism of “mental inner” vs. “physical outer,” far from being ontological, is, on Kant’s reckoning, fully as psychological as the dualism of tactual vs. visual access to reality, his ontology limits reality to physical substances and their properties. The

Breaking the Mold 21 mental is causally and in every other way fully determined by its physical underpinnings, and so is in principle fully explicable by science. In particular, nothing inherently distinguishes the temporality of the inner from the temporality of the outer so that the former is in principle explicable by science as fully and adequately as the latter. Indeed, I see no reason to think that Kant would have resisted the verdict of relativity theory that completely excludes subjective time from any final ontological accounting of nature. To be sure, this by itself precludes only substance dualism, not property dualism: the thesis that there are non-physical features of physical objects that are fully as real as their physical features.3 But since, again, the sole difference between the mental and physical in Kant’s philosophy is a psychological difference in the way they are experienced—via internal sense as distinct from external sense—this no more suffices for a fundamental ontological division of properties than the difference between visually and tactually sensing spatial properties does. Clearly, therefore, if Kant’s philosophy of mind falls anywhere on the physicalism–antiphysicalism spectrum, it is on the physicalist side. How can a self-described idealist like Kant possibly have espoused physicalism? The answer lies in his conception of subjectivity. As understood by physicalists and antiphysicalists alike, subjectivity is completely incidental to intelligence, contributing nothing essential to it, least of all the higher sort that is impossible apart from logically determinate thought. Qualia, for example, may be viewed as part and parcel of what makes a mind conscious; but intelligent? The famous query of ­qualia-ists—“what is it like to be a bat?”—presumes qualia to be present in minds lacking anything approaching human intelligence. Indeed, there is no reason to limit qualia to mammals: why not more primitive creatures (“what is it like to be a fly?”), or, for that matter, anything in any measure conscious—as panpsychists might wonder what it’s like to be a proton? The same non-intellectual character applies to everything else that has so far been mentioned as subjective: lived experience, Hume’s “I know not what,” subjective temporality, etc. Conversely, I know of no one, physicalist or not, who would limit intelligence to conscious minds on the ground that, in the absence of consciousness, intelligence would be lacking some essential ingredient. On the contrary, it is more and more accepted that intelligence needs to be understood in abstraction not only from the human psyche but from biology generally, i.e. without specifying a substrate of any kind (material, immaterial, or neutral monistic), so that a genuine theory of intelligence would be limited to mathematical logic, information theory, computation theory, complexity theory, and other relevant areas of mathematics. In short, there are two things about which non-elimitivist physicalists and antiphysicalists seem generally agreed: (1) intelligence, particularly the kind that depends on logically structured propositional representation, is fully independent of

22  Breaking the Mold consciousness, and (2) analyzing or explaining what is most distinctive of the one in no measure analyzes or explains what is most distinctive of the other. Kant, by contrast, not only denied that the subjective dimension of consciousness makes no contribution to intelligence, but deemed its role so vital that without it thoughts would admit of no properly logical combination or relation at all, leaving only the most primitive, least intelligent, logically unstructured species of mentation like Pavlovian association. This contribution can be found in the form of intellectual self-­consciousness—the ‘I think’ of Descartes’s “I think therefore I am”— which Kant termed the analytic unity of apperception (AUA). (For definitions of this and other Kantian terms see the glossary at the end of the book.) According to Kant, AUA psychologistically explicates the logical universality that makes possible genuine logical universals, i.e. the conceptual representations that constitute the building blocks of propositional thought. This universal self-consciousness in turn presupposes another form of self-consciousness, one that consequently must already be in place in purely sensible consciousness ahead of all thought, and so prior to and independently of propositional representation and everything it involves (universals, predication, affirmation, conditional assertion, quantifiers, etc.). He termed it the synthetic unity of apperception (SUA) and traced its possibility in turn to pure intuitions of sensibility exemplified by the form of juxtaposition whereby alone sense impressions can become spatial representations and the form of succession whereby alone internal impressions can become temporal representations. Kant’s analysis of intelligence in terms of self-consciousness thus proceeds through the following stages: given a pure sensibility equipped with intuitions of space and time, SUA is produced; given SUA, AUA becomes possible; given AUA, representations can be produced in consciousness that are fit for the kinds of logical combination constitutive of propositional representation; and given propositional representations, consciousness can attain the myriad forms of discursive intelligence and reason familiar to us, including science and mathematics. There is, of course, a great deal more to Kant’s a priori psychologism. Consciousness must also incorporate logical forms for producing propositions because AUA by itself suffices to yield only the raw materials for logically structured representation but not the forms requisite to combine them into actual propositions or combine the resulting propositions into complex propositions, inferences, narratives, etc. From these subjective functions,4 Kant’s psychologistic anatomy of the intelligent mind proceeds to showing (1) how, from the various forms of consciousness (SUA, AUA, and logical forms), additional equally subjective conscious representations can be produced—most notably, the categories (fundamental concepts of objects)—and (2) that these representations are essential to all extra-logical modes of intelligence, from cognition of nature

Breaking the Mold 23 in ordinary experience and science to mathematics, morality, and even estimations of beauty and ugliness in nature and art.5 Thus, for Kant, the task of the psychologistic philosopher begins with identifying the “underlying a priori principles of the possibility of thought itself” with an eye to showing “how the faculty of thought itself is possible,”6 and culminates in establishing that objectivity of every kind springs entirely from the subjective features of consciousness that make propositional representation logically possible—features that, in the final analysis, boil down to that most quintessentially subjective representation of all, the ‘I think.’ Clearly, a psychologistic conception of subjectivity centered on the grounding of logic in the ‘I think’ cannot be slotted into today’s generally accepted scheme opposing physicalism to the three types of antiphysicalism. All varieties of subjectivity not entailed or enabled by the Cartesian ‘I think’ (qualia, subjective temporality, lived experience, et  al.) Kant treated exactly as physicalism does, even going so far as to maintain that only a priori psychologism can ensure their complete physical explicability. At the same time, because the subjectivity constituted by apperception (SUA/AUA) becomes part of the explanation of the physical, it cannot itself be explained physically on pain of circularity. Thus, from Kant’s psychologistic point of view, the physicalist/antiphysicalist duality defines a conceptual landscape with no place in it for a subjectivitydefining consciousness that is indispensable to intelligence generally and to logic, mathematics, and physics in particular. And surely any scheme with no place in it for “the supreme principle” of Kant’s philosophy can hardly be deemed adequate.7

Reasons for Kant’s Exclusion What accounts for this failure? Perhaps the best answer is that philosophy remains a discipline dominated by schools. Today’s dominant school, certainly among anglophones, is commonly denominated analytic philosophy. Among its reigning dogmas are two that are particularly deleterious when applied, as almost invariably happens, to the interpretation of Kant: (1) the conviction that philosophical analysis should everywhere conform to, and when possible be guided by, mathematical logic, and (2) the conviction that psychologism is a fallacy. I shall briefly examine each in turn. Modern mathematical logic, although anticipated to some extent by Leibniz, is principally the creation of nineteenth-century mathematicians, above all Gottlob Frege (1848–1925). Frege devised a concept-script (Begriffschrift) to aid him in his endeavor to show that mathematics can be wholly derived from purely logical foundations, a project known as logicism. Along the way, Frege developed new techniques of analysis that he applied not only to mathematics but to vernacular language. Although the logicist project crumbled in the face of Gödel’s incompleteness proofs,

24  Breaking the Mold many twentieth-century linguistic analysts picked up where Frege left off. And while some, Ludwig Wittgenstein most notably, eventually soured on analysis inspired by mathematical logic and devised a new “ordinary language” approach to language and the foundations of mathematics, the vast majority of post-logicist analytic philosophers have adopted Fregean analysis as their template in approaching philosophical topics of every kind. Analytic philosophers tend to find little to like in the history of philosophy. The prevailing attitude is well captured by Wittgenstein’s comparison of philosophy prior and subsequent to the advent of mathematical logic to the supplanting of alchemy by chemistry.8 Before mathematical logic, there was, for all intents and purposes, Aristotle; and while Aristotelian logic may have represented a huge advance over what preceded it, it has long since come to be regarded as obsolete. Unfortunately for Kant’s post-Fregean reputation, he embraced Aristotelian logic wholeheartedly and accorded it a central place in his philosophizing. It therefore comes as no surprise that analytically trained philosophers tend to view Kant’s transcendental philosophy as a once formidable structure long since reduced to ruin, fit only for piecemeal salvage. As remarked in the introduction, salvage operations tend to take the form of more or less extreme makeovers in which contemporary analytic methods informed by mathematical logic are substituted for Kant’s own. Yet, all have in common a failure to ask a crucial question that, depending on the answer, could put the whole enterprise in doubt. The question is whether Kant’s conception of logic and what differentiates it from mathematics implies that post-Fregean mathematical logic is mathematics rather than logic. If the answer is yes, as I shall argue in part II, the alchemy/chemistry comparison fails, and there is nothing suitable for mathematical-logical makeover in Kant. To be sure, his way of distinguishing logic from mathematics is quite unlike the ways favored by analytic philosophers. Kant’s is neither logical, mathematical, nor linguistic but psychologistic and, in the final analysis, a matter of the different subjective representations each discipline presupposes.9 So, if—a big ‘if’ to be sure—his distinction between logic and mathematics has merit, then we must be ready to admit that, contrary to the alchemy/chemistry comparison, post-Fregean mathematical logic is to Aristotelian logic what arithmetic is to psychological philosophy: a different discipline entirely. Aristotelian logic is not the only instance of Kant being unfairly convicted of giving hostages to fortune by being supposed to have claimed that a particular scientific discipline could never develop beyond its state in his day. He is, for example, regularly labeled a dogmatic Euclidean for declaring Euclidean geometry to be a priori truth, valid necessarily and universally, and thereby seeming to preclude non-Euclidean geometry. Yet, there is nothing in Kant’s philosophy not fully accordant with a Riemannian geometry that subsumes Euclidean geometry within it but

Breaking the Mold 25 in no sense refutes it, which, after all, is the prevailing view today. The reason he nevertheless so often gets mischaracterized is that his account of space is commonly supposed to treat it as intrinsically Euclidean. This too, however, mistakes Kant’s thesis that the mind is equipped with the representation of a fully geometrizable space, sufficient for the purposes of geometry howsoever far it may progress, with the affirmation of a representation of a fully geometrized space of a determinate (viz. Euclidean) kind.10 Kant had no crystal ball with which to anticipate what Gauss, Lobachevski, Bolyai, Riemann, and their successors would do, much less how Einstein would apply non-Euclidean geometry to gravity. But he did not need one. Space, for Kant, can be of interest philosophically only insofar as it derives from the conscious subject; but it takes actual mathematical definitions, axioms, and postulates before the fully geometrizable space furnished by the subject can be converted into a representation of Euclidean or any other kind of properly geometrical space. Indeed, so little did Kant view subjective space as Euclidean that he analyzed it as something that all of mathematics, not just geometry, depends on, so that his subjective space is not just fully geometrizable but fully mathematizable as well. As such, the space furnished by sensibility is not only geometrically but in all mathematical regards indeterminate. Indeterminate, yet fully determinable: a space that can curve in non-Euclidean as well as Euclidean ways, a space of any number of dimensions rather than maximally three; a space able to accommodate Calabi-Yau manifolds; a space without which even mathematical logic would be impossible. This, at any rate, is what I shall try to show in Chapter 9, which, if I am correct, means that Kant was dogmatically Euclidean neither in his philosophy of mathematics nor in his philosophy of nature. A third case of falsely supposing Kant to have given hostages to fortune is the notion that he was a dogmatic Newtonian who analyzed nature and its laws in such a way as to require space, time, causality, etc. to be understood as classical physics did prior to Einstein and Heisenberg. The truth is that the understanding of nature and natural law that Kant traced back to the subject’s apperception can extend to physical theories of every kind, and includes nothing not perfectly compatible with the reality of spacetime, gravity waves, quantum entanglement, or any other scientifically validated reality. This is because Kant’s transcendental theory deals with nature at a more objective and fundamental level than any physical theory does or can. Thus, just as his transcendental theory of space leaves the geometry of pure space completely indeterminate and has nothing to say about the geometry of physical space, I will show in Chapter 10 that his transcendental (i.e. a priori psychologistic) theory of nature has nothing to say about which physical principles—Newtonian, Einsteinian, quantum mechanical, or any other—apply in the world evolved humans experience through their phylogenetically determined sense organs.

26  Breaking the Mold Behind these and similar misrepresentations of Kant lies a deeper one: the notion that Kant was anti-psychologistic. Anti-psychologism is at least as basic to analytic philosophy as mathematical logic, just as it is to many other schools of philosophy. Frege cut his philosophical teeth as a critic of psychologism, and his successors continue to have no truck with the notion that a concept used in logic, mathematics, linguistics, or science can have anything in its analysis that derives from the conscious psyche. Indeed, analytic philosophers (but by no means only them) regard psychologism not merely as false but as fallacious: illicitly substituting subjective psychology in place of the objectivity proper to everything that belongs, or can belong, to logic, mathematics, science, language, etc. It is consequently no surprise that well-intentioned analytically oriented philosophers place a premium on devising non-psychologistic ways of construing Kant’s attribution to space, time, substance and accident, cause and effect, number, propositional form, and much else besides to an origin in subjective consciousness. Nativism is the route typically taken, the assumption being that a representation that originates a priori in the subject is necessarily an innate representation. However, since being innate to the mind means that the conscious subject is determined conformably to the representation and not vice versa, nativism, contrary to how it may at first seem, is a way of denying psychologism, not affirming it. For to say that the conscious subject is innately endowed with representations of space, time, cause and effect, number, et al. is tantamount to saying that its conscious representations of them contribute nothing essential to their content; and since this in turn implies their contents’ intrinsic independence of consciousness, they count as mind-independently objective. In other words, nativism splits Kant’s question about the origin of representations in the subject into two quite separate questions: how the representation comes to be in us and how the representation itself comes to be. If the answer to the second question is non-psychologistic, then the answer to the first may be of psychological interest, but can have no properly logical, linguistic, mathematical, or scientific interest where the representation itself is concerned.11 Hence, innatist readings of Kant’s apriorism oblige one to suppose that he too would brand as fallacious any analysis of space, time, etc. that treats any of their essential contents as originating in subjective consciousness. This then is how the anti-psychologistic reading of Kant favored especially by analytic philosophers was spawned, and continues to (mis-)inform the views of philosophers and scientists alike today. Nor is that all. Since objectivism requires that the origin of the representations supposed to be innate has to be completely independent of the subject (on pain of psychologistic fallacy), their origin can only be explained in either of two ways. One involves positing a special realm of meanings, contents of thought, or ideas that is real and objective in the same way the physical realm is, or more so. The affirmation of the objective

Breaking the Mold 27 reality of representational content is often termed platonism because Plato originated the notion of a “heaven” where abstract idea archetypes exist on a separate, higher plane of reality than their physical and mental-representational ectypes. Seventeenth-century Rationalists such as Descartes, Spinoza, Leibniz, and Malebranche accepted the essentials of Plato’s account while shifting the scene of Platonic archetypes to the infinite intellect of the deity. And realism of this kind remains a widely held view today, not least among mathematicians, albeit typically with the cosmogonic and theological elements expunged. The main contemporary alternative to platonism is the view that a priori representations are nothing other than conventionally grounded bedrocks of language, which I shall henceforth call conventionalism. Conventionalism preserves the independence of a priori concepts vis à vis the individual isolated subjective psyche without hypostatizing their objectivity beyond their normative role in vernacular discourse and related social practices. Here concepts are best thought of as socially rather than psychologically innate in that we are born into societies with established bodies of linguistic and other conventions. Although most closely associated with the later Wittgenstein and others in the ordinary language tradition, the view is shared by many social scientists and psychologists as well. Since platonism and conventionalism exhaust the options that remain once anti-psychologism is adopted, charitably minded analytic philosophers inevitably try to situate Kant’s apriorism in one or the other of these traditions. They try but without much success. This, in my view, is because Kant’s psychologistic endeavor to explain intelligence in terms of consciousness breaks all existing molds so completely and fundamentally as to fall entirely off the grid defined not only by analytic philosophy but by contemporary philosophy and science of mind generally. Not recognizing this helps explain why instead of finding attempts to revise the conceptual landscape to accommodate Kant, we find only attempts to revise Kant to fit the existing conceptual landscape.

Subjective Origin of A Priori Representations ≠ Innatism Kant’s a priori psychologism takes its start from a fundamental rethink of the ‘I think’ of Descartes’s cogito ergo sum—“I think therefore I am.” While following Descartes in excluding all empirical content from its meaning, Kant went further. He stripped this purely intellectual self-­ consciousness of both the objective metaphysical content Descartes ascribed to it in his proof of the soul’s immateriality (substance dualism) and the unique epistemological significance he accorded to it as the touchstone of objective certainty generally (foundationalism). What remained is a core of purely logical meaning: the ‘I think’ as analytic unity of apperception (AUA). As an a priori, purely logical self-consciousness, the

28  Breaking the Mold ‘I think’ counts as a highly developed, specialized form that consciousness takes in intelligent minds, not something essential to consciousness per se. The same is true of the synthetic unity of apperception (SUA) that AUA presupposes: pre-logical consciousness is not inherently determined in the manner requisite for SUA but requires that pure formal intuitions of sensibility be incorporated into it. Thus, if one is to begin at the beginning, the exposition of Kant’s notion of subjectivity must start with consciousness as such, considered prior to and independently of its determination conformably to the conditions for SUA and AUA. Kant’s account of consciousness follows neither Descartes nor Leibniz. Descartes restricted consciousness to beings capable of propositional representation, thereby excluding animals, which Kant did not. Leibniz went to the opposite extreme and espoused a panpsychic monadology that extends consciousness to all beings without exception, not excepting plants and things normally considered inanimate, which Kant also did not. Kant’s position on consciousness is instead closest to Locke’s, who supposed consciousness may reach down even to such lower animals as oysters and cockles but no further. While it may seem odd to pair a philosopher who restricted philosophy to the a priori with a “blank slate” empiricist like Locke, this is more likely due to current misunderstandings of both philosophers than genuine differences. Take Locke first. His rejection of a priori concepts and principles is frequently confounded with a rejection of all innate faculties other than a memory good enough to retain a body of experiences sufficient for learning.12 In fact, Locke was a faculties innatist who attributed to the mind whatever capacities are needed to produce all the conscious representations (sensations, feelings, desires, thoughts, etc.) it has had, will have, and could ever have, denying only that any of the representations themselves are innate. For example, although my gustatory faculty may give me the capacity to savor the flavor of a fruit that grows on a plant in a faraway galaxy, since there is nothing in the human gustatory system that in any way prefigures that sensation (i.e. no pre-formed neural correlate), it cannot be said to be present in my mind until I have actually tasted it. Similarly, although humans living prior to the last ice age no doubt had all the faculties requisite to master higher mathematics and physics, since no actual mathematical or scientific representations are prefigured in those faculties (e.g. no neural correlates of Gödelian or Einsteinian concepts), none were present in their minds. In both cases, Locke would classify the representations as acquired rather than innate, even while affirming that the complex assemblage of neuro-anatomical structures requisite to produce them are innate. Thus, although the mind may be a blank slate representationally, it is a slate of the utmost neural complexity when it comes to mental faculties. Innatists like Leibniz rejected Locke’s reasoning as turning on a distinction without a difference: mental operations, particularly complex ones, are impossible to have in capacity without being innately equipped

Breaking the Mold 29 with the representational contents requisite to exercise them. A contemporary example is the notion that the human visual system could barely function, or not function at all, without actual algorithms genetically “hard-wired” into the neural connectome. Yet, we must not suppose that Lockean blank slate empiricism involves a dogmatic denial of representational innatism. Instead, it rests on the conviction that innatism should never be our default position. In contemporary terms, this is just to say that we should always proceed on the assumption that conscious representational contents require not only genetically determined neuronal capacities but the biological consequences of protracted interaction with the environment. Only if an intensive search reveals no evidence in favor of any such empirical account and compelling evidence against are we justified, in that particular case, of abandoning empiricism in favor of a nativist explanation.13 In the case of the visual system, it means that we first need to determine if its “programming” requires a certain plasticity. Suppose, for example, that visual representation of three-dimensional space requires the systematic coordination of visual inputs with tactual spatiality. In that case, rather than having preset (innate) algorithms, the programming of the visual system might have to wait upon tactual experience to select the visual-to-tactual neuronal correlations best suited to adapting vision to touch. The systematic intersensory correlations would then be exactly like the flavor of the fruit growing in a distant galaxy: the mind’s capacity to form them would not imply that they are in any way prefigured in (“wired into”) it. That is the Lockean “blank slate” view, its sole purpose being to oblige each philosopher and scientist to remain ever mindful, as Locke put it, “how much he may be beholden to experience, improvement, and acquired notions, where he thinks he had not the least use of or help from them.”14 Like Locke, Kant repeatedly affirmed faculties innatism while at the same time vehemently denying representational innatism, particularly in the case of the a priori representations at issue in transcendental philosophy. The question is how the slate can remain blank if the representations concerned are a priori and admit of no empirical plasticity. Kant’s answer has two parts. First, since sensory data have to be experienced to be known, any consciousness of these data is ipso facto not innate but acquired. Second, a consciousness of sensory data counts as empirical only if it includes the representation of anything that in any way singles out or differentiates one datum from any other, or some from others, whether by quality, relation, pattern of occurrence, or any other mark. Thus, a consciousness of sensory data that neither singles out any sensations nor differentiates them in any way counts as both a priori (not empirical) and acquired (not innate).15 That no contemporary philosopher or scientist of mind seems to have even so much as asked whether such a consciousness of sensory data is possible, much less necessary, for human-type experience is another reason this book was written. In Kant’s philosophy of mind, a priori sensory

30  Breaking the Mold consciousness is deemed necessary to bridge the gap between purely empirical consciousness of sensory data and the a priori SUA entailed by the AUA on which the possibility of logically determinate thought, and so too intelligence, depends. It is also what, for him, marks the difference between human and non-human terrestrial minds (though in light of progress in neurobiology and ethology since his time, I doubt that he would be quite so ready to affirm such a strict divide today vis à vis, e.g., early Homo, Australopithecus, hominoids, cetaceans, therapods, etc.). Since part of my task is to show how the bridge to intelligence provided by Kantian a priori sensible consciousness is reconcilable with evolution and the other currently prevailing scientific assumptions discussed in the introduction, the best way to approach it is via the strictly empirical psychological philosophies that set the stage for, and exerted the greatest influence on, his philosophy, for these are more straightforwardly reconcilable with those assumptions.

Notes 1. T 1.3.8 ¶16. E.g. playing an episode over in one’s mind “feels” a distinctive way, unlike every other mental activity, as unique as the flavor of pineapple. 2. This is not to say that there may not be some inherent limitation that prevents humans from succeeding in scientifically explaining subjectivity. The physicalist claim is not that scientists of our species must be capable of fully explaining the mind but that science itself is in principle capable of so doing. Thus, if bioengineers someday figured out a way to remove any such limitation from the human phenotype, nothing would prevent us from fully explaining subjectivity scientifically. 3. Exponents include the neuroscientist Christof Koch and the theoretical biologist Stuart A. Kauffman, both of whom espouse panpsychism, a form of property dualism that applies universally to all matter (including plants, cells, and even elementary particles), not just human and animal minds. 4. See KAIM chapter 10 for a discussion of logical functions and their subjective character. For simplicity’s sake, I will speak only of logical forms in this book, but they should be understood as synonymous with subjective logical functions, and so as part and parcel of Kant’s a priori psychologism. 5. Kant grounded morality on a subjective idea that takes the place of deistic and other metaphysical hypotheses in grounding the moral ‘ought.’ Estimates of beauty are grounded in a subjective relation between objective faculties of understanding and imagination that themselves are grounded in forms immanent to consciousness. Kant’s analysis of cognition of nature and mathematics will be discussed in chapters 9–10. 6. MFPNS 476n and CPR Axvii. Thus Kant deemed the ‘I’ of pure self-­ consciousness the foundation of logic itself: B131 and B133–4n. The problem of the possibility of propositional thought is the focus of the inquiry Kant termed the “subjective transcendental deduction of the categories”: see KAIM introduction to part III and chapters 9 and 12. 7. CPR B136; see also A117n, B133–4n, B134–5, B137, B139, and A158/B197. Kant is, of course, not the only historical figure to fall outside the contemporary conceptual landscape. For example, the empirical psychologism of William James in which the meaning of ‘but’ is a but-feeling, ‘if’ an if-feeling, and so on is likewise excluded. Rather than being due to ignorance, however, this is because empirical psychologism regarding logic is generally held to

Breaking the Mold 31 have been thoroughly demolished by nineteenth-century critics such as Frege and Husserl and has never to my knowledge been seriously revived since. Kant’s psychologism regarding logic and mathematics was not empirical but a priori, and that is the view I am claiming ignorance has caused to be omitted from the conceptual landscape over which contemporary philosophers and scientists range. 8. See Wittgenstein’s Lectures: Cambridge 1930–32. From the Notes of John King and Desmond Lee. Edited by Desmond Lee, Chicago: University of Chicago Press, 1980, p. 21. In his later years, however, Wittgenstein would probably not have been so optimistic and instead consigned philosophy to permanent alchemical status. 9. For logic only AUA, logical forms, and the purely sensible SUA they presuppose; for mathematics these plus some of the representations made possible by them (the pure concepts of the understanding and the sensible syntheses they effect in conjunction with pure intuitions of space and time). See chapter 9 and KAIM chapters 10 and 14-E. 10. See chapters  8–9, KEUU chapter  2, and KAIM chapters  4 and 6. In addition to Kant specialists and philosophers generally, the view that Kant was dogmatically Euclidean is regularly echoed by scientists. Psychologist John O’Keefe is a case in point: “One line of argument in favour of the nativist position would flow from the demonstration of the universality of the Euclidean metric in psychological spaces. . . . [T]his was used by Kant as an argument for the synthetic a priori nature of space when it was believed that only one geometry was possible . . . The subsequent discovery of non-Euclidean geometry  .  .  . punctures the half of the Kantian position that asserts that infallible knowledge of the physical derives from our spatial representations” (“Kant and the sea-horse: An essay in the neurophilosophy of space,” in Spatial Representation, ed. Naomi Eilan, Rosaleen McCarthy, and Bill Brewer, Oxford: Oxford University Press, 1993, pp.  44, 47). Physicist Lawrence M. Krauss misconstrues Kant similarly: “For Kant, space existed in the mind, as a backdrop for all of our experience. From his perspective Euclid’s fundamental axioms of geometry were a priori necessary features of a universe in which thinking beings could live. Kant felt that these axioms were not derived from experience or experiment, for if they were, they would merely be provisional, not absolute” (Hiding in the Mirror, p. 48). 11. Though innate representations may subsequently acquire a subjective overlay in consciousness (e.g. color and feel, qualia), nothing essential to them can originate that way, only superficial incrustrations that analytic philosophical precision demands be stripped away. In this regard, analytic philosophy is heir to the early modern method of definition inaugurated in the Fourth and Fifth of Descartes’s Meditations on First Philosophy. 12. This is as true of scientists as it is of many philosophers. Steven Pinker, for example, writes that “Locke . . . tried to explain human intelligence without granting it any innate organization” (The Blank Slate: The Modern Denial of Human Nature, New York: Viking, 2002, p. 18). Edward O. Wilson also probably had Locke (among others) in mind when he wrote of the “archaic view” that “all that evolution has achieved is an exceptional ability to learn, based upon an extremely large capacity for long-term memory. A different view now prevails: the brain has a complex inherited architecture” (The Social Conquest of Earth, p. 217). 13. Recourse to nativism other than as a last resort Locke branded “lazy”: see ECHU 1.3.16 & 25. 14. ECHU 2.9.8. 15. Borrowing from natural right theory, Kant characterized representations obtained in this way as original acquisitions: Discovery, pp. 221–3.

Part I

The Path to Kant Psychologism and Conventionalism

How much more refreshing was David Hume, the Scottish philosopher who held animals in such high esteem that he wrote that “no truth appears to me more evident than that beasts are endow’d with thought and reason as well as men.” In line with my position through this book, Hume summarized his view in the following principle: ’Tis from the resemblance of the external actions of animals to those we ourselves perform, that we judge their internal likewise to resemble ours; and the same principle of reasoning, carry’d one step farther, will make us conclude that since our internal actions resemble each other, the causes, from which they are deriv’d, must also be resembling. When any hypothesis, therefore, is advanc’d to explain a mental operation, which is common to men and beasts, we must apply the same hypothesis to both. Formulated in 1739, more than a century before Darwin’s theory saw the light, Hume’s Touchstone offers a perfect starting point for evolutionary cognition. The most parsimonious assumption we can make about behavioral and cognitive similarities between related species is that they reflect shared mental processes. Continuity ought to be the default position for at least all mammals, and perhaps also birds and other vertebrates. —Frans de Waal, Are We Smart Enough to Know How Smart Animals Are?

Introduction to Part I and Summary of Chapters 2–5 My plan in this part of the book involves drawing on the ideas of Kant’s British empiricist precursors to blaze the most direct, perspicuous trail to guide the reader to his philosophical standpoint. This is because the centrality of psychologism thereto places Kant squarely in the psychophilosophical lineage pioneered by Locke, advanced by Berkeley, and brought to its empirical-psychologistic outrance by Hume. Like them, Kant rejected the analytical approach favored by those who saw in mathematical definition the model for a better method of proving the extra-linguistic reality of familiar vernacular concepts, particularly their psychological reality. They sought instead to produce an “anatomy” of the psyche itself with an eye to determining whether notions expressible by words really are representable in consciousness; and if psychology

34  Path to Kant: Psychologism and Conventionalism indicates they are not, then no matter how rigorous-seeming the analysis, such notions have, and can have, only verbal reality. Berkeley extended the method into properly psychologistic territory when he contended that certain prima facie non-psychological notions might nonetheless owe essential ingredients of their content to consciousness, e.g. corporeal existence to sensation. And while Berkeley contented himself with isolated instances, Hume refined psychologism into a systematically applicable method that he believed capable of explicating the very “cement of the universe.” Chapter 2: Locke’s Oyster Since Kant did not, like Descartes, limit consciousness to beings with self-consciousness, his account of the latter rests on the foundation of a theory of consciousness as such. Here the first question that arises is where consciousness begins. The answer that tends to be favored today is qualia, formulated in terms of “what it’s like to . . .,” e.g., have one’s foot tickled. Yet, whereas qualia may be a satisfactory response to anyone who doubts that humans are conscious other than in the same purely physiological sense in which their bodies sweat, it is ill-suited to demarcating consciousness from non-consciousness. The view endorsed by Kant is the one developed principally by Locke, who contended that the difference coincides with the line between perfect insensibility and sensation of even the most minimal kind, such as even creatures as remote from humans as oysters may be supposed to have. Since everything else, self-­consciousness included, must therefore be regarded as an inessential add-on to the basic sensational consciousness platform, the Lockean psychologist is left with the task of explicating how, from its simplest beginnings in sensation, consciousness is able to take on ever higher forms. Starting with the barest minimum of sensation, proceeding to multiple sensations, multiple sensational fields, senses sensitive to the external environment as well as internally, thoughts that allow sensations to be recollected and imaged, and psychological activities such as discernment of, attention to, and comparison of the contents of sensational consciousness, the process culminates with the fully intellectualized consciousness of philosophers, mathematicians, and scientists. This is the conception of consciousness that was adopted and adapted by Berkeley, Hume, and Kant. And it also accounts for why their approaches lend themselves so well to being integrated with evolutionary science, which simply adds the long-term phylogenetic dimension to their synchronic and short-term ontogenetic analyses of sensational and higher forms of consciousness. Chapter 3: Berkeley’s Vision While the ostensible topic of the chapter is visual spatiality, its underlying focus is imagination, and in particular how a faculty that previous

Path to Kant: Psychologism and Conventionalism 35 to Berkeley had been regarded as limited to fantasies, dreams, hallucinations, and arbitrary associations like phobias first acquired the status of a genuine cognitive faculty, fully worthy of being set alongside the understanding. It has to do with the fact that sensible consciousness of physical objects requires the ability to see, touch, hear, smell, and otherwise sensibly apprehend one and the same spatial object, and so presupposes a consciousness in which the sensory fields of the various external senses are so tightly integrated that they function as a single, unified external sense. Berkeley’s analysis of sensible spatiality led him to conclude that such consciousness is possible only in and through the association in imagination of intrinsically non-spatial visual and other sensations with properly spatial tactual sensation, and impossible otherwise. To be sure, there are other Berkeleyan innovations that ought also to be factored in when reckoning with Hume and Kant, most notably his anti-abstractionism. But none seems to me quite so vital for comprehending his successors’ psychologisms as Berkeley’s associationist attempt at psychologizing the sensibly accessed sense-divide transcending external world (not to be confused with idealism since it leaves open the question of the mind-independent existence of external objects). Chapter 4: Hume’s Cement I noted in the introduction that Hume’s preeminent importance for Kant stems not only from having first raised the problem addressed in the Critique of Pure Reason but also from pointing the way to how psychologism might suffice for its solution. The chapter provides a synoptic view of Hume’s strictly empirical psychologism and the enriched conception of association at its heart. In contrast to Berkeley’s isolated efforts, Hume could and did apply his fortified psychologism systematically, to the point where it may plausibly be supposed not only to absorb into imagination all the non-mathematical cognitive roles formerly ascribed to the understanding, but to constitute a veritable “cement of the universe” itself, at least insofar as we are capable of becoming conscious of it. How this can be and how everything in Hume’s associationist theory of human understanding can be supposed to apply to non-human animal understanding as well are examined, as are some of its limits and shortcomings. For the latter in particular were indispensable to stimulating Kant to develop a systematic a priori psychologism that could at once supersede and subsume Hume’s empirical variety. Chapter 5: From Hume to Kant via Wittgenstein Hume’s skepticism stems from the recognition that there is more to linguistic meaning than psychologism is capable of explicating. Its locus classicus is his analysis of the concept of cause and effect. In ordinary and scientific discourse, cause and effect signifies the necessitation of

36  Path to Kant: Psychologism and Conventionalism the existence of one thing—object, action, or movement—by something else distinct from it. Outside language, however, Hume could identify no other notion of cause and effect than the one provided by customary association. Accordingly, he treated as merely verbal any and all ingredients of the meaning of this and other notions that feature in human discourse but cannot be explicated psychologistically. Language, for him, is essentially dependent on convention and thence, like conventions generally, so deeply embedded in our uniquely human sociality, cultures, and lifeways that it, deprived of the infrastructure they alone can provide, straightaway collapses into meaninglessness, leaving the understanding wholly at the mercy of the association-driven human-animal psyche. Since this excludes any view that accords the meanings of at least some of the notions employed in discourse an independence of all things human—i.e. platonistic conceptions of meaning from the Greeks to Frege and beyond—Humean skepticism may be condensed into the following thesis: any representational content that empirical psychologism cannot explicate convention can, and if convention cannot, nothing can. The consequence is that any notion philosophers, mathematicians, or anyone else supposes to have a meaning that transcends consciousness and convention, or is in any way independent of them, can be nothing more than unintelligible nonsense masquerading as sense. Granted that conventionalism is as much a part of the skepticism Kant set out to refute as associationist psychologism, one must then reckon with the fact that, by contrast with the richly detailed, subtly nuanced treatment Hume accorded the latter, the former figures only tangentially in his thought. Since this want has somehow to be made good if one is to properly understand and assess Kant’s refutation, I have opted to weave into my discussion the views of the philosopher I regard as conventionalism’s preeminent exemplar, Ludwig Wittgenstein (1889–1951). Whereas Hume made no pretense of being able to utilize association to explicate logic or mathematics psychologistically, in his later writings Wittgenstein made a particular point of incorporating these topics into his general treatment of language as inseparably bound up with human conventions, institutions, and lifeways. The resulting conventionalist extension of the scope of Humean psychologism transforms it into the perfect skeptical foil to Kant’s a priori psychologism. This double-edged skepticism is the topic of Chapter 5, including its application to the case that led Kant to credit Hume with waking him from his dogmatic slumber: cause and effect. Hume focused on cause and effect because he deemed it the most fundamental and indispensable notion in the entire human cognitive repertoire. The key to his analysis is the recognition that two of the features essential to the generally accepted definition of the relation—the distinctness of the relata and the necessity of their connection—are impossible to combine in a single act of thought, and so are logically and psychologically incompatible. If true, it follows

Path to Kant: Psychologism and Conventionalism 37 that the psychologistically inexplicable components of the definition must get whatever meaning they have wholly through language, which, in contrast to the representations of the individual isolated psyche, is constituted entirely through social convention, and so is not limited to what is or is not possible in a single act of thought. Yet, precisely because language is as inextricably bound up with our uniquely human sociality, culture, and lifeways as the institution of money and our common acceptance that pieces of metal or paper with a particular provenance are currency, no pretense can be made that the usual way of defining cause and effect is “objective” in any sense that transcends convention and its supporting social infrastructure. Thus the skeptical conclusion follows that, apart from customary association, the notion has, and can have, none but merely verbal validity. Kant, who was of one mind with Hume regarding the fundamental importance of cause and effect, was stopped dead in his tracks by the skeptical implications of Hume’s analysis. He further recognized that cause and effect is by no means the only notion traditionally accorded fundamental importance to objective representation by metaphysicians that is vulnerable to Humean skepticism. Moreover, unlike Hume, Kant saw that the same skeptical reasoning could be used to put mathematics in the same predicament as metaphysics. Since this makes Humean skepticism a threat to the entire conceptual edifice of human cognition, Kant sounded the alarm, calling on philosophers to set everything else aside to devote all their efforts to establishing that at least some concepts of objects are ‘objective’ in a sense that transcends both empirical psychology and convention but is nevertheless non-platonistic.

2 Locke’s Oyster

Where does consciousness begin? Panpsychists tend to hold that it is wherever there is sufficient complexity, regardless of whether that means attributing consciousness to inanimate matter, even at atomic scales. But most contemporary philosophers and scientists of mind suppose that consciousness starts with qualia—what it is like experiencing whatever one is currently experiencing—awareness of which is knowing in the sense that creatures equipped with language can report it. Kant, following Locke, Berkeley, and Hume, charted a middle course, extending consciousness beyond knowing, reportable kinds of awareness to include everything present to us in sensation, whether or not we do, or even can, discern and attend to it—but no further. For them, the difference between sensational consciousness and its absence marks the divide between consciousness and non-consciousness, so that even creatures so primitive as to be completely incapable of knowing consciousness still qualify as conscious provided they have at least one “small dull perception whereby they are distinguished from perfect insensibility.”1 Thus, in relation to the mindbody problem, everything beyond the barest minimum of sensation, qualia included, has to be regarded as an inessential extra.

Dark Consciousness Imagine you are present at the deciding game of the World Series (readers unacquainted with baseball may want to skip this paragraph and the next). The home team is behind 3–0 in the bottom of the ninth, the bases are loaded with two outs, and the batter has two strikes against him. The manager decides to send in the team’s slugger to pinch hit despite the fact he was so badly injured in a preceding game that he can barely walk, much less run. Before stepping into the box, the slugger points to the center field upper deck, sending the cheering crowd into paroxysms. The visiting team’s manager then goes to the mound to bring in his ace reliever. Other things being equal, he would tell his pitcher to walk the slugger. But the batter next up is the best hitter in baseball and the hottest hitter in the Series. By walking the slugger and bringing in a run, a

40  Path to Kant: Psychologism and Conventionalism single would tie the game and a double win it for the home team. Since the slugger is notorious for striking out, the manager tells his pitcher to pitch outside the strike zone hoping he’ll swing at bad pitches. The first two pitches curve well outside. Awaiting the next, the slugger crowds the plate. Seeing this, the pitcher decides to go inside, coming within a fraction of hitting the slugger, who, in his effort to avoid the pitch, aggravates his injury and calls time. The crowd is furious. Refusing to be pulled from the game, the slugger waves off the manager and then once again points to the center field upper deck, reigniting the crowd. After fouling back several more inside pitches, the visiting manager signals to the pitcher to throw the next one low and away to get the slugger to chase it. But rather than be walked, he literally hurls the bat at the ball, causing another foul and heightening the frenzy of the crowd. The manager signals the pitcher to put the next pitch in the same place. This is exactly as the slugger anticipates. With the crowd completely silent, the next sound is the crack of the bat crunching the ball and propelling it straight to where the slugger pointed. Needless to say, there is complete pandemonium as the fans rise as one and cheer at the top of their lungs. You too are screaming at full throttle, but so loud is the roar that you know you are cheering only because you feel it in your throat and chest, not because you can hear yourself through the din. Query: what are you hearing? More particularly, what is auditory consciousness here, what auditory discernment, and what auditory attention? You can hear nothing but the crowd’s deafening roar; but does that mean you are not conscious of tens of thousands of individual voices as well? Of course, you cannot auditorily discern one voice from another, much less attend to what any individual screamer (yourself included) is screaming. Yet, at the same time, if all the voices were suddenly to give out except one at the opposite end of the stadium, then, given suitable acoustics, you would still hear that voice, perhaps even make out what was being screamed (e.g. “we WON!!, we WON!!”). So, in order to hear the crowd’s roar, mustn’t you be hearing—i.e. be conscious of—each individual voice making it up, even if you cannot hear—i.e. discern or attend to—any of them? Or, alternatively, if the volume of all the voices except those in left field were suddenly reduced by two-thirds so that you could discern the difference, would you still not have to be hearing—i.e. be conscious of—the volume of all the voices making up the new distribution of sound, even without being able to discern or attend to the volume of any individually? Or, again, if the brain’s capacity for auditory discrimination could suddenly be enhanced to the point where it became possible to pick out any or even every one of the tens of thousands of voices in the crowd individually, doesn’t this entail that one was already consciously hearing each voice in the stadium—i.e. that their sounds were already present in one’s consciousness—prior to the enhancement? Or if you were suddenly deprived altogether of your capacity for auditory discrimination

Locke’s Oyster 41 and attention, would you not still perceive—be conscious of—the roar? So too memory: if deprived of the capacity to retain the perceptions of even the instant previous, would you not still consciously perceive the sound of the present instant? The upshot of these considerations is that auditory consciousness, strictly so called, presupposes neither auditory memory, discernment, nor attention. Auditory sensational consciousness is therefore primary: the only one of these modes capable, at least in principle, of evolving entirely on its own. The notion that there is more, indeed considerably more, content in conscious representation than can be discerned, attended to, remembered, or articulated in language is characteristic of the kind of analysis favored by early modern philosophers from Descartes to Kant. They used examples2 like the complex auditory consciousness just considered to make a number of points, starting with the importance of not confounding consciousness as such with other varieties of mentation such as conscious discernment and attention. For them, consciousness, properly so called, is much wider in scope than discernment and attention, and so too everything the mind does that requires them. They sometimes compared consciousness to a huge, minutely detailed map, only tiny portions of which are ever illuminated at any one time. The illuminated parts correspond to what we perceive when we discern and attend, and so are not just aware of but knowingly so. Where our doings are concerned (verbal and gestural included), the illuminated area is limited to such actions as we perform with conscious intention, deliberation, calculation, etc. And within that narrow beam of knowing consciousness early moderns further distinguished an even smaller zone where we know what we are aware of with sufficient clarity never to confuse it with anything else. In the Cartesian terminology most early moderns adopted, what we are conscious of unknowingly we perceive only darkly (obscurus, obscur, dunkel), our perception when we are knowingly conscious is bright (clarus, clair, klar), and what we perceive with such lucidity that we cannot confound it with anything else is clear (distinctus, distinct, deutlich).3 My focus in what follows will be confined to the first two. The notion of unknowing dark consciousness tends to have little traction today with mind scientists and philosophers. Psychologists, for example, are apt to describe a driver whose mind is fully occupied in a cell phone conversation as quite literally “unconscious” of the perceptions and actions whereby she controls the car. Clinicians speak of “blindsight” and similarly non-conscious mental activity. And cognitive scientists are wont to refer to “unconscious” attention, memory, and other ostensibly non-conscious mentation, thereby distinguishing ‘mind’ and ‘mentation’ from ‘consciousness’ by equating the latter with knowingly attending, remembering, acting, etc., especially as evidenced by one’s capacity to report it. Such practices may well be justified given how difficult it is to establish the scientific bona fides of anything at all regarding

42  Path to Kant: Psychologism and Conventionalism consciousness. But it should not be forgotten that this may simply be due to the limited technological reach of science today. If at some time in the future consciousness can be reliably detected by monitoring nervous systems, it may become possible to determine whether unknowing consciousness exists and, if so, to investigate it. In the meantime, it must be granted that science cannot prove that early modern philosophers were not at least partly correct in holding that we conceive consciousness too narrowly if we confine it to all and only such mentation as the subject undergoes or performs knowingly. We therefore have no choice presently but to resort to philosophy and do our best to determine where the line falls separating dark consciousness—mentation that is unknowing but nevertheless conscious—from genuinely non-conscious mentation. Early moderns were skeptical that the introspective reports of subjects are of use in this regard. Locke, for example, postulated that there may be actions, indeed entire ordered sequences of actions, that we perceive in ourselves, and so are present to consciousness, but which are performed so rapidly as to exceed our powers to consciously discern one from the next, much less consciously attend to them individually in sequence, with the result that ordered, highly complex conscious mentation attended to introspectively is apt to appear as a single instantaneous simple perception involving no activity at all (something undergone rather than done). Locke’s favorite example was seeing three-dimensionally. On the hypothesis that our only direct conscious access via sensation to three-­ dimensional space is tactual, it follows that if we had to depend solely on visual consciousness, the third spatial dimension would be no more consciously accessible than a fourth or fifth spatial dimension. To compensate, humans are wholly dependent on experience to teach them how to parse the two-dimensional inputs of visual consciousness three-­dimensionally by acquaintance with their more or less constant co-occurrence (or lack thereof) with properly three-dimensional tactual counterparts. Locke compared the process to reading: experience teaches us how to read the contents of visual consciousness three-dimensionally much as we learn to read narratives from characters on a page. We learn which two-dimensional visual data count (or not) as letters in the visual language of three-dimensional tactual spatiality, we learn which combinations of letters make three-dimensional sense (or not) in that language, and finally we learn how to extract real-time three-dimensional narratives from sequences of three-dimensionally meaningful two-dimensional visual combinations. Eventually we become so fluent in the language that we rely more on vision for knowing the world in three dimensions than on direct three-dimensional tactual perception, if only because we see so much more than we can ever get our hands (paws, flippers, beaks, antennae, etc.) on. So, just as with reading, each step in mastering the visual language of three-dimensional tactual spatiality requires conscious discernment and attention. In order to associate data of one sensory modality

Locke’s Oyster 43 according to their correlations (or want thereof) with data of another, we must, at least at some point in our lives, discern and attend to—become knowingly conscious of—these data as they are prior to their acquisition of the significance they ultimately acquire through this process. To be sure, once the visual language becomes second nature—“customary” was the term favored by Locke and his successors—attention is directed entirely at the narrative by which our bodies survive, thrive, or die, even to the point of losing the ability to shift attention back to see things in the same two-dimensional way we did originally—much as we lose the ability to hear speech in our native tongue (another of Locke’s analogies) as creatures incapable of language hear it (or as we ourselves presumably heard it in infancy, before taking our first step toward proficiency). But does being unable to attend to it imply that we are, quite literally, blind to it? Or does dark consciousness of those two-dimensional visual inputs remain as essential to seeing three-dimensionally as dark consciousness of the letters on a page seems essential to attentive awareness of the narrative? Many scientists today proceed on the assumption that consciousness begins with attentive discernment. Since what cannot be consciously attended to cannot then be regarded as consciously accessible at all, these scientists thus seem committed to taking the two-dimensional inputs of vision to be beyond the reach of consciousness, and so to have only non-conscious, purely neurological existence. Against this, upholders of Locke’s view can argue that if it is granted that three-dimensionality is restricted to touch, we could not relate what we are seeing to three-­ dimensional reality if we were not at some level conscious of its associations with tactual three-dimensional spatiality. But if these associations all take their departure from experience of the correlations between touch and the two-dimensional data to which vision is originally restricted, how could we make sense of what we see three-dimensionally if we were not at some level also conscious of these two-dimensional data as such? They are the letters in which the visual narrative describing three-dimensional reality is written. If we had no consciousness of them—were literally blind to them—then we would be equally blind to the narrative (so, too, if a driver absorbed in a cell phone conversation were literally blind to the road—had no sensational consciousness of it whatsoever—a potentially fatal crash would seem sure to ensue). Thus, according to Lockean empiricism, we must see—be darkly conscious of—the two-dimensional given of vision and feel—be darkly conscious of—the three-dimensional tactual associates just as surely as we see—are attentively aware of—the three-dimensional end-product of these experience-bred associations. Whether or not we accept this as a tenable account of three-­dimensional seeing, the conception of consciousness exemplified by it is surely worth exploring. The philosophical grounding of inaccessible consciousness starts with the kinds of examples considered earlier: the roar of the

44  Path to Kant: Psychologism and Conventionalism stadium crowd, reading, and three-dimensional seeing. The next step is the recognition that all consciousness involving the representation of sense-divide transcending spaces and/or times is complex and multi-tiered in precisely this way, but we are not there yet. First, I want to consider two further examples, one provided by Locke and another by Kant, to help us better understand how early modern psychological philosophers sought to establish the existence of dark consciousness. The example from Locke is the experience of taking in the demonstration of a mathematical theorem “with one glance” that otherwise “might very well be called a long one if we consider the time it will require to put it into words and step by step show it to another.” This concerns not the given of sensation but what Locke termed ‘reflexion’ (a.k.a. ‘internal perception’ and ‘internal sensation’). In Locke’s view, the mind of a skilled mathematician cannot take in a proposition AND be conscious of it as a demonstrated theorem, with the necessity and universality that implies, without also taking in the entirety of its demonstration in the premises’ proper sequence. She does not do this knowingly because the actions are performed too fast for introspection, but perform them she must or there could be no apprehending the proposition mathematically, as a demonstrated theorem, nor a fortiori using it as such. The general lesson Locke would have us draw is “how quick the actions of the mind are performed” and “seem to require no time, but many of them crowded into an instant.” And he means: the conscious mind. It is “done with so little notice” because “the facility we get of doing things, by a custom of doing, makes them often pass in us without our notice.” Custom renders notice of individual mental actions redundant, and sometimes even renders it impossible, but, as Locke and his successors understood it, must do so without eliminating them from consciousness altogether insofar as consciousness of these actions is essential to becoming conscious of their sequentially produced joint product—just as with seeing threedimensionally. In the case of mathematical demonstration, custom can only eliminate the need for noticing—discernment of its different steps, attention to them in proper sequence—if it preserves all the conscious representations requisite to apprehending the proposition as proven. So, even though custom may obliterate any introspectible difference between the access a mathematician has who has mastered a proposition’s demonstration and the awareness of a tyro, the sequential presence, albeit inaccessibly (darkly), of each of the demonstration’s steps in the former’s consciousness is essential to the difference in how they comprehend it: as a proven conclusion rather than merely a potential theorem.4 An example provided by Kant is the visual or tactual perceiving of a line. He claimed that to apprehend a line in sensation, it is necessary to form the image for oneself by drawing it in imagination. This, however, does not mean that the act of drawing it is introspectively accessible and involves a succession of individually discernible steps. Instead, so far as

Locke’s Oyster 45 attentive discernment is concerned, the act is instantaneous, and so is discerned as passive and receptive rather than active and constructive. Nevertheless, each step in the sequence must at some level be present in consciousness because a line cannot be represented without representing its length and length cannot be represented except by proceeding part-bypart in such a way that one is (darkly) conscious not only of the sequence and adjacency of the parts but of each as increasing (lengthening) what preceded it. Since the representation of successive lengthening essentially belongs to any representation of a line, whether visual, tactual, in memory, or in imagination, the introspectively passive and instantaneous recognition of a visible or tangible line in attentive awareness is at the same time a dark consciousness of the entire sequence of acts requisite to construct it and unrepresentable otherwise. And the same is true, according to Kant, for all objects, recognizing which involves consciousness of lengths of space and/or spans of time.5 Early modern philosophers regarded examples such as those so far considered as evidence that there is more to consciousness than what is discernible or attentively accessible in it. In something like the way physicists posit inaccessible dark matter because without its gravity electromagnetically accessible galaxy-scale phenomena could not be explained, early moderns posited dark consciousness on the ground that without it much that we are indisputably able to consciously discern and attend to would be inaccessible. If true, it represents something of an embarrassment for contemporary science of mind. The latter is largely restricted to the introspective reports of experimental subjects. To the extent that it is not, scientists have no reliable means to distinguish dark consciousness from genuinely non-conscious, exclusively neurological mentation. And while this embarrassment may one day be remedied, until it is, dark consciousness has the potential to nullify any analysis of mentation that fails to budget for it.

Phylogenetic Relativity The term “access consciousness” is sometimes used purely functionally to designate behavior that is environmentally responsive in genuinely cognitive ways. However, as I shall use it here, it counts as a synonym for what the early moderns termed “bright consciousness,” which may or may not be cognitive but is always conscious, indeed knowingly so, in contrast with dark consciousness, where discernment and attention are lacking and sometimes impossible. I  know of no distinction in use by contemporary philosophers or scientists that coincides with the early modern distinction, but I  can think of no better term to designate our access to the world through sensation and other modes of consciousness.6 Access consciousness in humans is intelligent. We are never aware of bare sensations as a primitive creature might be, or simple associations of

46  Path to Kant: Psychologism and Conventionalism such sensations as a somewhat less primitive one might. Through sensation our consciousness has direct access to a natural world populated by trees, streams, mountains, weather, seasons, sun, sky, moon, and stars. Insofar as other people inhabit it too, sensational consciousness also provides direct access to the social world of interpersonal and collective human behavior. Since the social world includes cultural artifacts and institutions, sensation in humans extends access consciousness into the worlds of spoken and written language, training and education, play and sport, music and dance, flirtation and courtship, hunting, war, trade, and so on, as well as the myriad spoken and written narratives that go with them: ideologies, religions, stories, songs, scholarly and scientific writings, etc., i.e. “memes” of all sorts. But how much of our representation of the world—i.e. the objective phylogeny-independently real contents it comprises—do the sensations through which we access it actually contribute? Is it any more, or different, than the contribution of the sensational consciousness of the most primitive creature, extant or extinct, to the “world” it experiences? If ever we are tempted to equate what we have access to through sensation with objective features of the natural and human worlds themselves, we only need to consider that, for all we know, the evolution of our sensory capacities might have taken a quite different path and equipped us—i.e. creatures otherwise like us—with senses so different that access consciousness would have no qualitative overlap with how it is at present (understanding ‘sensational quality’ broadly to include all observable features of sensation, e.g. not just visible color or hue but also its brightness, glossiness, high/low contrast, etc., as well as visual features for which there are no words). We would then have sensational access to a world with no visual, auditory, olfactory, gustatory, kinesthetic, or any of the other sensational qualities now familiar to us. Nothing would be loud or quiet, high or low, shrill or dulcet; hot or cold, humid or arid, heavy or light, soft or hard, solid or yielding, gentle or brusque, vertiginous, nauseating, or painful; aromatic or appetizing, stinky or rancid, harsh or mellow, sweet or bitter; etc. Access consciousness to the world would thus become as radically unimaginable to us, as we are now, as the sounds a trumpet makes are unimaginable to anyone born deaf or the color scarlet to anyone born blind. So why would we still want to say it is objectively the same world that we have access to presently? How could there be any ground for saying so in sensation in and of itself? How would the worlds of such radically incommensurable sensational access consciousness still be “the same” as our present natural and human reality? That creatures otherwise like us might have evolved with sensational access consciousness radically incommensurable with our own is, of course, pure speculation. It may well be that the only kinds of sensations possible are those through which humans access the world, so that it would make no difference if sensational access consciousness had

Locke’s Oyster 47 evolved on earth, elsewhere in spacetime, in any of the various multiverses posited by physicists, or anywhere else that access consciousness may be physically possible (if only by futuristic genetic or other kinds of engineering). But that too is speculative, and moreover smacks of being fully as parochial as pre-Copernican notions of earth’s cosmic centrality. In addition to all the ways sensory capacities might have evolved on earth but did not, there are also all the ways they could (or did or will) evolve on exoplanets, or any other physical environment favorable to the development of access consciousness in the universe/multiverse.7 Indeed, if one wanted to be as non-parochial as possible, one could extend the definition of “beings with access consciousness through sensation” to scales and dimensions that physics today, with its acquaintance with four but possibly not all fundamental forces, cannot even conceive: forces active far below the theoretical quantum minimum (the Planck length  =  1.61619926 × 10−35 meters) or at scales so great or in spatial dimensions so remote from the familiar three as to be beyond the range even of gravitational influence. For if physical reality cannot be presumed to be restricted to what exists through electromagnetism, the weak force, the strong force, and gravity, then why should the same not be true of sensational access consciousness as well? A truly non-parochial perspective must allow for indefinitely many ways beings otherwise like us might have conscious access to the world through sensation, each signifying a set of sensory capacities, sensations, and sensational qualities radically heterogeneous not only with our own set but every other set as well. There could be no resemblance whatsoever between the (somatically) exterior and interior worlds each species of such beings accesses through its unique panoply of sensations and the worlds accessed by all the others through theirs, ours included. If the same physical stimulus were to act on the senses of a member of each species, the object’s sensational appearance to each would be completely incommensurable with its appearance to the others. Which, then, of the sensational qualities constituting these appearances can be said to actually belong to the object independently of its effect on the particular set of sensory capacities a given observer happens to have? Which observed qualities are truly phylogenetically independent, and to that extent objective? Can all be? Only some but not others? Or none at all? How could it ever be decided? If everything in each species’ physical and social worlds is bedecked in sensational qualities completely incommensurable with the qualities bedecking everything in every other species’ worlds, it cannot but seem risibly parochial to insist that the sensational qualities whereby Homo sapiens accesses objects exhibit the world as it is objectively, while those whereby all differently constituted species access them have nothing to do with how the world is and everything to do with the phylogenetic engineering of their sensory systems.

48  Path to Kant: Psychologism and Conventionalism To be sure, a creature that possessed not just one but all possible sets of sensory capacities might have simultaneous access to all these incommensurable sensational appearances. But would there be anything in those appearances themselves—scrutinizable in them, analyzable from them—to indicate that they were appearances of the same object in the same world? For how different would things appear if the sensations of each set of sensory capacities instead provided access to radically incommensurable universes within a multiversal landscape? Or to no physical world at all? There is no telling from sensational appearances as such; their content simply does not include the requisite information. It might be thought that even if sensational qualities are not independent of phylogeny, then at least the fact of their difference counts as genuinely objective. For example, it may not be the blue per se of the sky above or the green per se of the field that provides access to the world but the qualitative difference as such—a difference that given a sufficiently different phylogeny might equally well have presented itself to consciousness through different colors, through sounds, or through sensations of a kind wholly unknown to us. Or, in the same way that visual and tactual qualities serve equally well to mark out one and the same shape despite the fact that the former do so entirely through differences of light and color and the latter through felt differences qualitatively incommensurable with these, so too creatures with entirely different sensory capacities could use differences of sensations completely unknown to us to delineate that same shape. One may therefore be tempted to conclude that differences of sensational qualities, if not the qualities themselves, are objective, and so are informative about the world in a phylogeny-independent manner. There are, however, compelling reasons to resist this temptation. Given sufficiently different phylogenetic histories (or bio- or other engineering creativity), one and the same stimulus might result in a difference in the quality but not the intensity of one sort of sensation, a difference in the intensity but not the quality of another sort, or no difference at all in a third. In that case, it seems reasonable to expect that a creature with only the first sort of sensation would attribute a change in quality to the object accessed through the sensation, a creature with only the second would suppose the quality of the object to have undergone a change in degree but not in kind, while a creature with only the third would suppose the object to have persisted unchanged. Since this divergence of judgment is rooted entirely in phylogeny, differences in sensational quality seem no more capable of affording insight into phylogeny-independent objective states of affairs in the world than sensational quality itself. A similar argument can be run for sensed temporal differences. What seems simultaneous to one creature might seem successive or separated by an interval to creatures whose senses were refreshed at faster or much faster rates respectively. Or what seems instantaneous to one creature

Locke’s Oyster 49 might seem lengthy to another for the same reason. For example, a creature that refreshed its visual inputs rapidly enough to see a bullet traveling through the air might see a straight line from gun to target if its internal sense of the succession of its own perceptions were refreshed at a slower rate; it might see the bullet leave the gun and slowly travel toward its target if both senses were refreshed with equal rapidity; or it might perceive a temporal interval but not see anything until the bullet hole appeared on the target if the refreshment of its internal sense of perceptual succession were extremely rapid but that of its sight were as it is in us. And so generally: if the patterns of co-occurrence and succession humans experience through their senses need not be replicated in other creatures, can we claim the least justification for believing that our temporal access consciousness to the world is objective and that of all other creatures is merely subjective, illusory, and deceptive? Surely not.

Neural Correlates of Consciousness: Psychological vs. Metaphysical Perspectives My assumption throughout, barring a compelling reason to suppose otherwise, will be that the existence of every sensation in access consciousness is numerically identical with the existence of some complex of neuronal events that I shall refer to as its neural correlate (if sensations exist or are possible by non-neuronal means, then this notion can be generalized to the somatic or, still more generally, physical correlate of consciousness). The neural correlate of any instance of sensational access consciousness should not be confounded with the latter’s cause. Causes and effects are, by definition, distinct existents. This is why it makes no sense to say that mountains are the causes of valleys or vice versa: since one cannot even be so much as conceived to exist without ipso facto conceiving the other to exist as well, they are in truth conceptions of one and the same existence, and so lack the distinctness requisite either to conceive one to be the other’s cause (effect) or to conceive the cause (effect) of the one not to be the other’s as well.8 Being identical in existence with the correlated conscious representation, it is similarly impossible to conceive the neural correlate of any instance of sensational access ­consciousness— or any other conscious representation for that matter—to be that representation’s cause. Of course, there the resemblance with mountains and valleys ends: whereas the latter identity is built into their concepts, that between a conscious representation and its neural correlate is purely existential and not conceptual at all (not at least given the present-day, state-of-the-art conception of physical reality derived from quantum and relativity theories). This is what makes the latter identity unique, and uniquely puzzling. Nevertheless, the special character of their identity makes no difference where its implication regarding causal relations is concerned: just as the causes (effects) of a mountain are at the same time

50  Path to Kant: Psychologism and Conventionalism the causes (effects) of the logically correlated valley and vice versa, the causes (effects) of a neural correlate, as I am characterizing it, are at the same time causes (effects) of the existentially correlated conscious representation and vice versa. For example, whereas everything currently appearing in visual access consciousness owes its existence to a variety of causes—the reflective properties of the objects seen, the opsins that respond (or not) to the reflected photons, the neurons activated (or not) by retinal pigment cells, the neurons activated (or not) by these neurons, and so on through primary and downstream visual systems—there is only one collection of cellular and intercellular events in this causal chain that being caused to exist is identical with the visual access consciousness itself being caused to exist. That collection, not its causes, is the latter’s neural correlate. If, thanks to a yet to be invented technology, the same collection of neuronal events could be caused directly, bypassing everything else in the causal chain, then the correlated visual access consciousness would exist even in the absence of photons entering the eye, opsins, and everything else that, in the normal course of things, causes it to exist. Conversely, if another technology existed that could prevent those events from occurring even if everything preceding them in the causal chain happened, then the visual access consciousness correlated to that collection of events also could not exist. It is in this sense, then, that I count the existence of any particular sensational access consciousness and the existence of the collection of neuronal events that constitute its neural correlate as, quite literally, one and the same. A conscious representation’s neural correlate may or may not be the same from occasion to occasion—e.g. there is no a priori necessity that a sensational appearance indistinguishable from a previous one has to have the same neural correlate or that a neural correlate indistinguishable from a previous one has to be correlated with the same, or any, sensational appearance. It might well be that in the context of the totality of activity in the central nervous system at any moment, different subsets of that activity may be identical with the existence of the sensational consciousness of the blue sky I see now and that of the sensational consciousness of the “same” blue sky I  saw even a moment before. Presumably there is some commonality between them, but that cannot be presumed necessary, for it is not an a priori truth. In any event, for my purposes in this book, all that matters is that there be some neural correlate identical in existence with the existence of everything (all the contents) in every conscious representation, sensational or any other. This raises a question. How can sensational access consciousness and its neural correlate be identical in existence and yet have nothing in common? The one consists entirely of physiological goings on in the brain, the other of color and light, hard and soft, loud and quiet, etc., i.e. the sensations through which human consciousness accesses the world. Even

Locke’s Oyster 51 a complete, perfect explanation of the former would not so much as hint at the existence of the sensational qualities familiar in all human access consciousness. Conversely, no analysis of the sensational qualities bedecking the natural and social worlds of access consciousness, however exhaustive, would betray the least hint of a neural correlate. Indeed, sensational access consciousness is not merely a physiological blank slate but a slate on which nothing physiological can be written, since, so far as we know, the conscious representation of physiological states is not possible by means of sensation alone but requires, in addition, logically structured propositional thought. Accordingly, sensational access consciousness no more points to a neural correlate than to a liver correlate, a correlate 10 miles from the body, or any correlate at all. If suggestive of anything, it is of an existence distinct from and independent of the body and everything in it—a spiritual ghost that scientifically uninformed experience teaches us is hosted by an organic machine. How then can sensational access consciousness, though one in existence with its neural correlate, share none of its attributes? In all other cases, a complex physical system—and none known is more complex than the human brain—shares all of its attributes with whatever its existence is identical to. An eddy in a stream has no attributes distinct from the material components whose physical and chemical attributes create and sustain it. The abiotic precursors of earth’s first single-celled organisms had far more complexly organized material components than an eddy but nevertheless none in any way distinct from the physical and chemical attributes of their components.9 The first true organisms would have had no attributes distinct from the physical, chemical, and physiological attributes of their components, and so on up the chain of biological attributes to the immediate antecedents of the neural correlates of sensational access consciousness. But then, seemingly out of nowhere, that changed: there arose sensational access consciousness that, though one in existence with its neural correlate, consists entirely of sensational qualities and is wholly devoid of physical, chemical, or somatic attributes. The identity of existence and heterogeneity of attributes seems to me the true core of what is often called the “hard problem of consciousness.” The problem of the physics, chemistry, and physiology of the neural correlates of sensational access consciousness is, to be sure, hard enough. But since the only attributes it concerns are somatic, it is a purely scientific problem that admits a purely scientific solution. Where matters become “hard” in a radical sense is with the advent of attributes identical in existence with cellular and intercellular events that are neither physical, chemical, nor physiological but psychological (conscious) in nature. Viewed metaphysically, it is nothing less than the mind-body problem, to which there seem to be only three solutions possible. First, one can swallow hard and accept the quasi-Trinitarian two-in-one logic of one existence that is at the same time two radically incommensurable realities, one with

52  Path to Kant: Psychologism and Conventionalism purely physical attributes, the other exclusively attributes of consciousness. Second, one can deny the identity of existence between sensational access consciousness and its neural correlate and embrace ghost-in-themachine substance or property dualism (two existents), with the option of going further still and espousing neutral monism (an underlying third kind of existent, neither physical nor psychical). And finally, one can affirm a single existence but deny the difference in attributes either by rejecting the reality of sensational access consciousness to embrace elimitivist materialism—a machine with no ghosts to haunt it—or by denying the reality of a neural or any other physical correlate to embrace elimitivist idealism—ghost all the way down. There is, however, an alternative approach that postpones consideration of the metaphysical question until the psychological implications of a neural correlate for sensational access consciousness have been more fully explored. It was pioneered by Locke, who complained that “we began at the wrong end” by “let[ting] loose our thoughts into the vast ocean of being” (ontology) without first occupying ourselves “to take a survey of our own understandings” so as to inventory their “powers and capacities” and fix “the bounds between the enlightened and dark parts of things, between what is and what is not comprehensible by us”10 (psychology). In updated terms, the idea is to use neural correlation to anatomize intelligent sensational access consciousness with an eye to determining which of its components presuppose which and, more particularly, which is presupposed by all the others but itself presupposes none. Identifying some such central core of intelligent sensational access consciousness permits the mind-body problem to be posed in its most basic form, in terms of consciousness at its most primitive, the very point where genuinely psychological attributes first arise and physiological ones cease to tell the whole story—which also happens to be the form in which evolutionists ideally should approach it. Finally, as we shall find when we reach Kant, Locke’s displacement of metaphysics in favor of the psychology of human understanding has within it the potential to forestall metaphysical approaches altogether by depriving them of the conceptual wherewithal to even pose the mind-body problem—which, as we shall see, is the import of a priori psychologism.

The Advent of Consciousness: The True Locus of the Mind-Body Problem If the neural correlate of one’s current sensational access consciousness to the natural and human worlds is identical in existence with a collection of cellular and intercellular events in the brain, then portions of that collection are presumably identical in existence with various components of that consciousness. The latter typically include various forms of memory, discernment and attention, comparison and correlation, imaging and

Locke’s Oyster 53 association, conception, propositional thought, and, last but by no means least, sensation itself. Insofar as any of these presuppose others but not vice versa, it may be eliminated from consciousness by rendering its neural correlates inoperative without affecting those that do not presuppose it. For example, there is no reason to think that sensational access consciousness and its neural correlates could not come into existence if the neural correlates of propositional representation were inoperative. Similarly for all forms of memory: if the relevant neural correlates were rendered ineffectual, there is no reason to think that would prevent sensational access consciousness from nevertheless coming into existence. After all, a primitive creature with sensational access but without memory, much less logically structured propositional thought, seems perfectly possible and may once have actually existed, and perhaps others like it exist today. If so, then neither the hard problem of consciousness nor the metaphysical mind-body problem depends on memory, conception, or propositional thought for its resolution, but solely on whichever of the remaining components of sensational access consciousness is most primitive. The question is how primitive, and, in particular, whether sensational consciousness remains possible even when access consciousness does not, i.e. completely dark sensational consciousness. One might pose the problem this way: could there be a creature that has consciousness of sensation but without any ability to consciously discern or attend to (i.e. access, focus on) anything within it? The baseball game example at the start of the chapter suggests that there could. Certainly, if it were no more problematic than sight with eyes that lacked lenses, then it would be possible for us, whether through a stroke, drug, or some diabolical technology, to have all conscious discernment and attention rendered inoperative, along with all consciousness downstream from them (memory, comparison, correlation, imagining, association, conception, etc.), and yet still have sensational consciousness. Locke, for one, was confident that this is not only possible but an accurate reflection of the mental life of many creatures—“what it is like” to be them: We may, I think, from the make of an oyster or cockle reasonably conclude that it has not so many nor so quick senses as a man, or several other animals; nor if it had, would it, in that state and incapacity of transferring itself from one place to another be bettered by them. What good would sight and hearing do to a creature, that cannot move itself to or from the objects wherein at a distance it perceives good or evil? And would not quickness of sensation be an inconvenience to an animal, that must lie still where chance has once placed it and there receive the afflux of colder or warmer, clean or foul water, as it happens to come to it? But yet, I  cannot but think there is some small dull perception whereby they are distinguished from perfect insensibility. And that

54  Path to Kant: Psychologism and Conventionalism this may be so we have plain instances even in mankind itself. Take one in whom decrepit old age has blotted out the memory of his past knowledge and clearly wiped out the ideas his mind was formerly stored with; and has, by destroying his sight, hearing, and smell quite, and his taste to a great degree, stopped up almost all the passages for new ones to enter; or, if there be some of the inlets yet half open, the impressions made are scarce perceived, or not at all retained. How far such a one . . . is in his knowledge and intellectual faculties above the condition of a cockle or an oyster I leave to be considered. And if a man had passed sixty years in such a state, as ’tis possible he might as well as three days, I wonder what difference there would have been in any intellectual perfections between him and the lowest degree of animals.11 After all, who among us can deny that his or her consciousness could not one day be reduced to the barest minimum of sensation, with that scarcely perceived (i.e. no spotlight of attention, no differences discerned), and nothing remembered, even for a moment? Yet, so long as sensible consciousness persists at all—“some small dull perception”—and we are “distinguished from perfect insensibility,” consciousness, and so too the mind-body problem, retains its purchase. I can therefore see no reason why the same question metaphysicians pose of humans could not with equal propriety be asked of oysters and comparably primitive or enfeebled creatures that yet are not perfectly insensible: is their consciousness one in existence with its neural or other somatic correlate—and, if so, are they distinct in attributes—or, if it constitutes a genuinely distinct existence, is the bond that unites them a miraculous “substantial union,” or some third (“neutral”) kind of existent, or something else? Had Locke been able to view the matter through an evolutionary lens, he might have hypothesized that consciousness first arose, probably as a spandrel, as the barest minimum of difference from non-consciousness. Since in the entire absence of sensation no consciousness of any kind seems possible (pace panpsychism), its terrestrial advent would presumably have coincided with the appearance of the first sensation. But would that first consciousness have been likely to feature conception or propositional thought? Imaging or association? Comparison or correlation? Memory of any kind? Discernment or attention, i.e. access consciousness of even the most attenuated sort? It seems most unlikely. It would have been purely sensational, hence dark, and, in all likelihood, simple (undifferentiated). Of course, it might be claimed that anything answering to this description cannot in fact be distinguished from “perfect insensibility” and is unworthy of the name ‘consciousness.’ Still, how could a creature with sensation, however dark and undifferentiated, be perfectly insensible, that is, as lacking in consciousness as a creature or thing

Locke’s Oyster 55 altogether devoid of sensation? Would we ourselves, if somehow stripped of discernment, attention, retention, and everything else like Locke’s decrepit old man, be reduced to the same status as non-conscious beings, animate or inanimate, if we still retained “some small dull perception”? After all, consciousness had to start somewhere, and no point of origin seems more plausible than the emergence of the first sensation, however simple, dull, and useless at the outset. If so, then sensation marks the fundamental dividing line between consciousness and non-consciousness, at least when the issue is viewed purely psychologically.12 Without sensation, consciousness of any higher kind—discernment, attention, comparison, et al.—does not seem possible. With it, no matter how attenuated, there is consciousness even if, in the absence of discernment et  al., it remains completely dark. Where the brain is concerned, this is just to say that the neural correlates of primary consciousness and the neural correlates of sensation are one and the same, whereas those of all other varieties of conscious mentation, from discernment up, are correlates merely of inessential add-ons, extras incorporated into the basic platform, but not of consciousness as such. Sensation, uniquely, is psychologically both necessary and sufficient for consciousness, so that its first appearance on the planet may reasonably be supposed to mark the advent of consciousness itself, whether its somatic correlate in that period (Ediacaran?) was neuronal, proto-neuronal, or non-neuronal.13 Consequently, nothing is added to the mystery of consciousness specifically as such by further specifying that recollected sensations exist in addition to actual sensations or any other type of mental representation—imagined outcomes, comparisons, associations, propositional thought, emotions, desires, volitions, etc. The question of consciousness, even in its ontological aspect, thus devolves into the mystery of sensation pure and simple. That, at any rate, is the view that coincides most closely not only with Locke’s psychological philosophy but also with those of his principal successors, Berkeley, Hume, and Kant.14 Their equation  of consciousness with sensation contrasts starkly with today’s predominant philosophical paradigm of consciousness: qualia. Qualia are typically described using the formula “what it’s like  .  .  .,” e.g.,  to win the Nobel Prize, speak Pirahã, taste spoiled milk, get high, recall a ditty, feel jilted by one’s lover, etc. These qualia should not be confused with sensational qualities. On the one hand, qualia are not limited to sensations—e.g.  the “what it’s like” quale one experiences when the answer to a complex mathematical problem suddenly dawns on one has nothing to do with color, smell, sound, or anything else that might properly be termed ‘sensation.’ On the other hand, sensations are not all charged with qualia: yes, in the case of “what it’s like” when I dive into freezing water, but no in the case of “what it’s like” to see any of the innumerable small patches of blue composing the sky apparent to me when I gaze at a nearby offshore island,

56  Path to Kant: Psychologism and Conventionalism and certainly not in the case of the lone dull sensation that distinguishes mentation like that of Locke’s oyster or decrepit old man from perfect insensibility. Indeed, qualia do not seem possible apart from access consciousness: conscious discernment and attention at a minimum, and, in many cases, memory, comparison, correlation, association, conception, and logically structured propositional thought as well (often including an understanding of language and other human conventions, customs, and/ or institutions). But if, as the naturalism espoused by the early moderns implies, sensation already of itself introduces something radically heterogeneous from, yet identical in existence with, its neural correlate, then the hard problem of consciousness arises even before factoring in “extras” like access consciousness. The relatively recent shift in philosophical focus from sensations to qualia to mark the divide between consciousness and non-consciousness therefore requires one to make two assumptions that earlier philosophers and many today might not accept: that there is no consciousness except access consciousness, and that there is no access consciousness without a “what it’s like” quale (or is the absence of qualia supposed to be itself a distinctive quale?). Let us return to the example of a driver so absorbed in a phone conversation that she cannot later recall anything present to her sight or touch involved in maintaining control of the vehicle during the interval. In terms of qualia, this is just to say that she was aware only of the quale “what it’s like to hold a phone conversation with so-and-so about suchand-such,” but not the quale “what it’s like to be driving.” Yet, does her lack of consciousness of the latter quale imply not just inattentiveness but a total lack of consciousness of driving? All that is required to say that she is visually conscious of the road ahead and tactually conscious of operating the vehicle is that her sensation suffices to distinguish her perception from perfect insensibility and nothing more. To be driving nonconsciously, none of the visual and tactual sensations involved in driving could be present in her consciousness, for which her eyes would have to have been shut (perfect visual insensibility), and she would have had to go completely numb to being in the driver’s seat operating the steering, foot pedals, et  al. (perfect tactual insensibility). That is why, from the standpoint of Locke and his successors, nothing could be more mistaken than to view consciousness through the narrow beam of attention (a mistake compounded by the common practice in the sciences of restricting consciousness to what human and animal subjects can unambiguously indicate through verbal or other behavior). In sum, for psychological philosophers in the Lockean tradition, human mentation is best thought of as a pyramid with the base consisting of the purely non-conscious sort—by far the largest portion of its volume— a dark conscious mid-section—vastly smaller but still substantial—­and finally a minute tip of access consciousness, where alone qualia are to be found.

Locke’s Oyster 57

Consciousness From the Bottom Up Reliance on qualia rather than sensation to mark the difference between consciousness and non-consciousness brings with it a real risk of misconceiving the mind-body problem in its most basic form: the minimum difference from non-consciousness. In contrast to qualia, sensations can in principle exist in complete isolation, whether as the solitary sensation of a creature like Locke’s oyster or a mind that senility has reduced to a solitary, dark sensational remnant. Even when multiplied and diversified, sensations have no unity in and of themselves, do not signify anything beyond themselves, are devoid of context and reference, and are in no way indicative of anything else. That is why Locke and his successors regarded sensations as simple perceptions: they can be known only by having them (consciously perceiving them), i.e. their content is not otherwise accessible; and in having them, nothing else is known, i.e. no further content, sensational or non-sensational, is thereby made accessible.15 This includes their neural correlates, regarding which sensations are complete informational blank slates. None of the inter- and intracellular complexity of their somatic correlates is manifested in the corresponding sensations, which may well be as utterly simple and uniform as Locke’s oyster’s lone “small dull perception.” For while it is true that physiologists can draw accurate inferences about neural correlates and their causes from what subjects report of their sensations, that is the fruit of expertise and research, not anything they or their subjects derive directly from those sensations, simply by having them. Thus, sensation marks the divide where the ever-increasing complexity of physical systems from the subatomic through the atomic, molecular, chemical, cellular, and intercellular, on up to its neural correlate, gives way to a psychological reality that is quite literally wiped clean of all that complexity.16 It is this elemental simplicity and primitiveness that explains why Locke and his successors regarded sensation as the true starting point of psychological philosophy. Their goal was to explain how, given only sensations as inputs, outputs can be produced from the simplest requisites of access consciousness (discernment and attention) all the way up to fully intellectualized representations of causally interconnected objects and selves existing in the natural and human worlds. Considered from an evolutionary perspective, it means going back to the very beginning of consciousness: the first sensation. Although we cannot know whether the first creature in which the first sensation appeared was oyster-like or quite different, the mutation that produced it may well have been adaptive for other reasons, making consciousness originally a spandrel that established itself in a population only by piggy-backing on a genetically or epigenetically related adaptive feature. If so, we can only guess at the kind of process that transformed the first sensational consciousness into an adaptation. Perhaps its presence in an aquatic creature happened to

58  Path to Kant: Psychologism and Conventionalism coincide with the precise point where it was benefited rather than harmed by closing a certain valve. Given a new mutation that caused the valve to close whenever the sensation occurred (i.e. whenever its somatic correlate was present), the sensation would coincide with a boost in fitness, and thus gain its own evolutionary raison d’être. A second step might be the advent of sensational variability. How this could come to be targeted by natural selection is not difficult to imagine. Suppose that members of the aforementioned species could benefit from an ability to gradually constrict and relax the valve rather than being limited to having it fully closed or fully opened. Suppose further that there was some obstacle, biological or environmental, that prevented the requisite regulatory system from evolving in a purely non-conscious fashion. In that case, if mutations occurred that enabled the sensation that triggered the valve’s closure to vary gradually from smaller to larger and/ or from duller to more intense, the sensation might step into the role of regulating the gradual constriction/restriction of the valve, not unlike a virtual reality controlling device regulating the valve of a machine. Under these circumstances, it is not unlikely that the somatic correlates of its varying degrees would become a target for natural selection—correlates that might otherwise so differ from one another that nothing purely physiological would otherwise have connected them so as to form a single, naturally selectable somatic complex. If so, it would be a clear example of consciousness causally driving evolution rather than being merely an epiphenomenal bystander. Further innovations in sensational consciousness can be supposed to establish themselves (via their somatic correlates) in similar ways: from differences in a single sensation regulating a single behavior to different sensations regulating different behaviors; sensations co-opted into performing a role in sensing conditions external to an organism and eventually evolving into different senses; sensations from different sense faculties combining to regulate behaviors better than each could alone; discernment of sensational differences permitting more finely grained behaviors; and, with the ability to attend to discerned differences, the emergence of sensational access consciousness and all it portends—evolutionary steps all overlooked insofar as access consciousness, or anything dependent on it such as qualia, is confounded with the advent of consciousness as such. For while the actual evolutionary path taken by consciousness from its dark sensational beginnings is unknown and quite possibly unknowable, the hypothesis that it was not merely an epiphenomenal bystander but rather to some extent a driver of evolution is more than merely speculative. A radical heterogeneity between the attributes of consciousness and those of its somatic correlates follows if, as suggested earlier, (1) sensations can be simple notwithstanding the complexity of their somatic correlates, (2) sensations are complete informational blank slates with respect to their somatic correlates, and (3) sensations contribute no

Locke’s Oyster 59 phylogeny-independent, objectively real contents to our representation of the world, including the part situated inside our heads. This radical heterogeneity with respect both to the world and to the brain increases the likelihood that attributes exclusive to sensational consciousness and lacking in its somatic correlates, their correlates’ causes, and the organs that support them (not to mention their molecular, atomic, and subatomic constitution) may have created unique possibilities for regulating and controlling behavior, some of which established themselves and spread through populations. If so, then sensational consciousness generally, far from being epiphenomenal, may well have played an active role in molding the brains of virtually every sentient creature that has ever lived, ourselves not least. Of course, if sensations are devoid of objective information, and so purely subjective, then they cannot suffice for experience of objects in the natural and human worlds. Objectivity must instead be introduced into access consciousness in a form specifically suited to it. And this, it turns out, consists in the integration of sensations with thoughts.

Objective Knowledge vs. Objective Understanding Under the general rubric of “thoughts,” I  classify all non-sensational, purely mental representations, including the many varieties of conscious memory, discernment and attention, comparison and correlation, imaging and association, and conceptualization for judgment and inference. This way of categorizing them serves to exclude two classes of representation: emotions, desires, volitions, and similar such affects, which early moderns treated not as thoughts but as a second species of sensation (e.g.  internal vs. external affects in the case of Kant, reflexive vs. sensate sense impressions in the cases of Locke and Hume); and discourse, i.e. linguistic thoughts, which, as convention based and so irremeably social, are not purely mental and so cannot properly be classified as psychological. So construed, thoughts should be regarded as no less identical in existence with their neural correlates than sensations, and, also like sensations, as complete informational blank slates with respect to those correlates. Since what makes sensations subjective is not only their blankness but the identity of their existence with that of their phylogenydependent neural correlates, this raises the question of how thoughts can ever be objective if they are no less blank and they too are indistinguishable in existence from their phylogenetically relative neural correlates. All extant conscious species on this planet, and possibly all extinct ones as well, may have evolved from a last common conscious ancestor. But at any time subsequent to this creature evolution could have taken any of innumerably different courses, some of which might have resulted in our having completely different sensations than the ones we have. The likelihood of there being creatures with sensations incommensurable with our

60  Path to Kant: Psychologism and Conventionalism own seems greater still if there were no last common conscious ancestor, particularly if sensational consciousness has evolved on other planets or in environments radically incommensurable with earth’s. Certainly, such evolution is possible, or at least bio- or otherwise engineerable. But if this could be the case with our sensations and their neural correlates, might it not be so as well with our thoughts and their correlates, including even those we regard as paradigmatically objective? Or is objectivity something more, or even different altogether from anything phylogeny dependent? Most philosophers today, particularly in the analytic tradition, would no doubt regard the notion of phylogeny-dependent objective thought as nonsensical, a contradiction in terms, or at least claim to be unable to conceive how such a thing might be possible. Mathematics is a case in point. Along with many mathematicians and scientists, they would insist that mathematicians of any other species, however different from us, could never disagree with human mathematicians, be it about the simplest problems or the most complex, provided all “got their sums right” (applied the rules as prescribed). Disagreement about whether the product of seven and five equals thirty-five would be possible only if there were a contingency under which the product could turn out differently; since there is none, there is but one possibility, and so only one correct answer for mathematicians of any species, making the equation a genuinely phylogeny-independent objective representation. Those who take a maximally realist view of logic would argue similarly regarding their objects and relations (logical operators, inference forms, etc.), as would physics maximalists regarding theirs (laws of nature, constants, fundamental forces, etc.), chemistry maximalists regarding theirs (atomic elements, molecules, compounds, etc.), and so on for maximal realists in other disciplines. The upshot is a maximalist epistemology of objects of thought that, while conceding that one species may be superior to another in ordinary or scientific knowledge of this or that kind of object, insists that all creatures think and know the same objects, and that insofar as belief can count as knowledge only insofar as it is true of its object, the knowledge of objects by one species of creature cannot fail to agree with the knowledge of those same objects by any other species. Since this is just to say that knowledge of objects is no less phylogenetically independent than the objects themselves, a maximally realist epistemology implies that knowledge of objects, together with everything—conceptual, propositional, and inferential—that enters into it, is fully as objective as the objects themselves, and so the epistemic antithesis of sensation. Conversely, the admission of even so much as the possibility of phylogenetic variability implies that no species’ beliefs can ever meet maximalist standards of objective knowledge, with the consequence that even the very notions of objects and objectivity themselves would become irremediably suspect.

Locke’s Oyster 61 In other words, from the standpoint of a maximalist epistemology of objects, the phylogenetic dependence of the components of knowledge is tantamount to the kind of radical skepticism à la Descartes that maximalists of every stripe typically deem untenable or even self-refuting. Yet, as attractive as a maximally realist epistemology may be, there are other ways of regarding matters that reduce its appeal. In the first place, insofar as it is difficult, if not impossible, to conceive the kind of sophisticated objective thought epitomized by science and mathematics as capable of taking place in the absence of vernacular language, it may have to be conceded to be no less dependent than language on human convention and the socio-cultural infrastructure and species-wide lifeways that make conventions—linguistic and otherwise—possible. Second, even setting aside the specter of phylogenetic relativity raised by language-dependence, there is the fact we have direct acquaintance with objective thought solely in our own species, and it is surely quite a leap to elevate what humans agree to be objective into what every other creature, actual or possible, would agree—indeed, have to agree—to be objective as well, no matter how different thought in its phylogeny was from ours otherwise. How can we presume to know this? That we can conceive of no other kind of objective thought than our own imposes no limit on what nature may or may not be capable of bringing about—no more than our inability to imagine sensations other than our own constitutes evidence that nature has not produced them, or at least could not, even if only with the aid of genetic (or other) engineering. Nor, third, is it enough simply to declare that the notion of phylogeny-dependent objective thought is self-contradictory without first taking the trouble to prove that phylogenetic independence is in fact essential to objectivity. For it is by no means obvious why such independence is indispensable, or even useful, in order for objective thought and knowledge to perform any of the roles they actually play in the lives of our species. Surely, all that ultimately matters for us is that we can agree about them and that they work as we intend, not that they also be true in any absolute, phylogenyindependent sense. Fourth, and finally, placing the human understanding of objectivity at the universal center of objectivity by affirming its phylogeny independence risks running afoul of the so-called “Copernican complex”: the principle that we should assume, barring compelling evidence to the contrary, that everything about us—our biology (neural correlates included), our planet, our solar system, our galaxy, etc.—is average in every respect (including being exceptional in a few), merely one variant along a spectrum of variants, and so no more entitled to be elevated to the status of touchstone than any other instance of the kind. If true of our biology et al., then why not of human objective thought as well? More debilitating for epistemological maximalism, in my view, are the difficulties involved in reconciling it with the evolutionary origins of conscious representations and their neural correlates. Is it likely that the first

62  Path to Kant: Psychologism and Conventionalism elements to emerge—the prototypes—of today’s panoply of objective representations were phylogeny independent? Or were they simply phenotypic additions that, in the context of existing non-objective sensations and thoughts, and so in a wholly phylogeny-dependent way, enhanced their possessors’ fitness? Is it not also likely that the same is true of each subsequent increment, all the way up to and including our own species’ cognitive access to objects in the natural and human worlds? To maintain the contrary seems to premise a radical break from the previous gradual, genomic alteration by genomic alteration, evolutionary development of consciousness, with neural correlates of objective representations instead arising, for all evolutionary intents and purposes, ex nihilo, with complete indifference to the existing phylogeny and ontogeny of consciousness. That these phylogeny-independent representations were adapted to their environmental/biological context would then have to be treated as purest serendipity, their appearance in a consciousness able to support and deploy them a saltation so massive as to be little short of miraculous. Though not impossible, from an evolutionary perspective, an explanation of the emergence of objective thought that does not premise its phylogeny independence from the outset seems more promising. And Berkeley’s account of our ability to utilize thought to integrate the data of the various senses so thoroughly as to see, touch, hear, etc. one and the same sense-divide transcending object offers an example of precisely that.

Notes 1. ECHU 2.9.14. 2. Leibniz instanced the sounds of water droplets colliding that make up the roar of a wave crashing against the shore, while Kant evoked the sound of the individual tones simultaneously struck by an improvising musician: we cannot discern and attend to the sounds of the droplets or tones individually but must nevertheless be conscious of them individually if we are to hear the breaking of the wave or the chord. The same analysis was applied to other sensory modalities and to intellectual representations like mathematical demonstrations and complicated concepts like virtue. 3. These terms are commonly translated as obscure, clear, and distinct. My translations better capture the sense of the terms in the original languages— senses that may have been more evident in early modern English ‘clear’ and ‘distinct’ but that have since largely been lost. 4. Quotations are from ECHU 2.9.10. 5. The line example can be found at CPR A102 and B137–8. Recognition (A103–4) typically involves only so weak a consciousness of the synthesizing activity responsible for generating the representation of the recognized object that we wrongly suppose that consciousness of the activity forms no part of consciousness of the object: see chapter 10. 6. My use of the term thus has nothing to do with Ned Block’s distinction between phenomenal and access consciousness (both of which would count as bright for early moderns) or Gerald Edelman’s distinction between primary and higher order consciousness (ditto).

Locke’s Oyster 63 7. A chemistry sufficiently complex to support life may also be able to support conscious life: “For example in their 2006 paper ‘A Universe Without Weak Interactions,’ Roni Harnik, Graham Kribs, and Gilad Perez put forward a universe that seems capable of creating preconditions for life and yet has fundamental particles entirely different from our own. In 2013, Abraham Loeb, of Harvard, argued that a primitive form of life might have been possible in the early universe. And recently, Fred Adams and Evan Grohs showed that if we vary several parameters in our theories at once, there are ways for stars to produce carbon other than the mechanism Hoyle predicted.  .  .  . These three examples show that a chemistry complex enough to support life can arise under circumstances that are not anything like the ones we experience, and our universe isn’t all that special. . . . That theories with dozens of nonlinearly interacting components give rise to complex structure is something I  expect to be the rule rather than exception.” Sabine Hossenfelder, How Beauty Leads Physics Astray: Lost in Math. New York: Basic Books, 2018, pp. 114, 260. 8. More will be said on the topic in chapter 5. 9. Addy Pross, for example, makes the case for viewing all biological phenomena as chemical, making biology simply a branch of systems chemistry (i.e. sub-branch of chemistry), in What Is Life? How Chemistry Becomes Biology. Oxford: Oxford University Press, 2012. 10. ECHU Introduction §7. Shifting psychology to philosophical center stage in place of metaphysics may well be the most significant of all of Locke’s numerous important philosophical innovations. See KEUU part I and KAIM chapter 2-A. 11. ECHU 2.9.13–14. I do not know of anyone who has argued for consciousness in oysters or cockles. Andrew Barron and Colin Klein find reasons to attribute consciousness to insects in “Insects have the capacity for subjective experience” (Animal Sentience, 2016, p. 100) and “What insects can tell us about the origins of consciousness” (PNAS, 2015), while Jon Mallatt and Todd Feinberg make a strong case for supposing that consciousness goes down at least as far as lampreys in The Ancient Origins of Consciousness. Cambridge, MA: MIT Press, 2016. It should, however, be noted that the notion of consciousness with which Mallatt and Feinberg operate involves referral (or aboutness), mental unity, qualia, and mental causation, and so includes a great deal more than what is required to distinguish sensational consciousness as such from perfect insensibility. Their conception instead seems closer to Hume’s model of associative consciousness discussed in chapter 4. 12. The absence of psychological grounds to affirm sub-sensational consciousness has not stopped it from being posited on physical or metaphysical grounds, as witnessed by the renewed popularity of panpsychism among scientists and philosophers. 13. The non-neural scenario is that sensational consciousness first originated with some other kind of cell as its correlate but was later supplanted by proto-neurons or neurons. If tissue of that cell type still exists, it has presumably evolved into something that no longer has anything to do with consciousness in most or even all extant species. 14. Some may object to my assumption that these philosophers would have countenanced neural correlates of consciousness, but it is important to recognize that I am discussing only their psychological philosophies to the exclusion of their metaphysical and theological commitments. Even in the latter regards, however, I see no insuperable obstacle to updating Locke, Berkeley, Hume,

64  Path to Kant: Psychologism and Conventionalism and Kant to include this notion. In Locke’s case, whereas his readiness to entertain the hypothesis that human thought may be an attribute of an otherwise purely material substance was hesitant and highly controversial, I have no doubt that neither he nor many of his contemporaries would have had comparable misgivings about supposing that if oysters and comparably primitive creatures are not perfectly insensible, their consciousness has a purely physical correlate. Similarly, since Berkeley saw no difficulty incorporating the complex workings of physical systems, including human and animal physiology, into his theologically inspired idealism, I cannot envisage him balking at the empirical discovery that conscious goings on invariably correlate to happenings in the brain. I think there are few who would hesitate to concede that Hume would likely have had no principled objection to neural correlates, while any doubts readers have respecting Kant should be addressed in this book’s second part. 15. Sensational simplicity is often mistakenly equated with not being further differentiable, especially by division. It is true that Locke and his successors regarded complex ideas as ultimately resolvable into simpler ideas, which, in this context, means that a point is eventually reached where further distinction of complex ideas into components (sub-components, subsub-­components, etc.) ceases to be possible (though even here one needs to be careful, e.g. Hume: “impressions of touch are simple impressions except when considered with regard to their extension,” T 1.4.4 ¶14). But while this is the most important sense “simplicity” carries in Locke’s discussion of complex ideas, it is a purely relative meaning. Its non-relative meaning, which may justly be deemed its root, or core, meaning, is the one described here. See KEUU chapters 5 and 16. 16. This may pose a problem for the integrated information theory (ITT) of consciousness espoused by Giulio Tononi, Christof Koch, and others. Whereas the somatic correlate of a lone, dull consciousness such as that Locke posited in oysters is no doubt representable as a highly complex, multi-level information system, the consciousness itself is perfectly simple and so, informationally speaking, completely null. This informational asymmetry creates a significant risk that ITT will either falsely attribute the same informationally rich content to the consciousness that it correctly detects in its somatic correlate or, conversely, falsely take the informational nullity of the sensation to be incompatible with its having the status of consciousness at all, notwithstanding its distinctness from perfect insensibility. That is why one should be wary of discarding psychological criteria of consciousness and relying exclusively on physical and information-theoretic ones: if consciousness did indeed originally arise with a single simple, dull sensation, this momentous step in the evolution of life on this planet would otherwise be entirely invisible to science.

3 Berkeley’s Vision

While the ostensible topic of this chapter is visual spatiality, its true focus is imagination, and in particular how a faculty that previous to Berkeley was regarded as limited to fantasies, dreams, hallucinations, and arbitrary associations like phobias first acquired the status of a genuine cognitive faculty, fully worthy of being set alongside the understanding. It has to do with the fact that representation of physical objects requires the ability to see, touch, hear, smell, and otherwise sensibly apprehend one and the same spatial object, and so presupposes a consciousness in which the sensory fields of the various external senses are so highly integrated that they function as a single, unified external sense. Berkeley’s analysis of visual spatiality led him to conclude that such consciousness is possible only in and through the association in imagination of intrinsically non-spatial visual and other sensations with properly spatial tactual sensations. It thus cleared the way for Hume and Kant to enrich and develop Berkeley’s conception to the point where the imagination could be accorded a central place in their comprehensive psychologistic treatments of cognitive understanding.

Eyesight to the Blind Were someone given sight who previously never had it, would they, literally at first sight, be able to distinguish a cube from a sphere, without touching them? William Molyneux first posed this question to Locke in 1688, and it continues to attract the attention of philosophers and scientists to this day. Its interest derives from the light it casts on the relation of vision to touch with respect to space. Granted that the skin together with the body it covers give one tactual sensation in all three dimensions at once, is the same true of vision? Two dimensions, not three, seemed to Locke and Molyneux all that can be gotten from visual sensation alone. After all, if we were to draw straight lines from objects at varying distances to each of our eyes, the lines would all be perpendicular; and since this means that we could see only their endpoints, not the lines themselves, we could not visually contrast their lengths so as to

66  Path to Kant: Psychologism and Conventionalism determine distance and depth of field. This is by no means to deny that vision enables us to make judgments regarding depth and distance based on past experience of the correlations of visual sensations with tactual. But only touch is capable of supplying this information directly, without depending on comparison or experience-based judgment. How then could a newly sighted person possibly distinguish a cube from a sphere at a time when he still lacks any idea of how what he sees affects his touch? Before answering, we need to consider further what the absence of all experience of visuo-tactual correlations means for the novice seer. He would have no idea whether different distances, orientations, heights, etc. have different colors (i.e. hues, values, and/or chromatic intensities), and if so what they are. He would be completely ignorant about whether and how different angles and intensities of lighting affect what he sees. He would be clueless whether or how what he sees would change were he to shift his eyeballs, move his head, stand up, turn around, walk, jump up and down, run, etc. Nor would he have the least notion of what wetness and dryness look like, the different looks of rough and smooth texture, whether or not solidity and its lack differ chromatically, whether heat has a special look, whether rashes and backache do, what a happy face looks like, whether facial and other bodily expressions of emotion manifest themselves visually, and if so how, etc. With no previous experience to teach him how to connect visual sensations to tactual, much less knowledge of what those connections are or even whether they exist at all (assume he has not been coached or otherwise primed),1 it seems all but certain that someone newly sighted would be completely unable at first sight to visually distinguish the same figures he formerly knew solely by touch. To Locke, a negative response to Molyneux’s question is important because of its incompatibility with the common assumption that the same three-dimensional figures present in tactual sensation are present also in visual sensation. For how could a newly sighted person fail to recognize the three-dimensional spaces made familiar by a lifetime of tactual experience if they really were identical? Wouldn’t his failure to recognize purportedly familiar shapes prove their identity to be entirely illusory? The key point is that where three-dimensional space is concerned, the two species of sensation are radically incommensurable: they have absolutely nothing in common. Again, this is not to deny that sighted people are as capable of recognizing three-dimensional configurations visually as they are capable of doing tactually. What Locke denied is that, in the visual case, such recognition is purely a matter of having the requisite sensation. Instead, three-dimensional visual recognition is built on the correlation of the visual with the tactual, the grouping together of visual sensations according to their tactual correlations, and the ability to recognize visual data three-dimensionally in essentially the same ways speech is comprehended and texts are read. In other words, visual recognition of the third

Berkeley’s Vision 67 dimension is a matter not merely of visual sensation but of thought as well: memory, comparison, correlation, association, and a species of comprehension analogous to the linguistic sort. Of course, while Locke may have denied the identity of visual and tactual space at three dimensions, he did not do so at two, with the implication that Molyneux’s man would have no difficulty recognizing the identity of visible squares and circles with tangible squares and circles at first sight, and so have no difficulty telling them apart without the aid of touch. More generally, Locke had no principled objection to the notion that the same representational data can enter consciousness via different sensational channels, which he termed “ideas of diverse senses.” So, while visual sensations may be intrinsically spatial up to two dimensions, the converse is not true: space is not intrinsically visual in the way sensations of color and light are (“ideas of a single sense”). Similarly, while tactual sensations may be intrinsically three-dimensional, three-­ dimensional space is not intrinsically tactual like feelings such as itches and twinges are. Instead, there is such a thing as “the abstract idea of space” that is common not only to vision and touch but to these sensations and mind-independent physical reality. Locke’s solution to the Molyneux problem was disputed on various grounds, but perhaps the boldest and most thought-provoking challenge was issued by George Berkeley in his first major work, A New Theory of Vision (1709), published when he was only 24. Berkeley doubted whether Locke had succeeded in fully abstracting from the contributions of thought to sense experience when he supposed that there are contents common to the sensations of different senses. Take the case of vision and touch. Suppose that something happened or was done to prevent any consciousness of correlations of visual sensations with tactual sensations (a stroke, drug, surgical intervention, etc.). Touch, broadly construed to comprise sensation in and beneath the skin (including kinesthetic—­ proprioceptive and vestibular—and interoceptive sensation),2 would still produce the feelings one has when shifting one’s eyeballs in their sockets, tilting one’s head, bending at the torso, standing up, sitting down, walking, running, turning, jumping, and all the rest. But visual sensations, being unrelatable to these feelings, would now be completely independent of them, and so, spatially speaking, entirely on their own. Berkeley’s question was whether even so much as one spatial dimension could in that case be represented by purely visual means alone. Let’s start with the third dimension. Could three-dimensional sense be made of the kaleidoscopically shifting collages of varying hues and highlights caused by bodily motion if deprived of even so much as dark conscious relation to tactual sensations? If binocular vision were severed from all feeling of bodily position, orientation, distances between its various members, and whether and how any of those members were moving, it might still produce the same look, the same “special effect,”

68  Path to Kant: Psychologism and Conventionalism we normally equate with seeing in depth. But could we then be conscious of it as the third dimension of space? Not being in any way like, much less identical with, the third dimension of tactual sensation, Berkeley did not see how we could be. Things might retain the look we normally associate in thought with the third tactual dimension. But with all connection to the latter severed, that association is entirely effaced, and with it the tactually derived three-dimensional spatial meaning we are accustomed to attach to the special look conferred on visible objects by binocular vision. And the same would presumably apply to tri- and higher multiocularity, including creatures that had enough eyes to see objects of any size from every angle at once: with all connection to tactual spatiality severed, only special effects would remain, not the spatial meaning customarily attached to them in thought. Nevertheless, even if Berkeley’s reasoning regarding the third dimension is conceded, most vision theorists would still probably follow Locke in resisting its extension to the first two visual dimensions. But consider what it would mean for seeing in two dimensions if the third dimension literally became invisible as a result of severing all connections between sight and the spatiality of tactual sensation. Would the width and height of a cube or a sphere be any more than looks, mere “special effects,” in the same sense as binocular depth? One might answer no on the ground that color borders and the shapes they demarcate would remain. But would they? It depends on how much of a spatial assist they get from moving one’s saccading eyeballs to direct one’s foveae along the color boundary AND from being able to connect concurrent visual sensations to the feeling of moving one’s eyeballs thus and so—tactually sensed eyeballs that in their turn are situated at a particular place in the tactually sensed head and tactually sensed body overall, i.e. in the tactual sensational field of bodily spatiality.3 The question is this: if all connection to the visual “boundary” as felt in moving one’s eyeballs to track it were severed, would a linear boundary still be visible? According to Berkeley, although the synchronic and diachronic differences of color and light would remain, the loss of all connection to tactual sensation in the eyes— and with it all connection to tactuo-bodily spatiality generally—would entirely denude them of the spatial meaning of “linear boundary” “dividing” distinct two-dimensional “areas” we are accustomed to attach to them in thought: [Locke and others] are tempted to think that flat or plane figures are immediate objects of sight, though they acknowledge solids are not. And this opinion is grounded on what is observed in painting, wherein (it seems) the ideas immediately imprinted on the mind are only of planes variously colored, which by a sudden act of the judgment are changed into solids. But with a little attention we shall find the planes here mentioned as the immediate objects of sight are not

Berkeley’s Vision 69 visible but tangible planes. For when we say that pictures are planes, we mean that they appear smooth and uniform to the touch. But then this smoothness and uniformity, or, in other words, this plane-ness of the picture is not perceived immediately by vision: for it appears to the eye various and multiform. . . . What we strictly see are not solids, nor yet planes variously colored, but only diversity of colors. And some of these suggest to the mind [tangible] solids, and others [tangible] plane figures, just as they have been experienced to be connected with the one or the other.4 In sum, on Berkeley’s view, if the Molyneux man’s mind were wiped clean of all associations of the visual with tactual spatiality, he would be as helpless to visually distinguish a square from a circle, or even a straight from a crooked line, as he would a cube from a sphere. The general thesis Berkeley defended is that, contrary to Locke, there is nothing whatsoever common to the sensations of diverse senses, be they sight and touch or any others. Suppose, for example, that we abstract everything specific to the sense of sight from a red triangle seen against a yellow background, and then abstract everything specific to the sense of touch from the feeling one gets running one fingers along the edges of a triangular table. Does there still remain an abstracted shape, with no features whatsoever specific to either sense, or indeed to any sense, yet common to both and recognizably identical between sensory modalities? First take sight. The sensational qualities specific to visual consciousness are color and light (hue, brightness, gloss, etc.). Berkeley reasoned that if visible shape were genuinely distinct from color and light, the latter could be entirely eliminated without thereby affecting our ability to see the shape. If we are inclined to affirm that this is indeed what would happen, it cannot be because there is some third thing besides color and light in visual consciousness that remains to delineate the shape. It is rather that we are so strongly influenced by our life-long habit of utilizing correlations with tactual sensation to demarcate shapes visually that, instead of excluding tactual spatiality as we are meant to do, we surreptitiously rely on it and confound what is in truth a tactual triangle with the sensing of something that is specific neither to sight, touch, nor any other sense. Only when the ghost of tactual spatiality has been fully exorcised—no easy task given the habits of a lifetime—can we recognize that the elimination of all color and light from visual sensation leaves quite literally nothing to be seen, i.e. renders us completely blind (perfect visual insensibility). Accordingly, even if there were a visual spatiality distinct from tactual spatiality, it would be indistinguishable from uniquely visual sensational qualities. And since this is just to say that visible triangles are no less unique to visual sensational consciousness than mauve and matte, Berkeley concluded that it can have nothing in common with, much less recognizably identical to, anything present in tactual sensational consciousness.

70  Path to Kant: Psychologism and Conventionalism Now consider touch. The sensational qualities unique to that sense include hard/soft, smooth/rough, solid/unresisting, warm/cool, wet/dry, tickling, itching, pinching, pressing, rubbing, stroking, as well as the feelings incident to the flexing or relaxing of muscles anywhere in the body, along with kinesthetic (including vestibular) and interoceptive (including nociceptive) sensations generally. If the triangle we feel were genuinely distinct from all touch-specific sensational qualities, then the latter could be entirely eliminated without impairing our ability to feel the shape. Yet, it seems clear that the elimination of all qualities unique to touch from tactual sensation leaves nothing whatsoever to be felt—no spatial or any other tactually accessible residue (perfect tactual insensibility). Since this is just to say that tactual spatiality is no less unique to tactual sensation than itches and pinches, it can have nothing in common with, much less recognizably identical to, anything present in visual sensational consciousness. Thus, even if there is something in visual sensations we persist in calling “spatial,” Berkeley insisted that it can have nothing more in common with tactual spatiality than the word, and “we can no more argue a visible and tangible square to be of the same species from their being called by the same name than we can that a tangible square and the monosyllable consisting of six letters whereby it is marked are of the same species because they are both called by the same name.”5 Two things follow from this reasoning. First, if visual and tactual sensation do not have a common spatiality so that only one, but not both, can be intrinsically spatial, then the other must borrow its spatiality from the other via its systematic connection to it in thought. Since, in Berkeley’s view, the evidence that tactual sensation is intrinsically spatial is overwhelming, it follows that vision is spatial only derivatively, at second hand (the same no doubt applies to the auditory spatiality of the echo-locating sensations of bats, the echo-sonar sensations of cetaceans, the navigational sensations produced by the electric organs of certain fish, the location-detecting sensations some species of snakes obtain from their infrared-sensitive organs, the role of smell in avian navigation,6 etc.). Second, if tactual spatiality is as specific to touch as the feelings one gets when applying one’s muscles to do something, it is only through touch that we or any other creature, terrestrial or not, naturally selected or engineered, can have conscious access to space and spatiality.

Touch Imagine there were creatures who coordinate visual appearances with bodily position and movements exactly as we do but whose brains produce entirely different visual appearances from ours. For example, when we look up, our brains produce the visual appearance of new sensation emerging at the (somatic) top, existing sensation shifting downward, and sensation already at the bottom disappearing. By contrast, when

Berkeley’s Vision 71 they look up, their brains produce the exact opposite visual appearance. Clearly, the two appearances disagree visually. The question is whether it is equally correct to describe their disagreement as spatial. If spatiality were intrinsic to vision, the answer would have to be yes: they would be looking down when we look up and vice versa. But if visual spatiality were instead a function of whatever visual appearance happens to correlate with bodily movement, then visual “looking up” is whichever visual appearance is correlated in a given brain with moving one’s body so as to reposition the eyes to point the foveae further in the direction of the body’s topmost part (the top of the head if one is standing or seated upright). In that case, our visual disagreement with these creatures would not be spatial at all, and there would be no spatial difference between what for them is looking up and what it is for us. Indeed, it would be possible for systematic visual disagreement between them and us to coexist with complete spatial agreement and vice versa. To determine which answer is correct—whether spatiality is an intrinsic quality of visual sensation or something it has only through its essentially arbitrary correlation with motor function—imagine hanging upside down, legs hooked around a tree branch. If you use your eyes to look up, you would, as previously, reposition them so as to point your foveae in the direction of your body’s topmost part, in this case the knees. Visually, however, what happens is just the opposite from before: your brain produces the visual appearance of new sensation emerging at the (somatic) bottom, existing sensation shifting upward, and sensation already at the top disappearing. Now, if spatiality were intrinsic to vision, not needing to be referred in any way to motor function, then the hangingupside-down visual appearance would be in spatial disagreement with the visual appearance looking up gives when the head is topmost. Yet, independently of such reference, there seems no reason to regard the one visual appearance as in any way spatially truer than the other. Indeed, this would appear to be the case generally: because no visual appearance contains the least information regarding bodily spatiality as an intrinsic part of its content, there can never be any reason to regard one visual appearance as spatially truer than any other, possible or actual, unless and until they are indexed non-visually to the body.7 The two, the visual appearance effected by the cause of its neural correlate and the body itself, are completely independent of one another, each capable of being represented as existing in the other’s absence. The implication, then, is that far from being intrinsic to visual sensation, visual spatiality of every kind—­directions, orientations, locations, postures, movements, situations, relations, configurations, volumes, and dimensions—is wholly derivative, a function entirely of visual sensations’ correlations to the body, apart from which the visual is altogether lacking in genuine ­spatiality—no more than high pitched vs. low pitched auditory sensation is a genuine instance of spatial up vs. down.

72  Path to Kant: Psychologism and Conventionalism This is the first component of Berkeley’s theory of vision. Its corollary is that if visual spatiality is derivative from motor function, then there must be a distinct, non-visual representation of space from which the spatiality of visual representation derives. Yet, so long as the representation of space remains a matter of correlations to body parts, their posture and movement, then the same problems arise as with vision. For example, our brains are so constituted as to produce a certain auditory sensational consciousness when sound strikes the left ear slightly earlier than it does the right and another auditory sensational consciousness when the reverse occurs, so that, spatially, the first is for us auditory leftward, the second auditory rightward. But there is no reason there could not be creatures otherwise identical to us whose brains produced precisely the opposite sensational consciousness in the same circumstances. Clearly, the two appearances disagree auditorily. But the question again is whether it is correct to describe the disagreement as spatial. If spatiality were intrinsic to auditory sensational consciousness, the answer would have to be yes. But if it is instead a function of whatever auditory appearance happens to correlate with what goes on in each ear, then auditory leftward is whatever auditory appearance is correlated in a given brain with sound reaching the left ear slightly before the right, and correspondingly for auditory rightward. In this case, our auditory disagreement with these creatures would not be spatial at all, and there would be no difference between what for them is sounding from the left and what it is for us. Indeed, just as with vision, it would be possible for systematic auditory disagreement between them and us to coexist with complete spatial agreement and vice versa. To decide between the two alternatives, it suffices to turn around and face the opposite direction, so that the same sound reaches the right ear before the left. If spatiality were intrinsic to auditory sensation, not needing to be referred in any way to the body, then the auditory appearance we get after turning around would be in spatial disagreement with the auditory appearance that preceded turning. Yet, independently of such reference, there seems no basis for regarding the one auditory appearance as spatially truer than the other. So too generally: because no auditory appearance contains the least information regarding the body as an intrinsic part of its content, there can never be any reason to regard one auditory appearance as spatially truer than any other, possible or actual, unless and until they are non-auditorily indexed to the body. Since this is just to say that the auditory appearance and the spatial position or movement of the body are completely independent of one another, it follows that far from being intrinsic to auditory sensation, spatiality of every kind—directions, orientations, etc.—is wholly derivative, a function entirely of auditory sensations’ correlations to the body, apart from which the auditory is altogether lacking in genuine spatiality. And the same reasoning applies to the far more sensitive hearing of echo-locating

Berkeley’s Vision 73 bats, the echo-sonar sensations of cetaceans, and all other modes of representation, sensational or not, that depend on correlations to body position, posture, and movement. Insofar as the spatiality of vision or any other sense is not intrinsic, a feature it has only correlatively, at second hand, there must be an independent source of spatial representation from which that spatiality originally derives. This representation must furthermore be specific to the spatiality of the body, i.e. representative of nothing else, since that is the only way to eliminate the kind of problem just described, stemming from the inherent arbitrariness of all correlation of sensations to the body. The physical spatiality of the body itself, however, cannot be the source of this representation. First, its mere existence does not magically call into being the neural correlates necessary to produce genuinely (intrinsically) spatial representations in consciousness. In their absence, the spatiality of the body is as absent from consciousness, even consciousness at its darkest, as the spatial structure of proteins in one’s spleen cells’ cytoplasm, and so cannot be the source from which vision and other senses derive their spatiality. Second, the spatiality of the body extends to parts and scales that contribute little or nothing to the spatiality of visual, auditory, or any other kind of sensation: internal organs and tissues, individual cells and their components, not to mention their molecular, atomic, and subatomic sub-components. Apart from consciousness and its neural correlates, there is nothing to define which physical spaces are and are not “parts” of a body, or even to define the body as a single unified “whole,” much less to select which physical spaces at which scales will and will not enter into the spatiality to which visual, auditory, and other sensations must be correlated in order to acquire their second-hand spatiality. The representation of the spatiality of the body relevant to the spatiality of vision, hearing, and other senses must therefore be given by a unique, specially dedicated modality of sensational consciousness continuously present while we are awake, even if only darkly. And according to Berkeley, this is none other than touch.8 Berkeley readily conceded that we rarely if ever discern or attend to the visually relevant feelings issuing from the eyeballs indicative of their position and whether, saccading aside, they are moving or being held in position; the feelings issuing from the neck, spine, and limbs indicative of their posture and situation relative to other parts; the feelings whether and how movable body parts are in motion or holding their position; and the feelings indicative of locomotion, balance, homeostatic state, et  al. Nevertheless, he held that these feelings are never entirely absent while we are awake, and form part of a greater tactual field that includes feelings in parts of the body that rarely if ever correlate to visual and other sensations. Together these feelings give us our representation of the spatiality of the body. Having no feeling of any of one’s cells’ mitochondria, mitochondria form no part of that representation. If the feelings in one’s

74  Path to Kant: Psychologism and Conventionalism limbs suddenly disappeared, they would not figure in it either. If meterlong antennae suddenly sprouted from one’s head with a full complement of neural correlates for feeling, they would forthwith become as much a part of the spatiality of one’s body as the eyes, lips, head, torso, and limbs. It may well be that attention is paid only to these feelings when some sudden impingement, pain, itch, or the like induces us to direct attention to a locus in the tactual field. After all, the energy expended in the neural effort requisite for discernment and attention is no doubt biologically costly, making it in creatures’ interest to generally restrict access consciousness to more pressing exterior-directed goals like escaping predators, obtaining nourishment, finding mates, navigating obstacles, listening, scanning, keeping quiet and hidden, etc. But even when feelings are consciously accessed, it is always attention directed to an itch located at the knee, a pain located in a tooth, a pressure located in the small of the back. Attention is seldom if ever directed at such feelings tout court, in a state of total detachment from their bodily spatiality, with no consciousness whatsoever, however inchoate, of any locus in the tactual field. Nor does directing attention away from the body mean the loss of all feeling (sensation) of the lips, hands, eyes, and the rest of the body (perfect tactual insensibility), so that we become in effect numb to even having a body, our minds as completely detached from their tactually accessible components as from inaccessible ones like the pineal gland, liver cells, mitochondria, and ribosomes. So far as Berkeley was concerned, the tactual field that comprises bodily spatiality is a classic instance of dark consciousness: a constant presence to feeling, an integral component of all waking consciousness, even when, as typically occurs, nothing in it is discerned or the focus of attention. Touch scales space directly to the body, be it human, baleen, or the sub-amoeba-sized Megaphragma mymaripenne wasp (assuming tactual sensation to be as widespread in the animal kingdom as visual and olfactory). It spatially articulates the body into parts and subparts, adjacent and separated, surfaces and interiors, configurations, postures, orientations, directions, and sundry spatial relations. Although not directly sensitive beyond the body’s surface, it also provides indirect access to exterior space via feelings of bodily movement, particularly locomotion. Accordingly, by the motion or application of a limb, beak, antenna, or other body part, (dark) tactual sensational consciousness enables creatures to take the measure of their environs, including (1) greater or lesser distance, (2) higher or lower situation, (3) erect or inverted orientation, (4) up, down, or sideways direction, (5) boundaries (size), (6) boundary contour (shape), (7) spatial distinctness and relation (outside/inside, occupying/containing), (8) fewer or greater numbers of the spatially distinct, (9) rest or motion, and, with all this, (10) tridimensionality. Thus does the intrinsic spatiality of touch become the basis for extending

Berkeley’s Vision 75 spatiality associatively in thought to intrinsically non-spatial senses like sight and hearing.9

The Spatial Mien The same could be said of space that Augustine famously said of time: “What is it? If no one asks me, I know; but if asked to explain it, I don’t.” We all, of course, know what “space” means—the kinds of things meant by the word as well as its secondary meanings (space on one’s schedule, etc.). But that is representation by agreement, socially established linguistic convention, that can tell us nothing about the space present to consciousness prior to and independently of language. Is this purely representational space common to humans and other creatures or unique to humans? Is there only one kind or several? If many, how are they different? Is there any that is primary and original in respect to which all others are merely secondary and derivative? If there is a primary spatial representation, is its primacy reflected in the meaning of “space”? And if it is not, what does that tell us about language and its reliability as a guide to conscious representation? Everyone presumably accepts that tactual sensation acquaints us with space most immediately, and few would hesitate to extend tactual spatiality to animals, even quite primitive ones. But that, as we have seen, is where Berkeley drew the line: touch, broadly construed, is the only sense capable of acquainting us, or any other creature, with the spatial; and that is because, on his account, space is as purely and exclusively tactual as itches, warmth, balance, and aches. Correspondingly, vision, hearing, or any other sense, known to us or not, can be spatial only via correlations of its data with touch, and so merely secondarily and derivatively. Granted, few concur with Berkeley. But this much at least seems true: the burden of proof lies on those who would extend spatial representation to kinds of sensational consciousness other than tactual with the same immediacy as touch, and, more so still, on those who would extend it beyond sense entirely to representation of unqualifiedly objective, mindindependent space and/or spatial entities. Berkeley identified four obstacles in the way of extending spatial representation beyond touch to vision or any other sense, much less beyond the senses, that he believed can be neither overcome nor circumvented. First, there is the complete qualitative incommensurability between tactual sensation and other senses. Vision, for example, presents us uniquely with intangible color and light, while touch presents us uniquely with invisible feelings. A sighted creature lacking touch could analyze its visual inputs till kingdom come without being able to extract the least notion of what anything feels like, including the shape, size, posture, and locomotion of its own body. Conversely, prior to being made to see, the Molyneux man

76  Path to Kant: Psychologism and Conventionalism can no more anticipate from tactual sensation what anything looks like, including space, than he can divine scarlet from hearing a trumpet or burgundy red from drinking the eponymous wine. Sensing space tactually gives not the least clue as to what the spatial looks like, just as sensing it visually gives no inkling as to what the spatial feels like; the former is as absolutely invisible as the latter is absolutely intangible. If it is objected that the same spatial vocabulary applies equally to both, Berkeley did not deny it. However, he attributed this not to any qualitative commonality between the visual and tactual, spatial or any other, but to (dark) consciousness of the systematic correlations between the sensations of the two senses, even at their most fine-grained, that make each a well-nigh perfect predictor of the other. We can reliably expect how visual appearances will behave if we feel our eyes turn leftward in their sockets, and we can equally reliably expect how the shape of an object we see will feel if we take it into our hands. How then can we avoid the conclusion that, vocabulary notwithstanding, visual and tactual space, considered in their own right apart from all correlations, are as incommensurable as feeling and color generally? Second, if sight and touch truly were identical in respect to space, then, like identicals generally, they should be intersubstitutable. But when the visual is substituted for a component of a purely tactual spatial manifold or vice versa, the spatiality is not merely affected but nullified: By the distance between any two points nothing more is meant than the number of intermediate points. If the given points are visible, the distance between them is marked out by the number of the interjacent visible points. If they are tangible, the distance between them is a line consisting of tangible points. But if they are one tangible and the other visible, the distance between them neither consists of points perceivable by sight nor by touch, i.e. it is utterly inconceivable.10 When I gaze at the horizon demarcated by sky and plain, I cannot substitute anything tactual for the blue of the sky—not because it is difficult, or because it obscures the relation, but because the very notion of such a substitution is unintelligible. We simply have no idea what to do—any more than we know how to substitute a piccolo’s highest tone in place of the blue to preserve the above–below relation of sky to plain.11 Such substitutions should pose no problem if the spatiality that vision and touch are supposed to have in common really were something immediately present to consciousness in sensation (a sensational commonality, a Lockean idea of diverse senses). Thus, it can only be because tactual and visual sensations are wholly incommensurable, spatiality not excepted, that the very notion of their intersubstitutability is unintelligible. Third, just as visual and tactual space seem to have nothing in common qualitatively or relationally, the arbitrariness of their correlations

Berkeley’s Vision 77 argues against their being isomorphic in any way. For example, if one feels one’s eyes turning leftward in their sockets, there is a corresponding visual sensation. Yet, is not necessarily that one; it is only because one’s neurons happened to be networked the way they are, or because of some other feature of one’s physiology, that that visual spatial appearance is correlated with that tactual spatial appearance and no other. If one’s brain were relevantly different, then instead of that visual appearance, it might be correlated to another, even one directly contrary to it, or to something non-visual, or to nothing at all. But if any visual spatial appearance will do as well as any other as correlate to a given tactual spatial appearance, then there can be no fixed, non-arbitrary correspondence between any components we might distinguish in them. Visual and tactual spatiality must therefore be conceded to be completely non-isomorphic.12 Finally, if, as suggested in the preceding chapter, sensational consciousness is a complete informational blank slate regarding the causality and constitution of its neural correlates, then it must be equally blank regarding their correlates’ spatial constitution and spatial relations (both to other neural correlates and generally). There is no cache of neural information secreted in dark sensational consciousness, and so too none regarding any spaces and spatial relations the neural correlates of visual and tactual sensations may have in common. Thus, if it is granted that there is no basis in tactual or visual sensations themselves—qualitative, relational, or isomorphic—to justify identifying visual and tactual space, then it must also be admitted that their neural correlates do nothing to make up the shortfall by introducing (non-sensational, physical) spatial information to consciousness via the conduit of sensation. Faced with these obstacles, Berkeley saw no escaping the conclusion that touch alone among the senses is immediately and originally (i.e. non-derivatively) spatial, that the space that presents itself to sensational consciousness is as specific to touch as itches and chills, and that visual and all other non-tactual sensations are immediately and originally nonspatial. Since this makes the latter wholly dependent for their spatiality on their (inherently arbitrary) correlations to tactual sensations, and therefore spatial only secondarily and derivatively, visual spatiality must be acknowledged to be fundamentally analogous to mien: As we see distance, so we see magnitude. And we see both in the same way that we see shame or anger in the looks of a man. Those passions are themselves invisible, they are nevertheless let in by the eye along with colors and alterations of countenance, which are the immediate object of vision: and which signify them for no other reason than barely because they have been observed to accompany them. Without which experience we should no more have taken blushing for a sign of shame than of gladness.13

78  Path to Kant: Psychologism and Conventionalism Making the intrinsically invisible visible is by no means limited to tactual spatiality: emotions, passions, desires, thoughts, intentions, and similarly invisible contents of consciousness are visible from the color, posture, and other features of a person’s mien, particularly the face. This, of course, is not because visible mien and invisible inner representational contents have some common quality, intersubstitutable relation, or non-arbitrary structural isomorphism. Nor is it because inner representational contents are any less informational blank slates regarding the causality and constitution of their neural correlates than sensations. Indeed, had humans evolved differently, frowns might have expressed happiness, smiles sadness, and laughter grief; or distractedness, attentiveness, and shock; or auditory differences rather than inner ones; or nothing at all. These connections are always arbitrary, never necessary, at least so far as our consciousness of them in ourselves and others is concerned. There thus seems no alternative to distinguishing arrays of color and light as originally and immediately visible (= the primary objects of sight)—visible from the moment we first open our eyes—from mien as visible only in a derivative, secondary, learning-dependent sense (“. . . barely because they have been observed to accompany them. Without which experience . . .”). But there is more to Berkeley’s analogy than that. It no doubt takes great visual acuity to see a blush. It is dependent on ambient lighting, it is invisible to the myopic, and, most importantly, it requires that one’s visual attention be directed at the face. Nevertheless, even under optimal viewing conditions, a creature with the same visual acuity as us but a complete stranger to the expression of emotion and other inner contents of consciousness in the mien of Homo sapiens would be quite incapable of using its eyes to see it as we do. If intelligent, it might learn to infer particular inner contents from visible features of the human body and predict certain behaviors in consequence, in much the same way we may infer that a tree’s roots are diseased from the sight of its foliage. Still, it would be what might be called “mien blind”: no more capable of visually experiencing a human redden with shame, grin with joy, politeness, or menace, look hesitant or thoughtful, etc. than non-equines are capable of experiencing a mare in estrus as a stallion does. It is a form of blindness that cannot be remedied by glasses or other ocular correctives. The problem is not that it is missing visual content but an incapacity to carve up and parcel out, both synchronically and diachronically, the same visual contents we have so as to optimize their correspondences to emotional and other, equally invisible inner representations. No doubt our species’ particular brand of sociality favored the evolution of neural mechanisms that enable us to discern the relevant contents, collect them together, and represent them in a single unified visual representation with a determinate inner signification. A creature equally intelligent but with sufficiently different lifeways would likely have evolved mechanisms for carving up and parceling out the same visual contents in ways so radically unlike

Berkeley’s Vision 79 ours as to be unimaginable by us. It is these different ways, then, that constitute the difference between genuinely seeing emotions and other inner representations directly in a person’s mien and merely inferring them from visible evidence (on the basis of past experience, i.e. observed correlations)—even if, in the final analysis, the former is “seeing” only in a secondary, derivative sense. What is true of mien seems equally true of the visual representation of the human face itself. After all, before one can see whether a person is smiling benevolently or malevolently, one has to see the smile; and, in general, to see the inner on a face, or even merely its outward beauty or plainness, one first has to see the face. This means carving up and parceling out visual contents so as to group together eyes, brows, lips, and the rest in a single representation; and to do so not just any which way, as a collection united merely by being attached to the front of the head, or even the same collection with its characteristic configuration, but as an ensemble with a unity all its own. There is substantial evidence that humans evolved neural mechanisms dedicated to precisely this purpose. If so, a different phylogeny might have resulted in creatures visually identical to us but with mechanisms that carve up and parcel out the same visual contents in quite different ways, with the result that they, e.g., see human faces as we see the front of the head of insects, or as we see a foot, a rocky outcrop, or something altogether unsuited to being collected together and represented as a unity in its own right (such creatures would be “face blind,” and so insusceptible to pareidolia, flashed face distortion, and similar such phenomena). Nor do these considerations end with human faces. Because the possible ways of carving up and parceling out the contents of the visual field are for all intents and purposes infinite, there could be creatures with neural mechanisms so radically different from ours that their ways of experiencing the very same visual sensational contents we have would seem to us to verge on pointless, random, or worse. For example, a creature might be naturally equipped or engineered with neural mechanisms that made it more natural for it to couple the upper curve of the visible right eye with the visible lower lip instead of the lower curve of the eye, still more natural to combine both together with the third toe of the left foot, and most natural of all to regard these three together with a tree visible in the distance as a unity in its own right. To such a creature, our ways of carving up and parceling out the very same visual contents would no doubt seem as pointless, random, or worse as its ways seem to us. Yet, given sufficiently different lifeways and/or conditions of life, who is to say that creatures as or more radically different from us in these respects could not survive and thrive as well as or better than our species? We would be inclined to describe the result of how they carved up and parceled out visual contents not just as mien-blindness or face-­blindness but as a generalized object-blindness. But they might in turn find it

80  Path to Kant: Psychologism and Conventionalism unaccountable how we could be so blind to all the objects they see and nevertheless somehow manage to prosper in the struggle for existence. Given that mien-blindness, face-blindness, and at least the possibility of generalized object-blindness are capable of coexisting with visual sensational contents identical to ours, isn’t the same possible with regard to the most fundamental of all secondary, derivative species of vision— blindness to visual spatiality? Visual spatiality is fundamental because it is presupposed by all other secondary, derivative species of sight but does not itself presuppose any of them. If I cannot visually experience faces I  cannot visually experience mien; if I  cannot visually experience eyes, nose, lips, et al. in their proper gestalt, I cannot visually experience faces; and if I cannot visually experience objects generally in the ways creatures with our phylogeny are wont to carve up and parcel out visual data, then I cannot visually experience eyes, nose, etc. But how could I visually experience any of these if I could not, in the first instance, spatially carve up and parcel out the true primary objects of vision—color and light—so as to visually experience their size, shape, situation, orientation, relation, distance, tridimensionality, et  al.?14 Visual spatiality, as Berkeley recognized, is the fundamental stratum of secondary, derivative visual representation atop which each subsequent stratum of secondary objects of sight is overlaid. All strata, visual spatiality not excepted, are constituted by their particular correlations with the invisible, and the tangible above all. For while mien-sight and face-sight incorporate emotions and other inner representations, it is their correlations with the face and body of tactual feeling that is their primary support (e.g.  anger is experienced as being as much an affair of touch, broadly construed, as non-tactual emotional affect). Similarly, object-sight depends largely on correlations with uniquely tactual features such as solidity and its lack, soft/hard pressure, smooth/rough texture, warm/cool, wet/dry, pains/ticklings/twinges/etc., and so on. So, it is hardly surprising that space-sight as well is wholly dependent on correlations with tactual spatiality. None of this is to deny that there are populations of neurons in the human visual system dedicated to detecting spatial patterns of various sorts, synchronic as well as diachronic—lines, angles, edges, depth, motion, and much else besides. But the use of spatial terminology by researchers is not evidence that visual contents are spatial independently of their correlations with touch. If, for example, the Molyneux man’s brain were rewired so that when first made to see, he felt the same tactual sensations we do when our eyeballs trace a circle whenever his eyes scanned the horizon and other phenomena we regard as rectilinear, would he not suppose himself to be viewing circles? And if everyone were wired as he was, would neuroscientists not regard the same neural populations they currently take to be dedicated to seeing rectilinearity to be allocated

Berkeley’s Vision 81 instead to circularity? If indeed touch prevails in all matters spatial, as Berkeley contended,15 the answer must be affirmative.

Vision as a Language of Tactual Spatiality As important as it is, the analogy with mien takes second place in Berkeley’s theory of vision to another object of comparison: language. The analogy starts with the recognition that language, like mien, is a secondary object of sense. Berkeley focused on spoken language (hearing) and so shall we, but it is worth noting that the same considerations apply to sign-language and writing (sight), and would likely continue to apply if language could be communicated olfactorily, gustatorily, or via sensations of a kind unknown to us. The primary objects of hearing are sounds. If the sounds are those of spoken language, then, according to Berkeley, the primary objects of our hearing would be the very same acoustic contents that would be present in the auditory consciousness of a creature with the same auditory acuity we have but altogether lacking in language (Homo heidelbergensis? H. habilus?). Nor would he regard it as an objection that we cannot not hear language when we attend to speech in our native tongue, for mien-sight is no less involuntary—just try not seeing the agony of someone you observe screaming and writhing on the ground, horribly mangled and hemorrhaging, after being struck by a car; and then try convincing yourself that you do not really see and hear his pain but merely infer it. English morphemes, words, phrases, sentences, and narratives as well as the contexts that inform them and the things to which they direct attention are no more primary objects of hearing than of sight, smell, or any other mode of sensational consciousness. Rather, hearing them requires the same kind of special, phylogenydependent carving up and parceling out of auditory contents that mien requires with visual contents, except that in the case of language it is directed at optimizing correlations not with emotions and other invisible inner representations but with intrinsically inaudible linguistic representations (which belong to the non-auditory, non-sensory sphere of conventionally established vernacular usage, and so are dependent on the social infrastructure in which convention-governed behavior is generally embedded: see Chapter 5). The importance of optimizing correlations with linguistic contents in order to hear speech is perhaps easiest to appreciate when it is lacking. Words that took hours to pronounce, producing sentences requiring days, would obviously not optimize the correlations between the primary objects of hearing and what one wants to say with them. Or speech that was so difficult to enunciate that its constituents were always mispronounced, if pronounceable at all; speech without enough parts or with too many parts to express what one wants to say unambiguously, if at all;

82  Path to Kant: Psychologism and Conventionalism speech used to express different things that sounded too similar to reliably distinguish, or distinguish at all; and so on. In general, vocalizations that could not be used to say anything we might want to say in a given situation, or in any situation that might arise, would fail to optimize the desired correlations. Indeed, if vernacular languages failed sufficiently badly in this regard, we might not even classify them as languages at all. After all, one of the principal reasons we would not classify mien-sight (“body language” included) as a language of emotion and other inner representations is because its capacity to express the goings on in our minds (feelings, thoughts, et  al.) is so limited and imperfect. One can imagine a world where nothing is hidden, where each thing going on in one’s mind is correlated to a distinctive, easily recognizable, undisguisable visible bodily posture, coloration, or movement. In that case mien probably would deserve to be classified as a visible language of the invisible inner. As things stand, however, while there are many visible signs of the inner apparent on faces and bodies, these signs do not rise to the level of a language because they lack “the articulation, combination, variety, copiousness, extensive and general use and easy application of signs . . . that constitute the true nature of language.”16 Much the same is true of auditory spatiality: although many of the primary objects of that sense are serviceable as signs of tactual spatiality (mainly of direction and distance), there is not, in humans at any rate, the kind of optimizable systematic correlation where for every tactually sensible space, even the most fine-grained, there is a distinct, easily recognizable primary auditory object. The same is even more true of olfaction and gustation. This leaves sight as the only viable candidate in humans for a genuine sensational language of tactual spatiality.17 It is worth remarking that, on Berkeley’s conception of language, visual spatiality is a better exemplar than vernaculars like English. The latter have to adapt to meet the new expressive needs resulting from the changing culture and lifeways of their speakers, and so differ semantically, grammatically, and even logically, over time as well as across space. Nor is any single vernacular, extant or extinct, capable of expressing everything all the others enable their speakers to express, much less all that vernacular language is potentially capable of expressing. Sight, by contrast, has been essentially the same in all human populations since the human optical system first evolved, and, at least where correlations with tactual spatiality are concerned, seems largely indifferent to changing cultures and lifeways. This means that everyone everywhere carves up and parcels out the primary objects of sight in the same way to achieve the same end: the optimization of correlations with tactual sensations and, most fundamentally, tactual spatiality. The result is a unique, easily recognized “look” able to visually mark out every space encounterable tactually, leaving none unexpressed or undifferentiated from every other, even at their most finely grained. It also means that the

Berkeley’s Vision 83 looks corresponding to features that, in tactual sensation, are overlaid upon tactual spatiality are analogously “overlaid” on visual spatiality— e.g. the look of wetness is never seen except as a feature of the look of a particular spatial expanse, say, a puddle of a certain size, shape, and depth; and, in general, no change of color or light is ever seen except as a variable feature of the look of some portion of visual space. Moreover, whereas vernacular languages are quite limited in the number and variety of sounds they employ, and their sounds need to be discrete if they are to be easily recognized and effectively transmitted, the visual language of tactual spatiality is unlimited in number and variety—e.g.  the way color and light differences shade off into one another is a feature not a bug. Thus, the visual signs used to signify tactual spaces not only have “the articulation, combination, variety, copiousness, extensive and general use and easy application of signs (all which are commonly found in vision) that constitute the true nature of language,” the language they constitute eclipses any vernacular, actual or possible, in all these regards. Berkeley attributed the fact that visual spatiality—and the visible generally insofar as it rests on the foundation of visual spatiality—is never regarded as a language to some of the very features that make it so exemplary a language. To convince us, he asked that we consider how vernacular language would likely be regarded if it too had the features that induce us to overlook or deny the linguistic nature of visual spatiality: “if there was one only invariable and universal language in the world, and . . . men were born with the faculty of speaking it, it would be the opinion of many that the ideas of other men’s minds were properly perceived by the ear, or had at least a necessary and inseparable tie with the sounds that were affixed to them.”18 To make the analogy still closer, suppose that when new expressive needs arose with changing circumstances, language faculty mechanisms would immediately produce sounds with just the right significations to meet these needs, produce the same sounds in everyone who felt those needs, and allow those hearing these sounds for the first time to immediately comprehend their linguistic significance. Suppose also that there were no non-linguistic sounds, i.e. that every auditory sensation without exception had linguistic significance. And to make the analogy with sight complete, suppose finally that either humans were the only animals with hearing or that all non-human animals with hearing also understood this language. Would it not then be all but inevitable that we would regard the sounds derived from our language faculty as no less inherently linguistic than what they signify, and so regard language as an immediate and primary object of hearing in the same way we—equally mistakenly—regard space as the immediate and primary object of sight? The point of the thought experiment is that in matters so fundamental as the spatial character of all things visual we can rely neither on introspective scrutiny nor on the presence or absence of local (e.g. cultural)

84  Path to Kant: Psychologism and Conventionalism variation to decide whether it is primary or secondary. Its nature as a language is belied neither by its universality, its being everywhere determined by the neural mechanisms that make up the human optical system, nor by the unrivaled extent of its articulation, combination, variety, copiousness, et al. Yet, if its having that nature is granted, then the original dilemma returns with re-doubled force. Either one must, like Kant (Chapter 8), find means to neutralize the implication of Berkeley’s arguments that visual and tactual consciousness have no qualitative identity, no relational intersubstitutability, no isomorphism of any kind, and are blank slates where their neural correlates are concerned, or, failing that, one will have no option but to grant his conclusion that the space that presents itself in sensational consciousness is as specific to touch, broadly construed, as a tingle or air hunger, and that visual and all other non-tactual sensations are primarily and originally non-spatial. And the implication of this is that the latter are wholly dependent for their spatiality on correlations to tactual sensations that, in the final analysis, are as arbitrary as the correlations between English vocalizations (or visible ASL signs) and English itself (ASL itself).

How Consciousness Becomes Cognitive via Associative Imagination Berkeley’s analysis of visual spatiality as a language of tactual spatiality is crucial to his determination of multi-sensory spatial consciousness as the doing of neither sense nor intellect but imagination. Prior to Berkeley, imagination was generally credited with two roles in conscious representation. First, it rearranges materials garnered from the senses and stored in memory to simulate unexperienced situations: a bird seen on the ground imagined as taking flight, fantasizing winning the lottery or being assassinated in the Roman Forum, dreams, etc. As such, imagination represents nothing but possibilities (not actualities), fictions (not facts), and fantasies (not reality). Second, imagination associates otherwise unrelated representations. Since this includes the power to associate sounds with language, imagination was credited with producing precisely the kind of signifying relation Berkeley had in mind when he compared visual spatiality to language. For given that the relation of auditory sensations (the primary objects of hearing) to linguistic correlates (secondary objects of hearing) is not founded on any resemblance, intersubstitutable relation, or isomorphism of any kind, and the sensations are complete informational blank slates regarding the neural, the two can have no unity other than the association in imagination whereby the occurrence of the one in consciousness immediately triggers a transition in thought to the other. Berkeley’s theory of vision fundamentally transformed the prevailing conception of the imagination not by changing it—indeed, its being

Berkeley’s Vision 85 unchanged is essential to the theory—but by extending imagination into the cognitive sphere, thereby for the first time crossing the line separating reality from fiction (where Hume and Kant would follow). This is because the ability to represent space as transcending the divide between sight and touch is indispensable to all cognition of the physical. If it is only in imagination that a space that would otherwise be a uniquely tactual representation is able to span the divide separating the tactual from the visual to become as much the object of sight as of touch, then not only does imagination deserve to be regarded as a source of veridical representation, it must be recognized as no less fundamental to the cognition of the physical than intellect. The core of Berkeley’s theory of vision is the idea that the transformation of the primary objects of sight into a language of tactual spatiality effectively fuses sight with touch into what is functionally and phenomenologically a single multi-sensory external sense. “External” here is interchangeable with “spatial,” and so applies just as much to our awareness of the space in and under the skin through touch as the space beyond our skin through sight. It is a “sense” in a broader sense than usual because it includes secondary as well as primary objects of sense, and so is constituted as much by imagination as sensation. Sight and touch together form a single external sense insofar as their correlations rise to the level of a language, where nothing tactual is visually inexpressible and everything visual can be carved up and parceled out so as to carry tactual significance (images like Escher’s highlight this precisely by contradicting something in our custom-bred visuo-tactual expectations). In this way, long before we are out of our cribs, the two senses have become so closely and completely integrated that it makes less sense to regard them as separate, independent senses than as a single, multi-sensory external sense, and no sense at all to confound visual spatiality with a species of inference (intellect) rather than genuinely visual experience simply because its objects are secondary and are as much the doing of imagination as sight, just as in seeing another’s acute embarrassment. Nor is this imagination-generated unitary external sense limited to sight and touch: if I see the same spatial objects I touch, I also hear their sounds to the left, right, above, below, near or far, and may smell them well enough to detect their distance and direction (e.g. aromas wafting upstairs from a ground-floor kitchen). How can hearing and smell be integrated into the same external sense as sight and touch when their primary objects are neither intrinsically spatial nor capable of rising to the level of a genuine language of tactual spatiality? Berkeley’s answer was that their primary objects can be incorporated into the visual language of tactual spatiality insofar as they admit of being carved up and parceled out in ways that optimize their correlations either directly to tactual spaces or to visual contents that have themselves been carved up and parceled out to optimize their correlations to tactual spaces. They thus become for

86  Path to Kant: Psychologism and Conventionalism all cognitive intents and purposes synonyms of their tactual and visual correlates, whereupon they can be integrated into the language to which the latter belong, not unlike the way innumerable Latinate synonyms of existing Germanic words were integrated into English after the Norman conquest. Tastes too can be incorporated into the language as ancillary signs insofar as they are represented as located in things (fig flavor in the visuo-tactual fig), and even consciousness itself insofar as it is situated in a visuo-tactual human or non-human animal head. The result is a single, all-encompassing external sense that enables one and the same spatial object to become an object to all our senses and even, in certain cases, to be represented as a locus of non-sensational features such as emotions, thoughts, and other inner representations. And it is precisely these sense-divide transcending spatial objects of associative imagination that are afterward attributed physical reality by the understanding and become the intellectual focus of all empirical cognition, scientific no less than ordinary.

The Associative Experience The experience of using the visual language of tactual spatiality is less dissimilar to that of using English than one might suppose. Berkeley thought this particularly evident in the case of reading: signs being little considered in themselves, or for their own sake, but only in their relative capacity and for the sake of those things whereof they are signs, it comes to pass that the mind often overlooks them so as to carry its attention immediately to the things signified. Thus, for example, in reading we run over the characters with the slightest regard and pass on to the meaning. Hence, it is frequent for men to say they see words notions and things in reading a book, whereas, in strictness, they see only the characters which suggest words notions and things. And by parity of reason, may we not suppose that men, not resting in but overlooking the immediate and proper objects of sight, as in their own nature of small moment, carry their attention onward to the very thing signified, and talk as if they saw the secondary objects? which, in truth and strictness, are not seen but only suggested and apprehended by means of the proper objects of sight, which alone are seen.19 If we understand the characters visible on a page as primary objects of sight (light and color sensations), then among the secondary objects they may have are linguistic ones. For practiced readers in particular, the mere sight of the characters is enough to trigger an immediate transition in thought to words and meanings, with attention (access consciousness) focused exclusively on the latter, while the former recede into

Berkeley’s Vision 87 the obscurity of dark consciousness. The same is true of all the myriad variety of secondary objects of sight, and none more so than the most fundamental stratum of secondary visual representation presupposed by all the others: visual spatiality. Here, too, those with sufficient experience of sight have only to open their eyes for the primary objects of vision to immediately trigger transitions in thought to the associated tactual “words” and their tactuo-spatial “meanings,” with attention focused exclusively on the latter and the former receding entirely into dark consciousness. It is thus exactly the same as reading: our sensations may be visual, but they are ignored in favor of their invisible associated counterparts, so that our thought is exclusively occupied with language—or mien, or tactual spatiality, or whatever other secondary object happens to be our present focus of attention. Here the question arises of the extent to which association is a conscious process. Nowadays it is generally understood more as a neural mechanism than a species of conscious representation, even to the point of deeming consciousness irrelevant to it. This, however, is not true of the associative imagination that features in Berkeley’s theory of vision and the general theories of cognition of Hume and Kant. The occurrence of one conscious representation—say, a visible square—may be neurally connected to another—a tangible square—such that the sight of the former immediately triggers a transition in thought to the other. But this merely describes a succession in time: a square seen succeeded by the thought of a tangible square. Since there are innumerable other representations in consciousness simultaneous with that thought, what is there to single it out from every other representational successor to the sighting of the square so that it, and only it, is “read” as the secondary meaning of the visible square? Some may answer that a constant experience of one followed by the other suffices to “mold” plastic neurons so as to cause the thought of the tangible square to occur whenever a square is sighted. But that does not answer the question. Given that all conscious representations are, neurally speaking, informational blank slates, neural molding can only account for the presence of the thought of the tangible square among the manifold representations concurrent with it in consciousness subsequent to the sighting of the square. However, unless the neural molding includes in addition a neural correlate for consciousness of their correlation, it cannot consciously single out the tangible square from concurrent representations as the secondary object suggested by the visible square and no other. The difference is the same as what makes the narrative in my thought not just the successor to my sighting of characters on the page but what singles it out as their secondary object, the word read from them: in both cases, neural molding cannot explain my consciousness of the one as a secondary object uniquely suggested by the sensation whereby I perceive the other—or, indeed, consciousness of any successor

88  Path to Kant: Psychologism and Conventionalism representation as the secondary object suggested by a given primary object, and by that primary object and no other. So far as Berkeley and his successors were concerned, there is only one way such “correlation blindness” can be corrected: whenever the occurrence of one representation triggers the immediate transition in thought to an associated representation, there is consciousness not only of the representations but of the transition from one to the other as well. Neurologically, this would mean that at the same time plastic neurons are molded to effect the transition, neural correlates of a consciousness of the transition must be generated as well, or no consciousness of the relation could ensue. As immediate, the presence of such transitions in consciousness is invariably too rapid to discern or attend to them; but that poses no problem since attention is far more profitably directed to the secondary objects to which the transition is made anyway. All that matters is that the transition be present in consciousness at all, however darkly, to ensure that the representations concerned in the transition are singled out in precisely the way required for them to be associatively related in imaginative consciousness. And what is true of a single transition Berkeley and his successors extended to sequences and series of them as well. On the now common mechanistic conception of association, by contrast, there is no need for consciousness of the transition: the neural, electronic, or abstract Turing machine simply makes the transitions and moves on as programmed. Yet, if the transition did not register in any way at the level of consciousness (i.e. had no dedicated neural correlate), how could one associate, say, the characters one is seeing with the linguistic content one is thinking so as to recognize the latter as text suggested by, and so read from, the former rather than simply a disconnected thought randomly popping up in consciousness? With nothing to associate them in imagination, there could be no consciousness of their relation, and the invisible linguistic content could no more be the secondary object “seen” in the visible characters than a non-conscious reading device “relates” the ink marks it scans to the sounds it synthesizes. Association thus requires there to be sufficient consciousness of the transition to single out the linguistic content from everything else concurrently present in consciousness as the relatum of the visual primary object. So, too, the case of association between the primary objects of sight and tactual spaces: in the absence of even the darkest consciousness of transitions from the former to the latter, no association between them could be effected at the level of consciousness; and since this is just to say that the primary objects of sight would have no conscious association with space, space would quite literally cease to be visually experienced (space-blindness, hence also object-blindness, mien-blindness, etc.). To be sure, they might still be functionally spatial at the neurological level. But with no ability to experience tactual spaces as secondary objects of sight, and so no consciousness of any relation between them, the result would

Berkeley’s Vision 89 be at most a variant of blindsight.20 There must therefore be sufficient consciousness of the transition to single out the thought of a particular tactual space from all representations present in consciousness concurrently as the relatum of the non-spatial visual primary object, but not so much that its perception distracts us from the spatial secondary object on which our attention is most profitably focused.21 Why is this important? Granted that something like the association in imagination posited by Berkeley and his successors evolved, its naturally selected raison d’être is clearly not so that humans might speak and read vernacular languages but, most fundamentally, to provide the adaptive advantages a visual, auditory, or other sensational language of tactual spatiality confers on humans and all other creatures that possess it. In particular, if Berkeley’s theory of vision is correct in its fundamental assertions, such a language is essential to a multi-sensory external sense through which one and the same sense-divide transcending object can be touched, seen, heard, smelled, etc. Since the representation of such an object is essential to the intellectual representation of physical objects (both ordinary and scientific), if consciousness in its guise as associative imagination is psychologically indispensable to the former, then it must be recognized as cognitively indispensable to the latter. This means that apart from associative imagination, sense-divide transcending physical objects could not be represented at all. In particular, there could be no purely neurological, non-conscious sensory representation of such objects since there would then be nothing to consciously relate the data of the various senses to one another, or to do so in such a way as to yield consciousness of their relation as constituting access to the same sense-divide transcending objects. Thus, on Berkeley’s analysis, sensible cognition— sensory representation of objective physical fact and real existence—is inseparably bound up with consciousness of associative transitions in imagination. One can only conjecture as to the reasons a multi-sensory external sense might have been favored by natural selection. Among the more likely, it seems to me, is this: in the same way the blanking of the slate that distinguishes the utter simplicity of sensations from the complexity of their neural correlates serves to greatly simplify a creature’s representational burden (Chapter 2), its ability to focus all its senses on one and the same sense-divide transcending physically real spatial object in imagination and understanding—instead of having to expend valuable access consciousness resources to independently keep track of the different, mostly non-spatial objects of its various senses—would do likewise. Since this in turn would enable access consciousness to be more efficiently and advantageously deployed, it would not only facilitate the creature’s interactions with the world but also open up possibilities for increasingly sophisticated behavior that might not otherwise exist. In this respect, the only difference Berkeley’s theory of vision makes is that the

90  Path to Kant: Psychologism and Conventionalism neural correlates of consciousness selected for inclusion in the relevant behavioral loops would include correlates of sensation-to-thought relating transitions. One final remark before we leave Berkeley. His philosophy is best known for its denial of both the intelligibility and plausibility of materialism, a position generally referred to as idealism. It is important, however, to underscore that there is nothing idealist about Berkeley’s theory of vision, which he deliberately, and rightly, segregated from the idealist portion of his teaching. The theory of vision implies only that our representations of sense-dividing transcending objects are dependent on consciousness, but leaves open the question of whether those objects themselves are dependent on it as well. Idealism proper was reserved principally for Berkeley’s Principles of Human Knowledge (1710), while its defense against the realist implications of Newtonian physics and the calculus that underwrites it can be found in works such as De Motu (1721) and The Analyst (1734).22 Since the only idealism that I am concerned with in this book is the Kantian physicalist variety, I shall therefore pass over these parts of Berkeley’s philosophy and proceed directly to the consideration of Hume’s transformation of Berkeleyan associationism into an explanation of objective cognition generally.

Notes 1. Before humans acquired language, anyone who was born blind but, through some fortuitous physiological event, suddenly acquired normal sight would have been wholly unprepared for his or her new experiences. It is of such minds that the theories of vision developed by Locke, Berkeley, and others were meant to apply, for their focus was not on what extensively primed subjects can infer from their new visual sensations but which representational contents are and which are not present in those sensations themselves. Insofar as infants attain the full gamut of visual experience long before they can be coached through language, we were all once in the position of correlating our neonatal visual sensations with our (probably prenatally established) tactual bodily spatiality. 2. I use “touch” in this broadly construed sense throughout. Helpful recent accounts of the various components of touch can be found in Touch: The Science of Hand, Heart, and Mind. Edited by David Linden. New York: Viking, 2015, and How Do You Feel? An Interoceptive Moment with Your Neurobiological Self. Edited by A.D. Craig. Princeton: Princeton University Press, 2014. 3. “[W]henever we make a nice survey of any object, successively directing the optic axis to each point thereof, there are certain lines and figures described by the motion of the head or eye, which being in truth perceived by feeling, do nevertheless so mix themselves, as it were, with the ideas of sight, that we can scarce think but they appertain to that sense” (V §145). E.g. “the actions of turning the eye up and down . . . suggest a very obvious reason why the mind should denominate the objects of sight accordingly high or low. And without this motion of the eye, this turning it up and down in order to discern different objects, doubtless erect, inverse, and other the like terms

Berkeley’s Vision 91 relating to the position of tangible objects would never have been transferred or in any degree apprehended to belong to the ideas of sight: the mere act of seeing including nothing in it to the purpose, whereas the different [tactually sensed] situations of the eye naturally direct the mind to make a suitable judgment of the situation of objects intromitted by it” (V §98; cf. §§16 and 97). Tactually sensed swiveling ears seem to perform much the same role in cats vis à vis auditory spatiality that tactually sensed swerving eyeballs do vis à vis visual spatiality in humans. 4. V §§157–8. See also V §§132–42. 5. V §140. 6. An accessible discussion of avian olfactory mapping can be found in The Genius of Birds. Edited by Jennifer Ackerman. Melbourne: Scribe, 2016, pp. 228–34. 7. This remains true even if the visual appearances are of the body itself. We might imagine a creature with multiple sets of eyes at the end of flexible tubes that enable it to observe everything its body is doing, including its other sets of eyes. But then, of course, there could be a creature similar in all ways except that different visual appearances consistently correlate to the same positions of its eyes and the same appearances consistently correlate to different positions of its eyes. Hence, even visual appearances of the body itself have no intrinsic spatiality, but instead require correlation with the body as it is in itself, i.e. non-visually. Something similar may be true even of insects: “Normally when a fly (or any animal) turns in one direction, the visual scene moves in the opposite direction. If this relationship between action and consequence is reversed by placing the fly in a flight simulator, the fly adjusts within 24 hours. Now when it wants to approach a promising target, it turns away from the target, and voilà, the target enters its field of view.” Principles of Neural Design. Edited by Peter Sterling and Simon Laughlin. Cambridge, MA: MIT Press, 2011, p. 99. 8. The idealism Berkeley espoused in the Principles of Human Knowledge undermines the basis for distinguishing the real physical spatiality of the body from tactual spatiality. My guide here, however, is the New Theory of Vision, where Berkeley was careful to seal off his theory of vision from his as yet unpublished idealism, which is neither advanced nor assumed anywhere in it. 9. It may be objected that not everything I attribute to Berkeley in this section can be attested by specific texts. But even if this were true, there are so few ways to go in elaborating the view that genuine spatial representation is 100% tactual sensation that claiming anything in this section as mine rather than Berkeley’s would indict me of philosophical plagiarism. That spatiality proper is exclusively tactual was Berkeley’s view is provided with scholarly support in KEUU chapter 14. 10. V §112. Also: “To objects of either kind we indifferently attribute the terms high and low, right and left, and such like, denoting the position or situation of things . . . but to define the situation of visible things with relation to . . . any tangible thing or vice versa is absurd and perfectly unintelligible” (V §111). 11. “Men speak in a high or a low key. And this, it is plain, is no more than metaphor or analogy. So likewise, to express the order of visible ideas, the words situation, high, and low, up and down, are made use of, and their sense, when so applied, is analogical” (VV §46). 12. “Those [visual] ideas that now suggest the various magnitudes of external objects to us, before we touch them, might possibly have suggested no such thing: Or they might have signified them in a directly contrary manner: so that the very same ideas, on the perception whereof we judge an object to be small, might as well have served to make us conclude it great. Those ideas

92  Path to Kant: Psychologism and Conventionalism being in their own nature equally fitted to bring into our minds the idea of small or great, or no size at all of outward objects; just as the words of any language are in their own nature indifferent to signify this or that thing or nothing at all” (V §64). But see note §17 later in the chapter. 13. V §65. 14. E.g. “a man born blind, and afterwards, when grown up, made to see, would not in the first act of vision, parcel out the ideas of sight into the same distinct collections that others do who have experienced which do regularly coexist and are proper to be bundled up together under one name. He would not, for example, make into one complex idea, and thereby esteem a unit, all those particular ideas which compose the visible head or foot” (V §110). 15. “[T]he extension and figure which is the proper object of touch . . . is principally, though not immediately taken notice of, when we look at any object” (V §59). 16. Alciphron 4th Dialogue §12. 17. Among the examples of optimization Berkeley offered is the following: “it must be acknowledged, the visible square is fitter than the visible circle to represent the tangible square, but then it is not because it is liker, or more of a species with it, but because the visible square contains in it several distinct parts whereby to mark the several distinct, corresponding parts of a tangible square, whereas the visible circle does not” (V §142). In the same way, an auditory tone composed of four tones of equal duration would be fitter to signify a tangible square than a single sustained tone, four tones of variable duration, or any other auditory pattern. “Fitness” here relates solely to the serviceability of a sign: visible circles could still be used to signify tangible squares, but we would find the sign harder to learn and deploy for the purpose than visible squares. 18. V §66. 19. Alciphron 4th Dialogue §12. 20. Berkeley’s theory of vision suggests an account of clinical blindsight like the following: if the associations with touch extend as far as the lower, primary levels of the visual system, then even if the higher levels requisite for visual consciousness cease to function because of, e.g.,  a stroke, the non-­ conscious parts of the visual system may retain enough of their tactual spatial meaning to enable a blind person to “see” space even without actual visual consciousness. 21. “So swift and sudden and unperceived is the transition from visible to tangible ideas that we can scarce forbear thinking them equally the immediate object of vision” (V §145). 22. For discussion see KEUU chapters 10–13.

4 Hume’s Cement

Hume’s preeminent importance for Kant stems not only from having first raised the problem addressed in the Critique of Pure Reason but also from pointing the way to how psychologism might suffice for its solution. This chapter  provides a synoptic view of Hume’s empirical psychologism and the enriched conception of association at its heart. In contrast to Berkeley’s isolated efforts, Hume applied his associationist psychologism systematically, so much so that it can plausibly be supposed not only to absorb into imagination all the non-mathematical cognitive roles formerly ascribed to the understanding, but to explicate nature itself by constituting a veritable “cement of the universe.”1 Nor was that the end of it, since he extended his psychologism into the realm of the passions with the aim of showing that association plays roles in human social and moral interactions akin to laws of nature. No less importantly for our purposes, Hume saw in associationist psychologism a theory of sufficient scope to explain human and non-human animal mentation through a single set of principles. Only after crossing the divide separating human nature from human artifice as exemplified in such uniquely human institutions as convention-based vernacular language did he find it necessary to narrow its scope to our species alone. Everything else, including our most sophisticated non-linguistic mentation, Hume believed, could and should be explained according to the same associative principles operative in other creatures.

The Affective Dimension of Association I have already remarked that it has become the norm in both philosophy and the mind sciences to regard association as a strictly non-conscious process. Hume, like Berkeley before him, rejected this on the ground that consciousness of a relation between distinct items can be consciously distinguished from consciousness of serendipitous concurrence or succession only if there is consciousness, however dark, of the relation itself; and this is true regardless of whether their concurrence or succession is due to non-conscious association or pure serendipity. For what matters

94  Path to Kant: Psychologism and Conventionalism where consciousness of relations is concerned is not what is or is not related in the brain or external reality but whether there is a relation present to consciousness in addition to the sensations or thoughts concerned in it. Such consciousness can exist only if it has its own dedicated neural correlate additional to the neural correlates of the related representations, any of which, whether sensation or thought, may be supposed to be fully capable of appearing in consciousness in the absence not only of the relation but of the other(s) as well. Extensive non-conscious neural processing may well be required to associate the neural correlates of distinct conscious representations, and moreover be replete with behavioral consequences. But unless that processing includes the production of a neural correlate dedicated to consciousness of the relation—which it might not, and probably does not in some creatures—the relation will be absent from consciousness. In that case, the related items will appear as separate and isolated from one another as each does from anything to which they have no non-conscious association or as they would appear if no non-conscious associative processing had taken place at all. Thus, for Hume as for Berkeley before him, association without (a neural correlate for) consciousness of the associating relation is, so far as consciousness is concerned, the same as no relation at all. Given that association can be a conscious process as well as a non-­ conscious one, what does the conscious part consist in? Berkeley situated it in consciousness of the transition from one associate to another, but since he seems never to have developed an account to show how consciousness of a transition constitutes consciousness of a relation, I shall henceforth confine my focus to Hume. Among the parameters Hume thought need to be taken into account is that the associating quality in transitions from one representation to another has to be so simple in nature that infants and animals, even quite primitive ones, can apprehend it. To him, this pointed to a feeling—not one that accompanies transitions of thought (as emotions may) but rather in the way the transition itself feels to the mind making it. He characterized it as a feeling of facile transition2 and contrasted it both with indifference, where nothing is felt in the transition and consciousness of relation is absent, and with difficulty of transition, which creates so strong a dissociation that a “sensible violence”3 is felt in the mere contemplation of the items concerned being related. Dissociation occurs particularly in cases of transitions contrary to transitions strongly favored by association, i.e. ones endowed with intense facility affect, while indifference is found in transitions where associative and dissociative qualities are feeble or altogether absent. Among the examples Hume used to illustrate what he had in mind is throwing dice. We strongly associate doing this with their falling to the table and rolling until stopping on one of their sides. The transition to any conceivable alternative—their remaining suspended in air, darting

Hume’s Cement 95 about like angry bees, turning into doves and flying off, or anything else we might imagine contrary to their association—is equally strongly dissociated from our thought of what happens whenever dice are thrown, though still purely affectively (“sensible violence”). Moreover, we feel this associating/dissociating quality in varying degrees, corresponding mainly to the frequency and constancy of past experience of particular outcomes. The relative rarity of the dice coming up snake eyes in comparison with their falling to the table and landing on one of their sides makes for far less facility being felt in the former transition, and consequently a far weaker association, than in the latter. The greater frequency of throwing sevens suffices for a more intense facility affect than the transition to snake eyes, but still yields a far weaker association than the maximally intense facility felt in the transition to the idea of their falling to the table and landing on one of their sides. For the latter, being supported not only by a frequent and constant experience but also by its close analogy with throwing, pushing, dropping, or otherwise impelling things similar to dice raises facility feeling to such a pitch as to make the association quite literally irresistible and unpreventable. The dice example illustrates a further feature of facile transitions of such importance that it earned association the status of supreme principle of Hume’s psychologism: the strength of our belief in the outcome is directly proportional to the strength of the association, and so to the intensity of the facility felt in the transition. No matter how long I may have watched people playing craps, I would not naturally form any beliefs regarding the probability of different combinations of faces occurring if their relative frequency of occurrence did not cause me to form associations between them and throwing the dice. It is only insofar as frequent experience causes associations to form and, moreover, to vary in strength according to the constancy of the experience relative to alternative outcomes that beliefs correlate to experience. If instead of frequent, constant experience we evolved to form associations and regulate their strength from entirely different causes, then such experience would be treated not as signal but as noise in belief formation. After all, who is to say that there is nothing “out there” that for us is noise but, in a differently constituted creature living in different circumstances, might cause belief-determining associations that enabled it not only to survive but to thrive as well as or better than us? So, my belief in the relative probabilities of different combinations of dice faces and my “bet-my-life” certitude that the dice will fall, roll, and stop on one of their faces are not based on experience at all, but determined entirely, and uniquely, by the affective intensity of the feelings of facility constitutive of associative transitions. And, in general, facility feeling is the primary basis of all our beliefs in matters of fact and real existence, while experience is never more than one among many causes of their having that basis, and so a phylogenetically relative determinant of belief that operates only indirectly, at one remove.

96  Path to Kant: Psychologism and Conventionalism The arch skeptic Hume must have delighted in taking down human rational pretensions a peg. Beliefs regarding matters of fact and real existence are based not on experience, but on the happenstance that experience in beings constituted like us is the occasion of intense facility feeling. If all such beliefs are based solely in such feeling, and the only way we can make them appear rationally respectable is by feigning that they are based on the causes of these feelings instead, then our pretensions to being in any way higher than animals are well and truly punctured. The causes of facility feeling might have been anything. In humans and similarly constituted beings, these happen to include frequent, constant experience of the sequence or co-occurrence of perceptions. But there are probably myriad circumstances that could have been utilized by evolution as triggers to produce (neural correlates of) facility feelings, and some—­including ones that do not involve frequent, constant experience at all—might have contributed to the fitness of creatures just as positively as our natural facility triggers do to ours. For them, frequent, constant experience would be a matter of complete indifference and could never be used suasively at all. Indeed, their modes of suasion would be a completely closed book to us, just as ours would be to them. But the difference, according to Hume, lies not in the grounds of suasion but only in the causes of those grounds. Belief itself, in all relational matters of fact and real existence, is grounded, in them just as in us, on how facile, indifferent, or difficult transitions of thought feel to the mind making them—and that is all. By shifting the basis of belief in relations from objective experience to subjective feeling, Hume moved the topic from epistemology to psychology where, instead of needing to be justified by evidence and to follow as a conclusion from premises, belief is determined purely affectively, by association-constituting feeling, and nothing else.4 Belief in relations is not, to be sure, the only kind, since we are equally determined by human nature to believe in the reality (actuality, real existence, factuality) of anything presented to consciousness in sensation, whereas everything present to us in thought, at least on human nature’s default setting, is regarded as merely fictitious (imaginary, merely possible, purely hypothetical). The default setting proviso is necessary because there is much that is present to us only in thought that we nevertheless believe to be fully as real as what we sense. Indeed, Hume saw it as the principal task of psychological philosophy to determine the conditions under which we come to believe in the reality of what we merely think but do not sense. His answer was quite general: in nearly all cases, and in all the most important ones, belief is accorded to thought only insofar as what is represented in it is associated, directly or indirectly, with something present to consciousness in sensation. Thus, if I see smoke pouring out of a building but do not see the fire within, their association prompts my mind to transition to the thought of fire with such facility that any other transition would produce a sensible violence; and since the smoke

Hume’s Cement 97 is presented to me in sensation and not merely in thought, the association prompts me not only to think there is fire within but to believe it really exists. So, it is thanks entirely to the psychological magic of association that I am led to fear that anyone inside the building is as much in danger of being incinerated as asphyxiated—notwithstanding that the fire that so rouses my dread is no less purely a contrivance of my imagination than thoughts of Pegasus or cities paved with gold. Memory is another example: only insofar as I strongly associate something currently present to me in thought with something previously present to me in sensation do I believe it the remembrance of something real rather than a fiction newly conjured up in fantasy. Hume took great pains to demonstrate that association with present or past sensation is essential to empirically based belief generally, including beliefs that things and parts of things continue in existence even when they are absent from the senses, beliefs that there is more to space and time than we are conscious of in sensation or memory, beliefs that observed regularities in nature are due to unobserved causes, belief that one’s conscious self persists and has persisted through all change in the contents of consciousness, and so on. For the present, I shall pass over the details of Hume’s analyses to focus instead on the fundamental psychological point: the reason creatures like ourselves, and presumably creatures far more primitive than us, are conscious of reality as encompassing more than what is apparent to them in sensation at the present instant is that they associate it in myriad ways with thoughts of things absent from consciousness AND association has the power to extend the belief we have in the reality of anything that is or has been present to us in sensation to its otherwise unbelieved associate(s) in thought—and to do so in direct proportion to the strength of the association, i.e. the intensity of the facility affect characterizing the transition from one to the other(s). Where association is strongest, creatures are as convinced of the reality of things they merely think as they are of the things they actually sense; where it is less strong, they are convinced of the probable or possible reality of the associated things they think; and where association is so weak as to be suasively ineffectual, they believe their thoughts mere fantasies with no place in the realm of real possibilities at all.

Belief Psychology in Human and Non-Human Animals The role of facility constituted association as the exclusive basis of belief in all matters of fact and real existence, and, more particularly, its power to extend the scope of what counts for us as real beyond what is present to us in sensation, led Hume to pose two questions: what is belief? and which associative principles are most important to it? In this section, I will examine his answer to the first; I will consider his response to the second in the next.

98  Path to Kant: Psychologism and Conventionalism Since sensations take primacy over thoughts and relations psychologically, and (presumably) evolutionarily as well, we first need to consider what induces creatures to believe anything present to them in sensation to be really existent (rather than merely possibly existent, fictive, etc.). Assuming that the first conscious creatures had sensations and nothing else, the question Hume would have us ask is whether there is anything proper to sensations themselves that supplies the information that real existence is represented in them, and therefore accessible through them. While previous philosophers seem to have taken for granted that there is, Hume thought not. Visual sensations are colors and their myriad modifications (bright/dim, shiny/matte, etc.), and nothing else. There is not, additionally, a special color or color modification that cues creatures into the reality of the scene painted in visual sensation. The same applies to tactual, auditory, olfactory, and whatever other sensations may exist: there is no special quality unique to any one sense or common to all that paints everything in sensation in the colors of reality and so persuades creatures to accept it as really existent (actual, factual, true). If belief is not intrinsic to sensations, then it must be added to them; and if not itself a sensational quality, then it must be something that somehow is added to our consciousness of them. The question is what. The first creatures to expand their psychological repertoire by accepting the sensations present in consciousness as reality did not have language. They may have been quite primitive, perhaps not even all that far advanced beyond creatures minimally distinguished from perfect insensibility like Locke’s oyster. If so, then adding reality to their consciousness of sensations can only mean adding some new affection—“feeling,” “sensation,” “sentiment”—that alters how they are conscious of them while not in any way altering what they are conscious of. In particular, just like the facility felt in transitions of thought, belief feeling is neither anything object- or property-like that can be observed by consciousness as things in sensation can (color, feel, odor, etc.) nor a feeling that accompanies consciousness of something in sensation like terror, curiosity, rage, etc., but is instead part and parcel of the affective dimension of consciousness itself—a “manner” of apprehending something present in consciousness. Belief feeling is accordingly free to imbue consciousness not only of sensations but also of thoughts associated to sensations, consciousness of internal objects like emotions and desires, and, in general, anything else capable of presenting itself as real—matter of fact or real existence—to the conscious psyche. Moreover, having the nature of affect rather than some special act of thought means that belief can vary in intensity. Thus, as belief affect rises from zero to maximum and then later falls back to zero, one and the same thought may start out unbelieved, later come to be believed to represent a real possibility (e.g. after its object is encountered in sensation once), still later become a real probability (many

Hume’s Cement 99 encounters), and finally attain so complete a certainty that the reality of its object cannot be doubted—only then, thanks to counterexamples and various other belief-diminishing factors, to gradually lose so much belief feeling as to end up as unbelieved as when it first arose. Viewed from an evolutionary perspective, the addition of (neural correlates of) belief affect may have manifested itself by strengthening the behavioral impact of the contents of sensational consciousness in some way. It need have produced only a slight selective advantage to establish itself in a population, after which it would have been available to the first creatures to form thoughts. The addition of thoughts might originally have been useful to creatures insofar as they were non-consciously associated with past sensations as memories or as ways of anticipating experience so as to take action—exploit a feeding opportunity, escape a predator, etc.—without needing to wait for the object prompting the action to present itself in sensation (when it may already be too late). But non-conscious associations would be all but useless behaviorally if creatures were to extend belief affect to associated thoughts willy-nilly: such creatures would flee as desperately from the memory or passing mental image of a shark as they would from a shark present in sensation. Nonconscious association could be made behaviorally beneficial only if some but not all of a creature’s thoughts are believed. And the thoughts that would be most likely to enhance fitness by being believed are those most closely associated with sensations. Yet, non-conscious association is a crude instrument: no reality (belief) can be attached to relations in consciousness unless those relations ceased to be non-conscious; and to be oblivious of the relations things have to one another makes for very limited, very primitive conscious mentation, far below the level where one can speak of “conscious intelligence” or “intelligent consciousness” at all. Accordingly, belief affect can really come into its own as an enhancer of fitness only when systematically correlated to conscious association, so that belief is primarily bestowed only on such thoughts as are consciously associated with sensations in relations, and then only in proportion to the strength of the conscious associative tie. Since conscious associative relations are purely affective in character, this is just to say that the intensity of belief affect follows association affect in precise proportion to the intensity of the latter. Now, according to Hume, association affect is facility feeling. But what is belief affect? His preferred expression for the feeling that constitutes belief in the real existence of anything present in sensation, thought, or “reflexion”—Hume’s umbrella term for passions, emotions, desires, and volitions—was force and vivacity. By “force” he meant the “firm hold” on the mind that things regarded as real have and things regarded as imaginary lack, and he used “vivacity” to emphasize their being “live” to us in ways fantasy never is (at least if we are awake and compos mentis). It cannot be emphasized too strongly that Hume never intended these

100  Path to Kant: Psychologism and Conventionalism terms to suggest that belief is a violent feeling like rage or terror, or to imply that a powerful fictional narrative cannot excite intense emotions (e.g. weeping uncontrollably at the movies). A representation composed entirely of dull, unvarying contents can eclipse even the most brilliant, variegated representation in force and vivacity provided the first is ­sensation—e.g. an overcast silent night scene remote from city lights and sounds—and the second is imagination—e.g.  a fantasized chaotic midday battle scene. In addition, our attention (access consciousness) may be fully absorbed by the events of an idle daydream, while co-occurrent forceful, vivid representations of sensation, reflexion, and associationenlivened thought lie buried in the deepest recesses of dark consciousness. A believing consciousness is simply one imbued with the feeling that is alone capable of triggering—i.e. that has evolved to trigger (via its neural correlates)—the kinds of behavior that distinguish what is accepted as real from what is regarded as mere fantasy. Beyond that, it is just as futile to attempt to capture it in words as it is with any other feeling: Were we to attempt a definition of this sentiment, we should perhaps find it a very difficult, if not an impossible task, in the same manner as if we should endeavor to define the feeling of cold or passion of anger to a creature who never had any experience of these sentiments. Belief is the true and proper name of this feeling and no one is ever at a loss to know the meaning of that term because every man is every moment conscious of the sentiment. . . . I say then that belief is nothing but a more vivid, lively, forcible, firm, steady conception of an object than what the imagination is ever able to attain. This variety of terms, which may seem so unphilosophical, is intended only to express that act of the mind which renders realities, or what is taken for such, more present to us than fictions, causes them to weigh more in the thought, and gives them a superior influence on the passions and imagination.5 I can think of no better illustration of Hume’s conception of belief as a species of affect than the famous story of Orson Welles’s 1938 Halloween night radio broadcast of H. G. Welles’s War of the Worlds. The novel was updated and compressed to be presentable as a series of news bulletins. Although much exaggerated in the telling, it is said that some people tuning in took what they were hearing to be real and, terrified, took flight or prepared to fight, whereas other listeners accepted what they were hearing as a fictional script read by actors and so felt nothing remotely sufficient to trigger their fight-or-flight response. Since the very same contents were present in the minds of both groups and since everyone brought more or less the same associative background to their understanding of them, the question Hume would have us ask is this: what do all the members of the one group have in common that constitutes their

Hume’s Cement 101 belief in Martian invaders and disbelief in actors reciting lines in a studio that is reversed in all the members of the second group? It will not do to say, e.g., that some members of the skeptical group tuned in specifically to listen to Welles’s program and recognized the voices, for this merely tells us the causes (observation, testimony, etc.) and determining grounds (associative relations) of their belief and disbelief, but nothing about the difference believing itself makes in their consciousness. For Hume, the answer has to take its start from whatever it is that is common to the minds of all the gullibles, however different the causes and determining grounds of their belief in invading Martians and disbelief in actors acting, and whatever is common to the minds of the skeptics, no matter how different the causes and determining grounds of their belief in actors acting and disbelief in Martians invading. Before venturing to answer, Hume would further have us consider what the nature of belief needs to be if it is to possess the power to rouse in us a terror sufficient to trigger the fight-or-flight response. Could it do so if, instead of an affect, it were just another act of thought, not essentially different from the acts of representing this or that, comparing this with that, combining the thought of this with the thought of that, analyzing or inferring this from that, and such like? Just as these latter actions are, in and of themselves, powerless to raise intense emotion in us, the same would be true of belief if it too were a purely intellectual act. Such acts can directly connect up with other representations only through logical, grammatical, or other equally intellectual modes of combination. By contrast, emotions neither admit of such modes of combination nor have the least need of them in order to connect up with other representations distinct from them. Accordingly, just as hatred easily elicits anger and comeliness lust, nothing in the least intellectual needs to be posited to understand how belief, construed as force and vivacity affect, arouses emotions in us, even ones as strong as the terror felt by some of Welles’s listeners. Hume therefore saw no option but to reject accounts of belief as an act of thought and accept that it is an affect immanent to our consciousness of objects in sensation and thought insofar as the latter are closely associated with the former. We thus can be quite sure how he would have analyzed the effect of the War of the Worlds broadcast: believers in Martians invading felt the very same thing believers in actors acting felt, but because only the first conception is associated with imminent danger, only the addition of belief affect to that conception had the capacity to rouse emotions powerful enough to elicit the fight-or-flight response.

Cause and Effect Explicated as Customary Association Hume’s purely affective accounts of associative relation and belief in the reality of anything present to consciousness in sensation or thought

102  Path to Kant: Psychologism and Conventionalism have an obvious bonus: they make it easy to regard human cognitive psychology as continuous with non-human animal cognitive psychology. From the Humean standpoint, human minds are animal minds, the same in all essentials as the minds of dogs, pigs, chimps, dolphins, and other creatures: believers in sensational reality because of sensations’ maximal force and vivacity affect, and believers in a merely thought reality extending from sensational reality in the case of all and only those thoughts strongly associated with sensations via maximally intense facility feeling. Indeed, there is no preventing Hume’s model of human nature from extending far and wide through related animal classes and even phyla. But can purely affective animal association fully account for human intelligence? Even without mentioning artificial adjuncts (vernacular language, etc.), isn’t there more in human nature that is requisite for humanlevel intelligence than that? We shall see in Part II why Kant answered yes. But the reasons underlying Kant’s answer are so entirely original that Hume should not just be forgiven but praised for rejecting the affirmative answers available in his day and making so formidable a case in favor of the proposition that the natural principles underlying human intelligence are no different from those governing non-human animal varieties. We need not consider his critiques of the various affirmative options in detail here. Suffice it to say that their upshot is that there is nothing naturally in human understanding that could not be in any nonhuman animal understanding as well, so that this changes only when one turns to what is impossible apart from human artifice, above all bodies of linguistic convention like English as well as the uniquely human social practices and lifeways in which English and all other vernaculars are embedded. The gulf that emerges when human nature is complemented by artifice is most obvious in the case of science and mathematics. Since these cannot be mastered, practiced, or communicated other than in the media of speech and writing, they attest that human artifice can far outstrip anything human minds are capable of naturally, much less non-human minds. Nevertheless, when we are regarded in our state of nature, stripped of artificial props like language, Hume may well be right in holding that human intelligence is not all that different from animal intelligence, particularly in the case of our nearest pre-linguistic ancestors and relatives (Homo heidelbergensis? H. neanderthalensis?). So far as he was concerned, we have only to recognize both how powerful a representational instrument conscious association is and the purely affective nature of belief in order to appreciate that there is neither need nor grounds to suppose that human consciousness naturally includes representations “of so refined and spiritual a nature that they fall not under the conception of the [imagination] but must be comprehended by a pure and intellectual view of which the superior faculties of the soul are alone capable.”6

Hume’s Cement 103 But what about the most sophisticated notions often attributed to human nature rather than human artifact: cause and effect, identity at and over time, external objects, space, time, and self-consciousness? To make his associationist psychologism plausible as an account of human intelligence, Hume had to show either that we are mistaken to suppose we have such ideas naturally (especially without requiring language) or that there is in truth nothing more to these ideas than what can be explicated purely psychologistically in terms of various combinations of sensations, thoughts, transitions between them, and the affects of facility and vivacity constitutive of consciously accessible relation and believed reality. With Hume convinced that all of these ideas are, at their core, natural and shared across a wide swathe of the animal kingdom, the bulk of A Treatise of Human Nature is consequently devoted to showing that they can and should be explicated purely psychologistically, and so as beholden only to human-animal phylogeny, not convention, much less any “pure and intellectual view” of the kind postulated by platonists like Descartes or abstractionists like Locke. To this end, Hume resolved the full range diversity and complexity of human-animal understanding into expressions of three fundamental associative principles: association by resemblance, spatial and temporal contiguity, and cause and effect. The first two remain at the level of appearances: human-animal nature is such that the transition from one representation to another that appears similar to it or appears in spatial or temporal contiguity with it feels easier than transitions to non-­resembling, non-contiguous representations. Human-animal nature might have evolved differently so that just the opposite were true or something else entirely, in which case resemblance and/or contiguity of appearances would be naturally regarded as noise rather than relational information, and something that for us now is noise might have served in their stead. Yet, even in creatures constituted like us the influence of resemblance and contiguity when acting on their own “is very feeble and uncertain.”7 Accordingly, Hume focused almost exclusively on the third fundamental associative principle: cause and effect. The most salient thing that distinguishes causal relations from the other associative principles is their absence from the level of appearances. Let’s take Hume’s familiar example of colliding billiard balls. Anyone previously unfamiliar with billiards but well practiced in the ways of nature would need only look at the implements of the game to know precisely what will happen as soon as someone propels the cue stick into the cue ball. This makes it all too easy to suppose that the sensible appearances of the stick, balls, the table and its sides, and, of course, the player wielding the stick are all one needs to associate these objects in their proper relations as causes and effects. But what if someone unfamiliar with billiards who had spent his entire life in zero gravity and had no notion of how things behave at the surface of massive bodies like the earth were to

104  Path to Kant: Psychologism and Conventionalism witness the same billiard hall scene—remotely, say in the form of a futuristic true-to-life moving hologram, projected in his familiar zero-gravity environment? His experience of objects that resemble the implements of billiards would lead him to expect that the cue ball would behave like the weightless bodies familiar to him, and so be vastly more likely to fly off the table and bounce off the walls and ceiling when struck by a stick than behave so farcically improbably as our experience as earthbound observers leads us to expect. But how could this be so if the causal relations between objects had anything to do with their appearances? If they did, then the zero-gravity man would be able to use his senses simply to read off their causal relations in precisely the same manner he uses them to read off resemblances like the balls’ spherical shape, smooth surface, glossy finish, etc., and contiguity relations such as which balls abut which others, which do not abut one another but jointly abut some third ball, and which abut neither one another nor any intervening ball. Since scenarios are easy to imagine where any of us could be this person, there seems to be no option but to conclude that causal relations are in fact insensible and that it is only life-long acquaintance with how objects behave in the environment(s) familiar to us that misleads us into supposing otherwise. Hume also invites us to consider how the world would appear to us if (the neural correlates of) all the causal relations we have become conscious of over the course of our lifetimes were suddenly erased from memory. We would no longer be conscious of any associations between the movements of our eyeballs, head, torso, or legs and visual changes, or of any of the relations between either of these and other sensory consciousness (auditory changes, mien, language, etc.). Consequently, we would lose all consciousness of sense-divide transcending objects such as billiard balls, and so, a fortiori, all consciousness of their collisions and other causal relations. Even our consciousness of our own minds in self-consciousness would be lost: we would be oblivious to what makes thoughts memories (viz. consciousness of their causal relation to contents of past consciousness); lose all consciousness of the existence of thoughts as the effect of the presence in consciousness of an associated sensation, memory, or thought; have no consciousness of the causal dependence of emotions on any of the other contents present in consciousness (e.g. the terror upon seeing a great white shark swimming directly at one); and in general lack even the darkest awareness of causal relations of any kind between any of the contents present concurrently or successively in consciousness, be they sensations, thoughts, or reflexions. The point of this Humean thought experiment is that while all appearances would remain exactly the same, unchanged from what they were before the neural correlates of causal consciousness were deactivated, all coherence and connection, even those linking present consciousness to the consciousnesses immediately preceding and succeeding would

Hume’s Cement 105 be wholly effaced and supplanted by complete causal obliviousness. By accepting the point, we grant that causal connections are never directly observable in sensible appearances and so must always be brought to consciousness by other means. Obviously, the accumulation of observations over the course of experience plays a role. But how could it play this role if past experience is supposed to consist entirely of past appearances? Since nothing new appears on the hundredth occasion that was not already fully evident on the first, Hume concluded that experience makes us aware of causal relations between contents appearing in consciousness not by adding something to their appearance but rather by changing how we are conscious of them. Repeated observation does this by engendering habits of mind, or, as Hume more usually termed them, customs. Repetition can be done artificially, by training or educating a teachable animal or person. But it can also occur naturally, as when resembling appearances recur in experience with sufficient frequency and constancy for a custom of association to become instilled. If the appearance is a sensational quality like yellow, then experience of enough appearances with that qualitative resemblance will engender a customary resemblance association of the kind Hume, following Berkeley, used to explain how particular thoughts become general (“[i]f ideas be particular in their nature and at the same time finite in their number, ’tis only by custom they can become general in their representation and contain an infinite number of other ideas under them”).8 But should the appearance be a contiguity relation (e.g. one billiard ball colliding with another), then if enough similarly appearing, similarly related things are encountered in experience, a custom will be ingrained that not only yields a general idea of the relation but also makes us conscious of it as a relation of cause and effect. How does customary association do this? “Custom has two original effects upon the mind, in bestowing a facility in the performance of any action or the conception of any object, and afterwards a tendency or inclination towards it, and from these we may account for all its other effects, however extraordinary.”9 If the custom is associative already, it will strengthen the “feeble and uncertain” facilitation of resemblance and contiguity relations to the point where the transition becomes “fixed and unalterable . . . solid and real, certain and invariable.”10 In other words, with the addition of custom to association, facility will be felt in the highest degree, yielding the strongest possible non-logical conscious connection representations can have, while at the same time raising force and vivacity affect to maximal intensity, so that both the connection and the objects related by it will be regarded as fully real even when not all of them are present in sensation (e.g. the cannons I think of but neither see nor hear when I espy a distant fireworks display). The psychological preeminence Hume attributed to customary association in both human and animal nature derives precisely from its ability

106  Path to Kant: Psychologism and Conventionalism to widen the scope of one’s consciousness of reality beyond sensations— human-animal nature’s default reality—to include things present only in thought. Generalizations take on the worth of really existent natural kinds (species, types), and contiguity relations become real connections capable not only of joining really existent particulars (singulars, individuals) presently appearing in sensation but also of connecting these to particulars in spaces and times present only in thought that, owing to their customary associations with sensations, are nevertheless regarded as equally real: hidden interiors, obverse sides, concealed parts, details obscured by distance, etc. Thus, the enhancement of facility and vivacity affect that results when customary association is at its strongest extends the boundaries of a creature’s world—the sum total of what it presently regards as real—beyond the narrow compass of sensation to include natural kinds in addition to particulars, particulars occupying spaces and times absent from sensation, and connections linking all particulars together in a single natural order. How does customary association give creatures consciousness of causal relations more particularly? Before we can be conscious of the connection between the existence of things, actions, or motions as causal, one further ingredient is essential: immediate temporal contiguity. If objects (things, actions, motions) come into existence simultaneously, then, even if they are spatially contiguous, one cannot consciously connect the existence of the one to that of the other as cause or effect no matter how strong the custom associating them. They may still be represented as conjoint effects or causes with respect to something else, but the existence of the one cannot be conceived to be connected to the existence of the other in any way. But we equally cannot become conscious of objects as cause and effect if any interval of time elapses between the (sudden or gradual) coming into existence of the one and the (sudden or gradual) coming into existence of the other. On the contrary, if the one exists for any time without anything further happening, then the other’s subsequent coming into existence cannot be its doing, but must instead be due, at least partly, to another cause. So, in order for anything to be even so much as a candidate for being connected with something else in a customary causal association, they must be immediately temporally contiguous. When immediate temporal contiguity is part of the consciousness created by a maximally idea-enlivening customary association, everything is in place for a creature to become aware of the association as one between a cause and an effect. “[C]ustom operates before we have time for reflection” to make the “objects seem so inseparable that we interpose not a moment’s delay in passing from the one to the other.”11 At its transitionfacilitating maximum, it “conveys the thought to its usual attendant” with such force that “it will scarce be possible for the mind, by its utmost efforts, to prevent that transition,” thus rendering it psychologically (not logically!) quite literally “impossible for it not to form an idea of the

Hume’s Cement 107 other.”12 In this way, customs intensify facility affect so greatly that the relation between the objects involved in the transition feels necessitated, becoming effectively the affective equivalent of logical necessitation. Moreover, insofar as vivacity follows facility in a degree proportionate to the intensity of the latter, it becomes equally impossible not to believe that the customarily associated object to which the transition is made in thought is fully as real as the object present in sensation that triggered the transition. This means that it is not just the objects concerned in the transition but, quite specifically, their existence—vivacity affect-constituted reality (actuality, facticity)—that is associated by it. More particularly, insofar as custom determines one to be always the immediate predecessor of the other, the earlier will be apprehended as necessitating the existence of the later, and this will remain true even if, on occasion, it is the later that is present in sensation and the transition is made to the earlier in thought—i.e. the experience-bred temporal order ingrained in custom will be regarded as real even if the order in which they are sensed is on occasion reversed. Insofar as “cause and effect” relation is supposed to signify that one thing has been necessitated to exist by the (coming into) existence of something else, customary association therefore furnishes all the necessary ingredients. The necessary-feeling transition and irresistible belief affect characteristic of the strongest customary associations between immediately temporally contiguous objects, actions, or motions are thus both necessary and sufficient for consciousness of a causal connection between them: [A]fter a repetition of similar instances, the mind is carried by habit, upon the appearance of one event, to expect its usual attendant and to believe that it will exist. This connection, therefore, which we feel in the mind, this customary transition of the imagination from one object to its usual attendant, is the sentiment . . . from which we form the idea of [causal] power or necessary connection. Nothing farther is in the case. Contemplate the subject on all sides, you will never find any other origin of that idea.13 The natural objection to any such psychologistic explication of cause and effect is that maximal facility and vivacity affect imbuing my consciousness of colliding billiard balls are not the same as a consciousness of those billiard balls themselves being related to one another as cause and effect. Yet, this objection can carry psychological force only if facility and vivacity affect draw sufficient attention to themselves to become manifest at the level of access consciousness. Far from being attention grabbing, however, they naturally recede into the background of dark consciousness and, even at maximum intensity, never emerge. This is obvious in the case of vivacity affect, the very nature of which is projective, so that we regard not the consciousness it imbues but whatever

108  Path to Kant: Psychologism and Conventionalism content is present to consciousness in sensation reflexion or thought as really existent. As for facility, anyone who has ever performed a well-practiced routine knows that as soon as the actions of the body or thought become second nature, no notice is taken of the transitions from one to the next—unless one happens to go wrong, in which case facility is interrupted by a “sensible violence” and one may have to refocus attention on one’s activity until facility is restored and attention can again be directed elsewhere. So, if we grant Hume’s theses that any consciousness actuated by association essentially involves facility affect and that association buttressed by custom boosts facility to maximal intensity, then there should also be no difficulty accepting the corollary that “[s]uch is the influence of custom that, where it is strongest, it . . . conceals itself and seems not to take place merely because it is found in the highest degree.”14 Facility feeling draws the least attention to itself precisely where its affect is most intense. Yet, if both the facility and vivacity affects constitutive of causal association are entirely absent from access consciousness, this leaves only the contiguous objects concerned in customary transitions on which to focus attention. Consequently, it is no surprise that “we suppose [casual] necessity and power to lie in the objects we consider, not in our mind that considers them, notwithstanding it is not possible for us to form the most distant idea of that quality when it is not taken for the determination of the mind to pass from the idea of an object to that of its usual attendant.”15 Since Hume deemed it a principal argument in favor of his account of human causal consciousness that it extends naturally to animals, let us briefly consider what that means. Imagine a mature firehouse Dalmatian that has had frequent and constant experience of fire preceding smoke in a wide swathe of relevantly resembling cases—fires of various characters and sizes in various places consuming various materials succeeded by smoke of various tints, odors, and concentrations. Hume saw no reason to doubt that the effect of repeatedly experiencing these resembling contiguity associations would be the same on its mind as it is on ours: the inculcation of a strong customary association between fire and smoke with the first represented as always preceding the second. Accordingly, when the fire truck pulls up 50 meters from an abandoned multistory factory with smoke pouring through its windows, it will be just as true of the Dalmatian as the firefighters and other human witnesses that they will find it not only impossible not to think of a fire in the building that preceded the smoke in the order of time but likewise impossible not to believe that fire, albeit unsensed, fully as real as the smoke they see and smell. On Hume’s analysis of customary association, this is just to say that the transition from the sensations of the one to the thought of the other will be characterized by a maximal degree of facility, with the result that the thought to which the transition is made will be enlivened to the point where its vivacity matches that with which the sensations

Hume’s Cement 109 are regarded. More particularly, the intensity of the facility affect makes the transition feel necessary to the dog’s imagination in the same way it does to ours—inevitable, irresistible, unpreventable—and the intensity of the vivacity determines that, in its access consciousness, the connection will not be just between fire and smoke but, quite specifically, between the existence of the one and the existence of the other. In other words, the dog will have precisely the same concept of causal necessitation that humans have—at least humans in a pre-linguistic state of nature. Finally, given that experience has taught the Dalmatian just as it has us that the existence of fire always precedes that of smoke, it will regard the existence of the fire as having necessitated the existence of the smoke even though it saw and smelled the smoke before it observed the fire—i.e. customary association will determine it to reverse in thought (associative imagination) the order in which they were sensed. Now, if all this is granted, is there anything to prevent us from ascribing to the Dalmatian the very same consciousness of the fire causing the smoke that we attribute to the humans? If we agree with Hume in answering no, it puts paid to the notion common among psychologists that animals have no ability to represent anything absent from their senses. For any creature equipped with powers to form mental images, associate them, and strengthen its associations by custom ipso facto has the ability to infer unsensed causes from sensed effects and unsensed effects from sensed causes, as well as distinguish their objective order in time from the order in which they are sensed whenever they disagree. Nor is their consciousness of unsensed reality limited to causes and effects. Strong customary associations between coexistent objects, qualities, and actions would enable creatures to accord the same real existence to the unsensed interiors of familiar objects they accord to their sensed exteriors, to the insensible parts of wholes sensed at a distance, to features inaccessible to some sense (because of intervening objects, being down wind, poor lighting, too much ambient noise, etc.) when associated coexistent features are accessible to another sense, and so on. There is therefore nothing to prevent creatures from being as fully cognizant as humans of realities they are not currently sensing, both past and present, as they are of the realities they sense.

Causal Association Put to Cognitive Use Once a creature has causal consciousness, it finds causes wherever there are recurring temporal sequences, i.e. everywhere and always. Resembling colors, feels, sounds, etc. are encountered repeatedly. Some of these are found to be always concurrent (sensation “bundles”), while others are found to be always temporally contiguous, and it is the latter that are invariably associated in natural human-animal consciousness as cause (for the accustomed predecessor) and effect (for the accustomed

110  Path to Kant: Psychologism and Conventionalism successor). Nor is it difficult to see how causal consciousness might boost a creature’s evolutionary fitness. For while virtually all animal species put non-conscious causal information to behavioral use, those naturally equipped with causal consciousness of a subset of that information will be able to develop new and better uses for it that would otherwise be unattainable, and so, to that extent, be likelier to thrive in relation to less psychologically well-endowed competitors. Before we consider the mental basis for such new and better uses, we first need to note how the Humean psychological mechanism of causal association can evolutionarily adapt to optimize the use of causal information. The need to do so is not difficult to discern. If, on the one hand, causal associations are not prevented from being formed too quickly, major, easily avoidable sampling errors are apt to be committed, possibly resulting in a creature’s premature demise. But if, on the other hand, they are not prevented from being formed too slowly, well-founded causal associations that might have guided a creature’s behavior sooner will be absent, perhaps precisely when needed most, again perhaps with fatal consequences. How then can creatures calibrate their causal consciousness to find the happy medium that maximizes fitness? Obviously not by discovering and applying the principles of statistics, Bayesian analysis, Judea Pearl’s causal inference engine, or anything else that not only requires sophisticated reflection but is also impossible to represent, so far as we know, without the aid of linguistic conventions (words, sentences, combinations of sentences into narratives, sometimes supplemented by specially tailored symbolic notations). Fortunately, the purely affect-driven mechanism Hume proposed in relation to creatures in a state of nature, prior to and independently of artificial enhancements like language, requires nothing of the kind. Feelings, by their very nature, wax and wane, and facility and vivacity are no exception. More particularly, here, as everywhere else association operates, vivacity follows facility: as soon as facility reaches sufficient intensity for a causal relation to be conceived, vivacity will have waxed to the point where the conceived relation will ipso facto be believed to be real. This, however, still leaves the rate at which they intensify to human-animal nature to determine. Creatures that intensified their facility and vivacity into causal associations after too few encounters of resembling sequences of objects to avoid major sampling errors would be more likely to be culled than ones that ripened these feelings more gradually, while creatures that intensified them too gradually to put experientially well-founded causal associations to use in the most timely fashion would also be more likely to be eliminated than those that did so earlier. The sweet spot in the middle is thus where facility reaches the ripeness of intensity sufficient to trigger vivacity at the rate best suited for (the neural correlates of) causal consciousness to guide a creature’s behavior in the manner most likely to enhance fitness.

Hume’s Cement 111 Since creatures with psyches optimally attuned for causal association would presumably benefit by increasing the amount of useful causal information at their conscious disposal, inquisitiveness of the relevant kinds might yield a further boost to fitness. For example, an urge to array causal associations in series could prove beneficial if the thought of the immediate associate of something presently in sensation were of no use in present circumstances, but the thought of something associated at one or more removes were. In that case, a creature that naturally arrayed its causal associations in series and then integrated these into larger causal networks could greatly expand its intelligence (pun intended) regarding the world. And the more integrated and extensive networks the better, as the human case bears out: [the system of realities causally connected via custom] peoples the world and acquaints us with such existences as by their removal in time and place lie beyond the reach of the senses and memory. By means of it I paint the universe in my imagination and fix my attention on any part of it I please. I form an idea of Rome, which I neither see nor remember, but which is connected with such impressions as I  remember to have received from the conversation and books of travelers and historians. This idea of Rome I  place in a certain situation on the idea of an object which I call the globe. I join to it the conception of a particular government, and religion, and manners. I  look backward and consider its first foundation, its several revolutions, successes and misfortunes. All this and everything else which I believe are nothing but ideas, though by their force and settled order arising from custom and the relation of cause and effect, they distinguish themselves from the other ideas which are merely the offspring of the imagination.16 Take away the veneer of human artifice represented by spoken and written language, culture, and social organization, and the universe that creatures of our species would paint in their imaginations by means of integrated series of causal association presumably would probably differ little if at all from the universes of other hominoids, cetaceans, birds, and much else besides. “The globe” for a given animal species might not extend to the next valley or even the next copse, but it might also be genuinely planet spanning (e.g. for pole-to-pole migratory cetaceans and birds). Causal association here would be between the place and time of present sensation, the thought of a certain place remote from it, and the association-based causal series formed from tactual sensations of bodily movements, i.e. how much crawling, walking, burrowing, flying, swimming, etc. gets the creature from here to there via a given route. This in turn would be naturally integrated with the various obstacles, dangers, food sources, etc. its experience teaches it that each route is likely to offer.

112  Path to Kant: Psychologism and Conventionalism If only because their description would require many words, its considerations seem complex. Yet, the affective nature of association, in which the strongest associations are easiest and most irresistible, would obviate the need for effortful ratiocination of the kind we think of in connection with working out solutions to complex, multi-faceted problems. On the contrary, once a creature has a mental map of the world correlated to its mode of locomotion, causal association would do the rest: each route from the place of present sensation to the destination in thought will present itself, excite whatever emotions and desires thoughts of the promises and perils of each naturally do, and thereby determine volition to a decision, all within a veritable blink of an eye, quite possibly without ever breaching the surface of dark consciousness to draw the least attention to itself. Nor does the utility of customary causal association end with forming and navigating by mental maps. Encountering new, unfamiliar things and situations can put creatures in extreme danger. Because they are new, they have no relation to the creature’s existing network of causal associations. Some reactions can be hardwired (fight or flight, etc.), but new, unfamiliar things and situations can also present opportunities for a creature. The happy medium is to access all and only those opportunities that can be exploited within a range of acceptable risk. To assess the risk, causal association not only can be used but is often indispensable. The key to applying it is the kind of natural inquisitiveness that seeks out analogies between the new and the familiar. Insofar as everything resembles everything else in some respect or other, the search for analogies starts from whatever sensible resemblances (i.e. resemblances in appearance) the new and unfamiliar has to the customary and familiar—color, smell, feel, mode of locomotion, etc. If there is some insensible causal property associated with all familiar things that share these sensible resemblances (e.g. the ability to escape up trees), then associating it with a new object or situation that shares most or all of these resemblances no matter how unfamiliar it is otherwise not only will serve to dissipate a creature’s confusion and anxiety but may also suggest a course of action that could prove beneficial or even essential to its survival. On Hume’s analysis, analogical association starts from an experiencebred custom strong enough to facilitate the transition to the point where a connection between the sensible resemblances and the insensible causal property is both conceived and believed to be real. Equipped with such customary associations, a creature (1) will find it easy, even irresistible, to transition in thought to whatever causal properties are suggested by the sensible features observed in new, unfamiliar things or situations, (2) will consequently not only think these connections in the new instance but enliven their conception to the point of belief in their reality, and (3) will thereupon think and act on the assumption of the reality of the causal properties. Of course, if the unfamiliar object or situation exhibits too

Hume’s Cement 113 few of the resemblances associated with a causal power, the analogy may be too weak to rouse sufficient facility and vivacity feelings to determine a creature’s thoughts and actions. But even if insufficient to determine them, a weak analogy can still enhance these feelings enough to influence a creature’s thought and action, quite possibly to its benefit—and even slight enhancements to fitness can transform a species in the fullness of evolutionary time. Here, it may be queried whether Hume’s assimilation of human causal representation to animal causal association goes as far as such general abstract causal notions as laws of nature and universal forces. Motive forces, for example, engulf all creatures from the first moment of their conscious lives. But are any non-human animals even darkly cognizant of their commonality in such diverse phenomena as leaves falling in the wind, rain pelting, torrents cascading, coursing, and eddying, wings flapping against the air, feet pushing against sand, hard clay, or pavement, stones hurled, and, in general, wherever there is movement, resistance to movement, and effort applied to overcome it? It may seem that representations as general and abstract as motive force are so far beyond the psychological capacity of non-human creatures that they can have no consciousness of them. Cognizance of them may even be thought to require language, an artificial confection no different from other conventional practices, and so no less uniquely human. But if instead human nature already of itself equips us to conceive and believe causal notions of higher generality, and human nature is accepted as simply one among many variants of animal nature, what could prevent non-human animals from conceiving and believing them too? Hume saw no difficulty in explicating highly generalized causal notions, including nature and natural law, purely psychologistically, and so in a fashion extendable to animals. After all, there are two representations that, as essential ingredients of all consciousness of causal connections, are common to all representations of causation: immediate temporal contiguity and the real, though purely affective, necessitation produced whenever customary associations of temporal contiguity are at their strongest. These are universal to consciousness of causal relations however diverse their particular relata—raindrops, snapping jaws, waterfalls, footfalls, wings flapping, etc. Since these resemblances are encountered with more than sufficient frequency and constancy to form a customary resemblance association, any creature capable of customary association not only can form a fully generalized idea of causal relations but cannot help doing so. And since virtually all causal relations in nature that present themselves to access consciousness involve movement, or at least resistance to it, animals (including us) obtain the concept of motive force as a free bonus, as it were, with their general concept of causation. Of course, universal causal concepts are no good to a creature unless they are put to use in the universal representation of nature and natural

114  Path to Kant: Psychologism and Conventionalism laws. Hume focused on the two applications of the generalized concept he considered most important to human and animal fitness: the principle of the uniformity of nature and the general causal maxim. Any creature with a fixed, unshakeable belief in the uniformity of nature will approach new experiences in the certainty that the causal necessities and impossibilities familiar to it from the past will hold of the present situation and any it encounters subsequently. This conviction means that instead of being paralyzed in thought and action by uncertainty when confronted by things and situations so unfamiliar as to have only weak analogies with familiar ones, creatures can still approach them in the assurance that at least the general causal laws familiar to them will continue to hold, and so think and act accordingly. Similarly, any creature with a fixed, unshakeable belief that everything that begins to exist must have a cause of its existence will believe that there are insensible causes hidden behind every sensible beginning of existence. This can improve fitness if the creature is impelled by this belief, in any matter of sufficient concern to it, to devote the time, energy, and potential risk required to ferret out elusive causes and/or effects. If such searches were to pay off often enough to be a net benefit, then better searchers in a population will be advantaged relatively to poorer ones and non-searchers. Certainly, there can be no doubt that hypothetical creatures that naturally believed the contrary of the general causal maxim—that no causes exist that past experience has not made familiar to them—are unlikely to prevail over those that naturally believe that all beginnings have causes, familiar or not. Given the utility of both the uniformity principle and the general causal maxim, the question therefore arises whether creatures, ourselves included, are naturally equipped to conceive and believe them. Hume answered yes in both cases. The uniformity of nature boils down to the uniformity of causes: causal association makes its possessor conscious of the existence of one thing—object, action, motion—as necessitating the existence of another and, equivalently, the impossibility that the latter exist unless immediately preceded by the existence of the other. This combination of necessity and impossibility also has a temporal dimension: if the existence of X is necessary to the existence of Y and that of Y impossible without that of X, then whenever X exists Y’s existence must always immediately succeed it such that Y never exists unless immediately preceded by X’s existence. A causal necessity is consequently a causal always and a causal impossibility a causal never. Thus, “even brute beasts” that know causal connections from customary association ipso facto recognize the uniformity of causes through time, and so, like humans “long before [they are] out of [the] cradle,”17 are naturally equipped to conceive and believe the uniformity of nature principle in its full generality. The general causal maxim that everything that has a beginning of existence must have a cause can be conceived and believed by simple

Hume’s Cement 115 analogical association. A creature, finding that all causal relations it has experienced are both contiguous and connected by affective necessity, analogically universalizes the correlation to all contiguity and thereby is induced to believe there to be a contiguous cause of every beginning of existence, whether the contiguous objects are sensed or not and whether they are present, past, or future. For although there is nothing intrinsic to the idea of successive existence that requires it to be conceived and believed to involve causal connection, the force of customary association is more than strong enough to make up the logical shortfall psychologically. Of course, if a creature started encountering successions of existence it knew to be causeless, the relation conceived in the general causal maxim would fade, and belief with it. But since no creature in a state of nature can ever know anything not to have a cause, the analogy is naturally impervious to refutation.18 Thus, all creatures capable of customary causal association are in a position to benefit from belief in the general causal maxim by undertaking searches for unknown causes or effects whenever motivation to do so exists. Here, however, the specter of what Hume termed “contrariety” rears its head. Yes, A causes B, but sometimes A is succeeded by C, D, or E rather than B, and sometimes B is preceded by F, G, or H rather than A. Yet even one such exception is enough to undercut both the uniformity of nature principle and the general causal maxim—or rather prevent them from being conceived and believed in the first place. So how did Hume factor in contrariety, much less do so in such a way as to render these principles extendable to non-human animals? Hume saw nothing to prevent creatures from using their imaginations to parcel out the components of sensation in thought, both to combine them in new ways (fantasy) and to separate one or more components from the others for purposes of comparison. Comparison enables them to single out what remains constant through time from what does not and so to sort things into kinds by means of constant resemblances between subsets of their sensible features. Since sorting in this manner is sensitive to how far into the details of a things’ sensible appearance access consciousness (= discernment + attention) can penetrate, it also enables creatures to sort components into kinds by their sub-components to one or more iterations. So, associative generalizing enables creatures to relate things at multiple levels of generality (species, genus, etc.), resulting in hierarchical classification. Hume’s dice example illustrates this perfectly: we know, with absolute certainty, that throwing the dice will cause them to fall to the table, roll, and stop on one of their faces, but have no such certainty about which face they will fall on. Here, then, are two causal relations at different levels of generality: the causal necessity that thrown dice behave as thrown dice always do and the probabilistic gamble that you’ll roll a seven or eleven. So, too, when stalking lions are ready to emerge from cover and give chase to a herd, they are conscious, albeit

116  Path to Kant: Psychologism and Conventionalism purely affectively, that, once they do emerge, as a matter of causal necessity, their prey will flee; by contrast, they are conscious of no necessity of how the pursuit will play out, much less which member of the herd they will end up targeting, chasing down, and devouring (if any). And so too generally: thanks to multiple levels of classification, affective causal necessity is never absent from a creature’s experience at higher levels, no matter how permeated with randomness “at ground level” it may seem, i.e. at the lowest, least abstract level or levels of classification. With contrariety understood as everywhere embedded in necessary, universal causal laws of nature, it consequently poses no obstacle to creatures conceiving and believing in the maximally general, highest level causal principles of the uniformity of nature and a cause of every beginning of existence. Thus, Hume could legitimately claim to be the first psychologist to completely assimilate human with animal understanding in matters so abstract and universal. Accepting Hume’s use of a hierarchy of sortals grounded in customary resemblance association to explain how contrariety does not prevent creatures from establishing and maintaining unshakeable belief in the two fundamental principles of causal experience requires only one thing: that we first accept, if only provisionally, that causal concepts and inferences are purely affective in character in humans and non-human animals alike. If one does this, and further agrees with Hume that many animal species give every indication of being guided by belief in both the uniformity of nature and the general causal maxim, then there seems no alternative but to conclude that they have at least dark consciousness of these principles. For if, as Hume’s psychologistic explication of cause and effect mandates, facility and vivacity feeling are among their essential constituents, consciousness-independent representation of these principles, as of causation generally, is no more possible than it is for creatures to suffer pain even if no sensation of pain is present to their consciousness. Thus, insofar as their scope seems to extend beyond that of pain to consciousness-independent contexts, it is not human-animal psychology, conscious or not, but socially grounded logico-grammatical conventions governing word use that make it possible.

From Associative Causal Relations to Associative Identity Relations: Selves and Bodies Hume regarded human infants and other newborn creatures as being in the same situation as persons like those posited earlier, in whom the neural correlates of all the causal relations they have become conscious of over the course of a lifetime were suddenly eliminated. The brain in which they have yet to be activated and the brain in which they have all been deactivated would be equally unconscious of the causal relations between present thoughts and past sensations that make the former

Hume’s Cement 117 memories rather than mere mental imagery. They would also be oblivious to the causal dependence of visual sensations and their changes on tactual bodily spatiality (felt eyeball positions/movements, felt head positions/movements, etc.), and so conscious of neither visual spatiality nor sense-divide transcending objects. And they would be unconscious of causal relations between objects generally, from the most particular to the most universal, both externally (e.g. billiard balls) and internally (e.g.  will), with the result that their perceptions—sensations, thoughts, and reflexions—would all seem entirely “loosened,”19 causally connected neither to one another nor to anything else. If (the neural correlates of) all consciousness of causal relations is no different from (the neural correlates of) consciousness informed by customary contiguity association, then all infant creatures need for causal consciousness is to form customary contiguity associations. Since resemblance and contiguity are the only associative relations that can be formed on first acquaintance, without requiring the frequent, constant experience required to instill customs, these relations are the raw material from which customs generally, and customary causal associations more particularly, must take their start in the infant mind. What then is the first frequently encountered, invariable resembling temporal contiguity that appears to the not-yet-causal consciousness of infant creatures? It is an association that makes its first appearance from the second moment of a creature’s conscious life, where each sensation is immediately succeeded by a closely resembling thought that is thereafter added to its stock of mental imagery. With further experience, the creature also discovers that the order is never reversed: the thought always succeeds the sensation it resembles, never vice versa. Since this resemblance among directly observable temporal contiguity relations has perfect constancy, zero contrariety, and therefore gives rise to the strongest possible custom, it thus provides the infant mind with its very first consciousness of causal necessitation.20 Long before we are out of our cradles, and before non-human animals reach a comparable point in their lives, additional customary causal associations will have formed: consciousness of the past in memory, consciousness of visual spatiality and sense-divide transcending external objects, as well as internal consciousness of ideas giving rise to passions and emotions, these to desires or aversions, and those in turn to volitional action or inaction, with all leaving behind resembling imagery to add to one’s stock of thoughts. Presumably analogical association will be in place as well, along with generalizations sufficient to reckon with contrariety. Thus, Hume did not hesitate to attribute to infant minds (dark) consciousness of both causal uniformity (the uniformity of nature principle) and the causal necessitation of every beginning of existence (the general causal maxim). But what about consciousness of external objects existing in the external world, especially when absent from the senses? Here it is crucial to

118  Path to Kant: Psychologism and Conventionalism recognize that there is more to “external” than spatiality. Two or more things exist externally to one another if they exist independently of one another—e.g. one need not cease to exist if the other does. If one of the things is a perceiving consciousness, then to be conscious of the other as external, it must be conscious of it as existing independently of itself. Hume took this literally, as independence of self-consciousness, with the implication that consciousness of external reality, and so too of the external world in general, presupposes consciousness of self. It is not difficult to see why. Before something can be represented as external to something else, each must first be represented as external to me. Can ‘me’ here be understood as my body? Hume thought not. The body is not literally contained in consciousness, but rather in space. It is merely represented in consciousness, by proxy as it were, through tactual sensations together with all non-tactual sensations correlated to them (insofar as one sees one’s limbs or flippers, hears one’s larynx or syrinx, smells one’s bodily odors, etc.). Since these are all fleeting conscious representations that would themselves be represented as internal and dependent if there were a representation of the self in consciousness, it follows that “ascribing a real and corporeal existence to these [sense] impressions, or to their objects, is an act of the mind as difficult to explain as that which we examine at present.”21 So, before the body can serve as a criterion for distinguishing externality from internality (outside/inside my body), something else must serve as a criterion for representing the externality of the body itself in the first place (outside/inside me). In the absence of consciousness of self, what remains to fill that role? There are only the particular sensations, emotions, thoughts, and other representations that a creature would be conscious of in itself if it were self-conscious. These, however, are one and all ephemeral, continually displacing others in sensation or thought and being displaced in turn. To employ any of them as the internal/dependent existent relative to which anything else counts as external/independent would mean that before and after its existence nothing could be represented as external/ independent, while everything in between would count as external/independent whether it was genuinely so or was so merely for want of true self-consciousness. For example, in the absence of self-consciousness, my migraine would count as external to the objects I see or touch because it is capable of continuing to exist even if I close my eyes and cease touching whatever I just was. Similarly, my desire to drink the glass of water I am holding would count as external to those same objects if I continued having it after shutting my eyes and putting down the glass. In order to be conscious of anything present to me in sensation as external/independent, I need to be able to represent it as external/independent in relation to pains, desires, emotions, and everything else I normally would regard as internal and dependent; yet, to so regard the latter, I need to be able to represent them as existing internal to and dependent on myself—i.e.

Hume’s Cement 119 I need self-consciousness. For the same reason, in order to be conscious of anything in sensation as external/independent before and after as well as during my sensing of it I need to be conscious of a self that endures while sensations and other representations come and go to provide the kind of criterion of internality/dependence adequate to represent true externality/independence. Self-consciousness, as Hume understood it, is a consciousness of a self present from the first consciousness of one’s conscious life to the last. Its neural basis is whatever had to come “on line” in my fetal brain that was identical with my coming into existence as a consciousness, which then had to stay active for me to continue existing, and that had to be destroyed for my last consciousness to be terminal, i.e. the end of me. What consciousness of this enduring identical self does that neural basis yield? Hume’s answer starts from the recognition that new sensations are continually displacing present ones, and, in the process, laying down new memories and imagery for thought. In thought, new associations are continually made that, if they become customary and involve succession, yield consciousness of new causal relations of varying degrees of generality. On top of this, maturation, memory loss, radical changes in a creature’s situation (e.g. being confined in prison or a zoo), and other lifealtering circumstances affect a creature’s psyche no less than its physique, so that “the same person may vary his character and dispositions as well as his [sensations and thoughts] without losing his identity.”22 The question therefore becomes the following: whether, notwithstanding all this flux, there is “any single power of the soul which remains unalterably the same, perhaps for one moment.”23 Hume generally used the terms ‘self,’ ‘personal identity,’ ‘mind,’ and ‘soul’ interchangeably, so I shall too. Given the continuous change in the contents of consciousness from moment to moment throughout our lives, he thought no idea of an identically enduring self possible prior to and independently of associative imagination. But unlike his predecessors and virtually all of his successors, Hume regarded associative imagination as sufficient for consciousness of self, and therefore did not hesitate to extend such consciousness from humans to animals, including, no doubt with a smile, prideful peacocks.24 Why them? Self is the object of pride, just as it is of shame and a number of other “passions of the soul.” This is one reason why many would regard the attribution of pride to peacocks as outright anthropomorphism. But what if all that is required for selfconsciousness is customary association? In that case, the idea of self, just like that of causal necessity, would be nothing more than an amalgam of relation-constituting facility and belief-in-reality-constituting vivacity affects, and so may be presumed to be well within the capacity of nonhuman animals. If, in addition, Hume was correct that an idea of self is essential to being able to consciously distinguish external/independent from internal/dependent reality, then we not only can but should

120  Path to Kant: Psychologism and Conventionalism attribute it to any and all creatures that show evidence of being aware of the objects they sense as external and independent of themselves. Indeed, in that case, a refusal to grant that peacocks in full display are most likely feeling the same thing about the same object that humans feel about themselves when they too are in full display might better be characterized as a case of what Frans de Waal calls “anthropo-denial” than prudent wariness of anthropomorphism. What idea of self capable of being extended to avians does customary association yield? Hume’s explanation starts from the customary causal associations creatures introduce into their perceptions early in life that makes them (darkly) conscious of thoughts as memories, of visual changes as effected by tactually felt bodily movements, of sensations and memories causing the mind to transition to belief-imbued thought of their associates, of believed thoughts triggering emotions and passions, of these rousing desires and aversions, of those in turn prompting volitions to act, of tactually sensed actions then causing visible, audible, and other sensible effects, and of their effects in turn initiating a new cycle of consciousness of the ever-changing causal flux of one’s perceptions. The objects connected in these relations may change from instant to instant and vary over time, the causal connections themselves may be multifarious, and the causal character and dispositions affecting the entire causal system may alter dramatically through maturation, senility, chronic victimization, imprisonment, or other life-changing events. But through it all one thing stays the same: so long as customary causal association remains continuously “on line,” never having to be reformed from scratch, consciousness of causal necessity affect—an amalgam of intense facility and vivacity—will be continuously refreshed, absent only when consciousness itself is (dreamless sleep, coma, etc.). It thus constitutes a single sustained affective disposition beginning with the first customary causal association between perceptions formed in a creature’s infant mind and ending only when the neural organ of consciousness is definitively snuffed out (unless the physiological basis for customary causal association is somehow lost before then because of, e.g., a stroke). And (dark) consciousness of this sustained affective disposition is, according to Hume, none other than the consciousness of an enduring self common to human and non-human animals alike.25 None of this is to say that the self is explicitly aware of itself as being at the locus of any of these causal relations. As noted earlier, facility and vivacity seldom if ever emerge from dark consciousness, so that, at the level of access consciousness, causal powers are always attributed to the objects perceived, never to the consciousness perceiving them. Of course, once dark self-consciousness is in place and it becomes possible to distinguish externality/independence from internality/dependence, some of the objects involved in causal relations will be regarded as internal and dependent, e.g.  causes such as the will as manifested in voluntary

Hume’s Cement 121 movements and thoughts (e.g. daydreaming), and effects such as pleasures and passions. But prior to actual consciousness of that distinction, the self is simply the darkly perceived sustained affective locus of all consciousness of causal relations in the objects present to consciousness, whether sensations, thoughts, passions, or volitions. It is thus an idea of self that “arises from consciousness,”26 with complete indifference to the contents that present themselves to it. Indeed, memories can gradually all be forgotten and replaced by entirely new ones once, twice, or any number of times without that terminating the existence of the self—just so long as customary causal association never at any point in time has to be entirely rebuilt from scratch. For in that case necessity affect will be sustained through every conscious moment, interrupted only when consciousness itself is, and terminating only when the death of the brain destroys all customary association at a stroke. The upshot is that there is nothing very personal about the Humean self. In theory, it could exist in different bodies, live entirely different lives, exist in radically different times and places, all with no memory of any of the others, and still be the same self, provided only that there was causal continuity sufficient to sustain necessity sentiment through every successive instant of consciousness. In this respect, it is closer to the notion of the soul than to personal identity as usually conceived. My soul is what would still be me even if, as a fetus, I had been magically transported to a different womb and born into an entirely different reality (time and/or place) than the one familiar to me. Indeed, all that is necessary is the continued existence of the part of my fetal brain that constitutes my associative imagination: if that could be sustained in the body of a creature of another species, an android, or any other suitable housing, my soul might have lived a life radically different from my present one while still retaining its—my—identity (which requires only that affective causal continuity be sustained through the continual flux of causally related perceptions). All of which is just to say that the self explicated in Hume’s account of personal identity is really just the foundation on which all I think of as me is erected, not the totality. And, more particularly, it is that in me which enables me to represent some of the objects I experience as having an existence distinct from me, and so as both external and independent. I shall spare the reader a full résumé of Hume’s lengthy psychologistic account of conception and belief in external objects. Suffice it to say that his focus was exclusively on the “vulgar” conception of body common to human adults, infants, and a wide swathe of the animal kingdom. This conception directly equates objects present in sensation, as well as any objects in thought associated with their existence, with sense-divide transcending external/independent physical reality: the shoes I  feel and see on my feet are physically real, so too the vindaloo I see, smell, taste, and feel, the person I  associate with the operation of the lawn mower

122  Path to Kant: Psychologism and Conventionalism I hear outside, and so on. It contrasts with the “philosophical” view of physical reality, according to which sensations and thoughts are all internal/dependent existents, tactual and visual included, and only the unperceived objects causing them (including causes in my brain) have external/ independent existence. Rejecting the latter as a merely “palliative remedy,”27 Hume focused on the vulgar view, where the principal challenge was to explain how we are able to believe two contradictory things at once: that all our sensations and thoughts are dependent existences internal to the self and that some—tactual sensations of space, visual and other sensations correlated to them, as well as any and all of their associated objects in thought—are external to the self and independent of its existence. Again, suffice it to say that Hume’s account purports to show that associative imagination (facility and vivacity) is all the mind requires to conceive and believe both things simultaneously without being in the least troubled by, or ever noticing, their incompatibility.28 For present purposes, the important thing is the part played by tactual sensations of bodily spatiality in this contradictory duality. Represented as internal/dependent with respect to the associatively generated self, tactual sensations of one’s body fuse with self-consciousness to yield the conception and belief that the mind is embodied and the body ensouled. But represented as external/independent with respect to that same self, the body of tactual sensation is believed to be one physical object situated among other such objects, all comprehended within a natural world governed by particular causal laws, the uniformity principle, and the general causal maxim. Thus does one’s subjective personal identity become one’s objective human identity—or lion identity, eagle identity, orca identity, shark identity, etc.

The Achilles’ Heel of Empirical Psychologism Hume’s empirical psychologism defends the proposition that, apart from linguistic convention and its uniquely human social infrastructure, our psyche is in no significant wise different from other animal psyches. Of course, far from denigrating us as lower than we suppose—mere uncomprehending brutes—his point was instead that animal consciousness is in all respects on as high a level as ours is naturally (i.e. artificial enhancements like language aside). If true, anthropomorphizing is not the principal danger facing the psychologist but its contrary, anthropo-denial. All animals capable of associative consciousness understand themselves and their world causally just as we do, generally no less than particularly; they too use their idea of self to represent objects external to themselves, including those not currently present to their senses; and they too fuse their personal identity with tactually apprehended bodily spatiality to become self-conscious embodied (hominoid, cetacean, avian, etc.) persons existing in a physical world, subject everywhere and always to ­natural laws.

Hume’s Cement 123 And why should we be surprised at sharing our cognitive psychology with so large a swathe of the animal kingdom given that our anatomy endows us with so many of the same basic sensations, passions, desires, and volitions as well? To Hume, human no less than non-human animal understanding is fully as affective in nature as these other components of conscious mentation, being, in essence, entirely a matter of experienceactuated, idea-enlivening customary association: Experience is a principle which instructs me in the several conjunctions of objects for the past; habit is another principle which determines me to expect the same for the future; and both of them conspiring to operate upon the imagination make me form certain ideas in a more intense and lively manner, than others which are not attended with the same advantages. Without this quality by which the mind enlivens some ideas beyond others (which seemingly is so trivial and so little founded on reason), we could never assent to any argument nor carry our view beyond those few objects which are present to our senses. Nay, even to these objects we could never attribute any existence but what was dependent on the senses and must comprehend them entirely in that succession of perceptions which constitutes our self or person. Nay farther, even with relation to that succession, we could only admit of those perceptions which are immediately present to our consciousness, nor could those lively images with which the memory presents us be ever received as true pictures of past perceptions. The memory, senses, and understanding are, therefore, all of them founded on the imagination, or the vivacity of our ideas.29 What principally sets modern humans apart is convention, particularly the linguistic kind. Other creatures may have various social understandings with one another, but nothing that remotely approaches the kind able to support linguistic convention and the highly sophisticated sociocultural infrastructure it requires. More particularly, there has never been and probably never could be mathematics, science, or technology, or any other intraspecifically transmissible corpus of public knowledge of the least complexity, were it not for bodies of linguistic conventions such as English, some including writing, that are practiced by more or less complexly interlinked human individuals and populations. Conventions no doubt require a specially equipped psychology.30 Nevertheless, they are by their very nature artificial social constructs that can never subsist independently, in individual, socially isolated psyches. Thus, nothing that depends on the artifice of convention can do anything to set human nature apart from animal. I will reserve the topic of conventionalism for my discussion of Hume’s skeptical challenge in the next chapter. To conclude the present one, there

124  Path to Kant: Psychologism and Conventionalism are two further points that need to be mentioned to round off the examination of Hume’s empirical psychologism. 1. That the moon’s gravity is responsible for the tides, that the same force causally explains planetary orbits, that a certain enzyme catalyzes a particular chemical reaction in eukaryotic cells, and similar such causal beliefs are not associative in nature, even for their originators, and so might seem counterexamples to Hume’s principle that belief (vivacity) in matters of fact and real existence always follows associative relations (facility). His term for these was “philosophical causal relations,” which he readily admitted exist, but reconciled with his principle by contending that their belief-producing capability is parasitic on associative “natural causal relations,” i.e. customary causal associations.31 This is in the first place because philosophical causal relations depend both for their discovery and for their proof on repeated encounters with resembling contiguous appearances. Yet the only thing that prevents resembling and contiguous appearances from disappearing into the background noise of perception and makes us even darkly conscious of them as related is the facility human-animal nature determines us to feel in the transition from one to the other and so marks them off from all non-­resembling, noncontiguous ones. In the absence of this natural associative foundation, is it conceivable that we would even notice, much less base belief on, these or any other peculiarities in the appearances of things? Or, alternatively, if human-animal nature determined us to associate on the basis of features of appearances that presently are processed as noise rather than information? Clearly, without a foundation in natural associative relations of resemblance and contiguity, it seems inconceivable that philosophical causal consciousness would even be possible, much less capable of eliciting belief. But the primary reason Hume regarded philosophical causation as parasitic on natural is the indispensability of the notion of necessary connection to all concepts of causal relations. After considering all the alternatives (other than the one on which Kant would alight), Hume concluded that there can be no consciousness of any associative transition of thought as necessary—inevitable, unpreventable, irresistible—absent strong customs. So, even if we could be conscious of regularities (i.e. resembling contiguities) purely “philosophically,” i.e. without depending on associative facile transitions, lack of the custom-actuated necessity affect would preclude consciousness of the existence of any one thing as necessitating the existence of any other, or of the impossibility of the second existing had the existence of the first not preceded it. Since this is just to say that there is no way to form the idea of a causal relation between distinct existences (objects, actions, motions) except by means of strong customary associations, it follows that we could never believe such relations to exist otherwise, either naturally or philosophically, since “we can never have reason to believe that any object exists

Hume’s Cement 125 of which we cannot form an idea.”32 Only by building atop a foundation in natural causal relation can belief in causally necessary connections be extended into more “philosophical” realms, as happens when we abstract and generalize so far beyond those foundations that facility is lost, especially insofar as artificial means are involved—language, collaborative social organization, mathematics, technology, etc. (“philosophical decisions are nothing but the reflections of common life methodized and corrected”).33 The upshot of Hume’s analysis of philosophical causal relations as wholly parasitic on natural for their cognitive efficacy is to confer on association a genuinely cosmological dimension: It will be easy to conceive of what vast consequence these [associative] principles must be in the science of human nature if we consider that so far as regards the mind, these are the only links that bind the parts of the universe together or connect us with any person or object exterior to ourselves. For as it is by means of thought only that anything operates upon our passions, and as [contiguity, resemblance, and causation] are the only ties of our thoughts, they are really to us the cement of the universe.34 2. It will forever stand as testament to Hume’s exemplary character that he publicly owned up to a possibly fatal flaw in his associationist psychologism as soon as he recognized it: [A]ll my hopes vanish when I come to explain the principles that unite our successive perceptions in our thought or consciousness. I cannot discover any theory which gives me satisfaction on this head.  .  .  . I must [therefore] plead the privilege of a skeptic and confess that this difficulty is too hard for my understanding.35 Hume’s quandary is manifest in any transition of thought. Transition implies a succession of perceptions in one and the same enduring consciousness. Insofar as association requires consciousness of such transitions, it follows that one must already be able to represent consciousness as continuing identically through time before associative imagination is even possible. Since the problem is hugely compounded when consciousness of myriad such transitions is required, as in Hume’s psychologistic explication of the self, there can be no escaping the conclusion that personal identity through time cannot originally arise from sustained associative affect, but must instead have a pre-associative basis that has nothing to do with facility and vivacity affects. That is where Hume threw up his hands. His philosophical principles obliged him to reject all metaphysical explanations of the unity of consciousness, whether dualist, monistic idealist, or monistic materialist. He was instead irrevocably committed

126  Path to Kant: Psychologism and Conventionalism to a purely psychologistic approach for which vulgar understanding is the understanding. Thus, upon discovering that there is an even more primary self-consciousness than the one yielded by association, and seeing no way to explain it on his principles, Hume saw no option but to concede defeat. Yet, he did not presume to pronounce the problem “absolutely insuperable; others, perhaps, or myself, upon more mature reflection, may discover some hypothesis that will reconcile these contradictions.”36 Alas, Hume never announced a new hypothesis. Nor did anyone else before Kant, who seems to have been the first post-Humean to appreciate why one is even needed.

Notes 1. Conclusion of Hume’s 1740 abstract to the Treatise. 2. “The very nature and essence of relation is to connect our ideas with each other, and upon the appearance of one to facilitate the transition to its correlative” (T 1.4.2 ¶34); “the nature of relation and that facility of transition which is essential to it” (1.3.1 ¶3); “easy transition is the effect, or rather essence of relation” (1.4.3 ¶3); “the very essence of these relations consists in their producing an easy transition,” so that rather than observing “some real bond” among our perceptions we “only feel one among the ideas we form of them” (1.4.6 ¶16); etc. 3. T 1.3.11 ¶4 and ¶11. 4. This is true of only relations that concern matters of fact and real existence, not logical and mathematical relations, where belief does not come into it except in application (applied mathematics depends on causal association: T 2.3.3 ¶2). So far as properly epistemological matters are concerned (justification, propositional truth, etc.), Hume’s associationist account of belief removes them from the sphere of human nature and re-situates them in the sphere of socio-cultural artifacts, and, more particularly, conventionally grounded practices, especially linguistic ones. Hume’s conventionalism will be examined in the next chapter. 5. E 5.2. ¶¶11–12. Like most of Hume’s remarks on belief, his focus in this passage was on how unbelieved thoughts come to be believed. At T 1.3.5 ¶6, however, he makes explicit that belief conferred on what is present to the senses and memory is of the same nature: vivacity affect. (Here and elsewhere I draw indifferently from Hume’s Treatise and Enquiry because I see no significant substantive divergence between them.) 6. T 1.3.1 ¶7. 7. T 1.3.9 ¶6. 8. T 1.1.7 ¶16. The idea is that once an associative resemblance relation becomes ingrained custom, any new stimulus suited to the relation will trigger the custom and so be comprehended as an instance of the custom-defined kind. This is the empirical psychological conception of logical universality that Kant’s (Chapter 7) was intended to supersede. 9. T 2.3.5 ¶1. 10. T 1.3.9 ¶¶ 6–7. 11. T 1.3.8 ¶13.

Hume’s Cement 127 12. T 1.3.12 ¶14 and T 1.3.11 ¶11. The custom “forces us to survey such certain objects, in such certain relations . . . by a natural transition which precedes reflection, and which cannot be prevented by it” (T.3.11 ¶4 and T 1.3.13 ¶8). 13. E 7.2 ¶28. 14. E 4.1 ¶8. 15. T 1.3.14 ¶27. 16. T 1.3.9 ¶¶ 3–4. 17. E 1.4.2 ¶23. Human-animal nature determines belief in causal uniformity to exist in “the most ignorant stupid peasants, nay infants, nay even brute beasts.” 18. Quantum probability may be a counterexample, but no creature can ­conceive— much less affirm—quantum mechanical concepts in a state of nature, without benefit of language, considerable education, and highly advanced technology. 19. Appendix on personal identity to A Treatise of Human Nature ¶11. 20. The custom relates only to the contents of thoughts, not to their arrangement. Thus, imagining a city paved with gold is not a counterexample since everything in the fantasy (i.e. its contents) will have been previously encountered via the senses. Even so, association being empirical, Hume not only did not preclude exceptions, he even proposed one. Suppose there happened to be a shade of blue that one had never seen despite having seen all the neighboring shades. If the latter were all laid out from lighter to darker with a gap left for the missing shade, Hume thought it likely that one could supply the missing shade by means of imagination, even without memory. Yet, to count as a contrary instance to the custom it would have to actually happen (the merely logical possibility of its happening is not a counterexample to the experience on which the custom is founded). And even if such a case did occur, one would be highly unlikely to notice that the particular shade of blue in one’s thought was not a memory but something imagined with no sensation precedent, and so would be certain to continue in the full conviction that everything in thought resembles something that had formerly been present in sensation. 21. T 1.4.2 ¶9. The clause is preceded by this: “properly speaking, ’tis not our body we perceive when we regard our limbs and members but certain impressions which enter by the senses.” 22. T 1.4.6 ¶19. 23. T 1.4.6 ¶4. 24. T 2.1.12 ¶4. 25. See T 1.4.6 ¶¶ 6 & 18–19, and discussions in KEUU chapters 16-F and 17-B. 26. Appendix on personal identity to A Treatise of Human Nature ¶11. 27. T 1.4.2 ¶46. Hume held that the philosophical “double existence” view “contains all the difficulties of the [“single existence”] vulgar system with some others that are peculiar to itself.” For a detailed exegesis of Hume’s accounts of the self, external objects, and their relation, see HTC part III and KEUU chapters 16–17. 28. Hume held that there are many cases where the vulgar not only can but cannot help believing contradictory things, most likely because some beliefs are so essential to human existence that human nature has immunized them not only from doubt but even from being noticed. This is discussed under the rubric “natural dialectic” in HTC, especially in the book’s conclusion. 29. T 1.4.7 ¶3. 30. See chapter 5 note §2. 31. “[T]ho’ causation be a philosophical relation . . .’tis only so far as it is a natural relation and produces a union among our ideas that we are able to reason upon it or draw any inference from it” (T 1.3.6 ¶16). The parasitic character of philosophical causation becomes even more evident if, like Hume, one

128  Path to Kant: Psychologism and Conventionalism further accepts that only natural causation can provide us with conscious memory, representation of sense-divide transcending physical objects, consciousness of the soul and the external/internal distinction it makes possible, belief in the uniformity of nature, the maxim that every beginning of existence is caused, and a host of other indispensable custom-based associative causal beliefs acquired perfectly naturally, long before we are out of our cradles, without the least “philosophical” assistance. 32. T 1.3.14 ¶36. For further discussion see HTC chapter 5-E. 33. E 12.3 ¶25. 34. Conclusion of Hume’s 1740 abstract to the Treatise. 35. Appendix on personal identity to A Treatise of Human Nature ¶¶ 11–12. 36. Ibid.

5 From Hume to Kant via Wittgenstein

Hume’s thesis that associative psychology is the cement of the universe so far as our consciousness of the natural and human worlds is concerned was a gauntlet thrown down before theorists of mind of his and subsequent eras. It reduces all non-linguistic human conception, judgments, and inferences to processes fully as affect-driven and devoid of propositional or other logical structure as presumably drive non-human animal intelligences and govern both human and animal passions. In so doing, it effectively cuts the ground from under all claims of human exceptionalism since the universe represented by the human psyche becomes, in all fundamentals, no different from that of animal psyches generally. The challenge facing Hume’s opponents, then, was to show that associative psychology suffices in neither regard, and so to identify what makes human intelligence psychologically unique—or at least more exceptional than Hume would have it. The Humean challenge is two-sided. The first is the claim considered in the previous chapter that the explanatory power of associative psychology has been grossly underestimated, being nothing less than the cement of every creature’s representational universe, humans included. The second, which I shall explore now, is his thesis that conscious mind by itself is incapable of supporting any other kind of representation. This relates particularly to propositional thought, its logic, and the languages that are its vehicle: any representation that depends on these is not a mental representation at all—a consciousness with dedicated neural correlates—but something else entirely. In particular, all are, in Hume’s view, dependent on convention, and therefore irremeably sociological.

Conventional vs. Psychological Representation Behavior is governed by established conventions when there is “agreement . . . without the intervention of a promise . . . [as] men who pull the oars of a boat do it by an agreement  .  .  . though they have never given promises to the other,” as “gold and silver become the common measures of exchange,” and as “languages are gradually established by

130  Path to Kant: Psychologism and Conventionalism human conventions without any promise.” Conventions arise whenever a person implicitly comes to rely on another to respond to an action he or she performs with a particular answering action, both are sensible of “a like interest” in perpetuating the routine, and that interest “is mutually expressed and known to both.”1 When others, observing them, conform their behavior to theirs because they share, recognize, and make known the same interest, the convention spreads; and if its spread is far and deep enough to endure in a population across generations, it becomes established convention. If the artifice of established convention is taken away so that everything is left to nature to regulate, oarsmen cannot then avoid chaotic motion, gold cannot serve as a medium of exchange, and, above all, language becomes a complete nonstarter. And generally: whenever nature fails to regulate behavior but a need for behavioral regulation is felt, human social conventions artificially fill the void. Linguistic conventions are a species of established convention that fit seamlessly into a web of other social practices that together constitute a society with its own distinctive culture, all embedded in species-wide, uniquely human lifeways. Conscious representations, by contrast, are correlated not to social networks but to neural ones, and so are as informationally blank where conventions are concerned as they are regarding their neural correlates. From Hume’s perspective, the propositional thought characteristic of human populations is on all fours with crows flying in murders and dolphins swimming in pods: collective social practices no more discoverable from individual participants’ neurons than from the consciousness correlated to them. To be sure, putting socially constituted conventional meaning to use involves considerable neural activity, including much that is conscious.2 But that does not make the conventions themselves any less incapable of existing in an individual isolated consciousness, independently of the infrastructure of collectively established practices. Hume’s view of language3 is thus in complete harmony with the far more developed conventionalism of the twentieth-century philosopher Ludwig Wittgenstein, who, in the Philosophical Investigations and other late period writings, equated propositional thought— science and mathematics not excepted—with conventionally constituted moves in language games embedded in distinctively human forms of life. It is vital to recognize that the thesis defended by Hume, Wittgenstein, and others concerns human understanding, not human knowledge (Chapter 2). It does not reduce the truth of uncontroversial factual reporting, established science, and demonstrated mathematics to mere “truth by convention.” What a home run is may be determined by baseball’s conventions, but when a batter hits a pitch into the left-field upper deck it is factually, in objective truth, not merely conventionally, a home run. Similarly, when mathematicians devise demonstrations or scientists theorize, observe, and experiment, the knowledge that results is factual, objective truth, not true merely by convention. Nevertheless,

From Hume to Kant via Wittgenstein 131 by contrast with sensations and associative relations, mathematical and scientific understanding is fully as convention-dependent as the understanding of games like baseball, and so just as impossible unsupported by the infrastructure of collectively established human social practices. Yes, mathematics and science have their own special coinages—integrals, matrices, constants, fermions, bosons, et  al.—but then so does baseball. No one has ever taken a single step in mathematics, conceived and executed experiments, programmed a computer, built a nuclear reactor, or done anything else in the least scientific except by relying, implicitly and explicitly, on the broader language spoken by their contemporaries and handed down by their forebears. Mathematical and scientific understanding, like convention-dependent human understanding generally, are social inheritances comprehensible only in and through the broader language in which they are indelibly embedded, the social infrastructure on which language depends, and the species-wide lifeways that define uniquely human behaviors. Indeed, linguistic convention is so indelibly bound up with our humanity that Wittgenstein famously averred that “if a lion could speak, we could not understand it.”4 This book’s brief does not allow me to provide a full résumé of human social conventions and the myriad ways they are essential to language. It is sufficient to recognize that linguistic representation is not bound by the constraints that bind psychological representation to all and only such contents of consciousness as one’s own, strictly personal experience has disclosed, with no more relation or unity among these contents than single, isolated acts of representation are capable of conferring. Linguistic convention could not be more different, as Wittgenstein’s comparison of language to a city illustrates: Do not be troubled by the fact that [my two invented languages] consist only of [primitive building site] orders. If you want to say that this shows them to be incomplete, ask yourself whether our language is complete;—whether it was so before the symbolism of chemistry and the notation of the infinitesimal calculus were incorporated in it; for these are, so to speak, suburbs of our language. (And how many houses or streets does it take before a city begins to be city?) Our language can be seen as an ancient city: a maze of little streets and squares, of old and new houses, and of houses with additions from various periods; and this surrounded by a multitude of new boroughs with straight regular streets and uniform houses.5 Languages are like cities in being multi-generational collective enterprises that exist to accommodate and facilitate all the activities taking place within them. Cities do this with bricks and mortar, paving and cable, vehicles of various kinds, multifarious distribution systems, sundry services, communication networks, and much else, all planned and regulated

132  Path to Kant: Psychologism and Conventionalism at the micro- (individual), meso- (group), and macro- (citywide) level. Languages do it through conventions—rules everyone tacitly agrees to abide by. Linguistic conventions are a particularly strong building material in that, almost uniquely among conventions, everyone has an interest in always conforming to them and no one is ever tempted to flout them, since to do so is to veer straight into incomprehensibility; and who would ever want what they take the trouble to say not to be comprehended? Equally importantly, linguistic conventions are myriadly versatile and endlessly adaptable. Some are constitutive of meanings, others of grammar and syntax, and still others of phonology, but that barely scratches the surface. Linguistic conventions are constitutive of orders, requests, desires, statements, and questions. They constitute the discourse of teaching, learning, and testing, of story-telling, reporting, and jesting, of legal briefs, business prospectuses, and economic programs, of competitive sports, board games, and military drills, of theatricals, dances, and chorales, and so on and on. They are constitutive of the formulae used in calculus and chemistry, rocketry and artillery, bookkeeping, and metric conversion. Linguistic conventions of myriad kinds, connected in numberless, interlocking, multi-layered ways, make possible all the cooperative, competitive, and other kinds of coordinated verbal behavior that together go to make up language. And like the city in which we are born, conventions are not something we each individually have to create from scratch like experience, but are a common inheritance passed down from generation to generation, silently modified in manifold ways to suit the expressive needs of each. Nor does the analogy end there. Crucially, just as everything that takes place in a city is situated at a particular node in a citywide network and determined accordingly, every act of conventional representation takes place at a convention-constituted node in the wider network of language and is informed thereby. Thus, according to the nodes to which other conventions directly connect it, the passageways they provide to remoter nodes, and all manner of language-wide connectivity in which they embed it, each such representation has additional, multi-layered meaning, grammatical relatability, and syntactic orderability conferred on it. This makes conventionally constituted meaning, logic/grammar, and syntax inherently holistic,6 and so inherently impossible to capture by purely psychological representations which, by their very nature, are not situated in or informed by a preexisting, meaning- and structure-conferring network. On the contrary, the inviolable rule in the case of the latter is that no content, relation, or order exist except in and through the single acts of representation responsible for bringing them into psychological being. Psychological representations are thus prototypical hermetic isolates, as confined to the individual, isolated, wholly private consciousness of the psyche doing the representing as their neural correlates are to the brain responsible for producing them.

From Hume to Kant via Wittgenstein 133 Nor is there any equivalent in psychological representation to the land cities require—land of the appropriate sort for building cities atop—or its linguistic analogue: the wider social infrastructure and species-wide lifeways on which human conventions generally and linguistic conventions more particularly wholly depend. Instead, the representing subject has to produce the underlying land for its structures—the understanding that supports them—along with the structures themselves. This is just to say that psychologically generated representations have only so much and no more meaning or structure than the subject doing the representing puts into them: they connect up with others only insofar as the subject produces a connection between them, have complexity only insofar as the subject synthesizes them from simpler ingredients, and combine to form unities only insofar as the subject fashions a unity from them. Nor is there ever anything other than fallible memory to ensure such subjective accretions to representation get preserved. Thus, in the absence of the holistic representations constituted by linguistic conventions, there is little if anything to distinguish human understanding from non-human animal varieties—just as Hume maintained. None of this is to deny that there exists a certain continuity between psychological representations and at least some socially constituted conventional ones. For example, conscious representations distinguishing between one, two, many, big and small, more or fewer, etc. seem to lie within the scope of non-human animal understanding. Although not conventional, and so lacking the logico-grammatical character that fits them for linguistic use, they are on a continuum with convention-based elementary representations of quantity and number that have this character. Similarly, representations can be generalized purely consciously by means of customary resemblance associations—the sortals of humananimal natural consciousness—without requiring conventions or being in any way logically or grammatically fitted for inclusion in propositions. Nevertheless, such generality seems continuous with the logically and grammatically structurable variety characteristic of convention-based linguistic generality. Or, again, the loss of vivacity affect from a previously believed idea—disbelief in the reality of what is represented in it— is on a continuum with the logical act of denying a proposition—disbelief in the truth of something that cannot be represented except via linguistic propositional logic and grammar. And much the same is true of conscious and convention-based logically and grammatically structurable representations of sensations like pain (“Ouch! That hurts!”), sensibly detectable patterns like shapes (“That is a cube, not a sphere”), as well as emotions, desires, and much else. But however great the appearance of a continuum, it should not obscure the very real and fundamental difference between conscious representations in the individual, isolated psyche and culturally established, socially constituted conventional representations. There is nothing in any

134  Path to Kant: Psychologism and Conventionalism individual’s consciousness corresponding to the latter. Even such conscious representations as may be needed to operate with conventions (joint attention et  al.)7 are only ancillary to one’s actions’ conformity to the socially established norms that shape both others’ expectations of one and one’s expectations of others. Stripped of all relation to existing conventionally prescribed verbal and non-verbal social practices, conscious representations lose whatever linguistic meaning and logico-­ grammatical fitness their seeming continuity with non-mental representations might be supposed to confer. For intrinsically and of themselves, they are altogether devoid of all properly linguistic attributes, and so are, in the most literal linguistic sense, senseless. Conversely, a machine devoid of consciousness that was programmed to perfectly imitate how humans operate with their conventions would rightly be adjudged to have mastered the same non-psychological representations we do; and if it were designed to look and act like a human, we would probably find it impossible not to accept it as one of us—even if we were convinced that it lacked everything that makes conscious representation possible. In short, for Hume, and Wittgenstein after him, conscious representation and linguistic propositional thought are not two species of the same genus but different things altogether, one part of nature and purely psychological, the other irreducibly sociological and as artificial as military protocol and rules of the road.

Humean Skepticism Accepting that psychological and conventional representation cannot be understood as species of a single genus does not prevent one from acknowledging that linguistic conventions are no less perfectly fitted to expressing what’s on our minds than heavy fur-lined gloves are perfectly fitted to keeping hands warm in a Siberian winter. It is thanks to all that language makes possible that there were shoulders for Newton to climb upon to see further, just as it is language that enables any of us to see so vastly much more than did our pre-linguistic ancestors and non-linguistic animal relatives. But this is not because language is some kind of mental enhancement like an additional sense or psychological capacity would be. Instead, it is merely a toolbox of cultural implements that allows knowledge to be pooled and expanded further, faster, in more ways, and more reliably than it otherwise could. In particular, we must guard against being deceived by the near perfect adaptedness of linguistic tools to our expressive needs into taking them as proof, or evidence of any kind, of the existence, or even the possibility, of a corresponding conscious representation. Indeed, from Hume’s point of view, the dependence of linguistic representation on convention is proof of precisely the opposite. A case in point of the need for caution is the method of definition favored by Rationalists like Descartes, Spinoza, and Leibniz.8 They sought to allay

From Hume to Kant via Wittgenstein 135 any suspicion that a term is merely an artifact of language, with none but verbal validity, by supplying a description capable of satisfying the kind of strict canons of definition that guide mathematics and science. Such definitions were purported to be proof positive that the term in question has extra-linguistic, objective meaning—a meaning in principle graspable by the psyche even in the absence of language and capable of being instantiated without reference to language or anything else inextricably bound up with human sociality. That is precisely what Hume questioned. He doubted whether descriptions in language, however rigorous and canonical, can ever constitute evidence of the psychological realizability or language-independent instantiability of a representation. The locus classicus of Hume’s critique is the putatively objective (consciousness-independent) concept of cause and effect, regarded by many in his day as so entirely unimpeachable as to be indubitably objectively valid. While descriptions of the concept varied, virtually all converge on the following features: If X is the cause of Y, then the existence of X necessitates the existence of Y such that it is impossible for Y not to exist if X does and impossible for Y to exist if X does not. If experience seems to offer counterexamples, it is merely because X or Y or both have been misidentified—i.e. either one’s candidate for X is not the (complete) cause of Y or one’s candidate for Y is not the (complete) effect of X. Yet, even if we never successfully identify the cause and/or its effect, since it is self-evidently impossible for any object, action, or movement to come into existence without some cause necessitating its existence, every existent Y must have a unique existent X as its (complete) cause, and every existent X must have a unique existent Y as its (complete) effect. If only for argument’s sake, let’s grant that this definition captures something central, even indispensable, to the objective concept of cause and effect. How, Hume would have us ask, does that prove that a purely psychologically generated conscious representation corresponding to the concept is possible? Hume’s argument that it does not and cannot—that it is not any kind of evidence at all for the concept’s psychological realizability—starts from the observation that consciousness of a causal relation between any two items is possible only if they can be represented as existing separately. We cannot, for example, represent mountains as causes of valleys or valleys as causes of mountains because mountains cannot be conceived at all without also conceiving adjacent valleys and vice versa. Only items that can be conceived to exist without the conception of the one’s existence requiring the conception of the existence of the other are candidates for causal relation. If I  can conceive the dram of whiskey I  drank last

136  Path to Kant: Psychologism and Conventionalism night separately from the mass extinction at the end of the Permian, then I can conceive that very dram as having caused it. If I can conceive of leaping 10 miles into the air, descending 500 miles through solid earth, and returning faster than light speed to where the leap began, then I can conceive myself performing this precise sequence of actions. Conversely, since I  can conceive fire to exist without having to conceive smoke as existing, I  am free to conceive fire as the cause of rose fragrance and tequila as the cause of smoke. I am similarly free to conceive ipecac as the cause of earthquakes rather than vomiting, or the dust mote currently wafting past as the cause of the Big Bang. Anything can be conceived to cause anything so long as consciousness of the existence of the one does not logically necessitate consciousness of the existence of the other.9 What Hume noticed that no one previously had, or at least not recognized as important, is that the distinctness presupposed by the foregoing definition of cause and effect as an objective relation cannot be combined in thought (i.e. psychological representation) with the necessity of the relation. The separate conceivability of X and Y implies the conceivability of one existing without the other, or indeed either existing without anything else ever existing, which is precisely what representing X and Y as necessarily connected in an objective relation of cause and effect precludes.10 This sets up the following predicament: for X and Y to be candidates for causal relation, their existence must be distinct in a sense that precludes causal relation, whereas not to be distinct in this way precludes them from being candidates for causal relation in the first place. Or, more generally still: if a representation is psychologically impossible, then its logical possibility is moot, and if it is logically impossible, then its psychological possibility is moot. By contrast, the interlocking, multi-layered semantic, logico-grammatical, and syntactic representational structures possible within the preexisting framework constituted by linguistic conventions are not in the least constrained by psychological impossibility. So, if the definition of the objective relation of cause and effect is supported therein, then even if its psychological impossibility is conceded, it is linguistically just as meaningful as the equations of calculus and the laws of physics—or, for that matter, my horoscope from last Tuesday, the Ten Commandments, and der Ring des Nibelungen. After all, if, as seems to be the case, the psyche has nothing to take the place of convention in the economy of r­ epresentation—­nothing able to perform the same or comparable ­functions—then no representation dependent on convention is ever, strictly speaking, possible purely psychologically. All of which simply serves to underscore Hume’s general point: the definability of a concept has nothing to tell us about its possibility as a psychological representation. And more particularly: unless one can refute Hume’s skeptical argument against the psychological possibility of the objective concept of cause and effect, any hopes placed in canonical description as a method of proving the extra-linguistic meaningfulness of this or any other representation that requires that necessity

From Hume to Kant via Wittgenstein 137 be combined with distinctness in a single act of thought must be acknowledged to be completely unfounded.

Kant Stopped Dead in His Tracks Hume’s skeptical reasoning struck Kant like a thunderbolt. He credited it with waking him from his “dogmatic slumber” by showing that the concepts of objects and their relations he had thitherto regarded as most indubitable of all—not just cause and effect but all the purportedly objective categories of traditional metaphysics—are in fact all psychologically impossible inasmuch as each posits an extra-logical necessary relation between items presupposed as distinct. The concept of substance-accident is a case in point. To conceive X to be an accident of Y, X and Y must be conceived as distinct in existence. For if there were no more difference in their existence than there is between that of mountain and valley, the relation would be purely conceptual (logical) and not a real relation of existence at all. Yet, if conceiving X to be an accident of Y implies that the existence of X depends on that of Y, then the existence of X must be conceived to be necessarily connected to that of Y, i.e. impossible apart from it. Since this contradicts the presupposed distinctness of the existence of X from that of Y, the category of substance-accident is likewise revealed to be fully as vulnerable to Hume’s skeptical argument as that of cause and effect. But what alarmed Kant most is that the same reasoning, unbeknownst to Hume, extends to the necessary relations affirmed in mathematics. We need look no further than “2 + 2 = 4” to understand how. No one disputes that the equality between the determinations on either side of ‘=’ is a necessary relation. But Kant, donning his Humean skeptical hat, noted that the representation of the sum of 2 and 2 is as distinct from the representation of 4 as both are from the cube root of 64, the difference between 1,043,742 and 1,043,738, and an infinite number of other numerical representations that partake of the same necessary relation of equality. So, the question again becomes how representations presupposed as distinct can possibly be necessarily related. To see why they cannot, simply take any pair of representations from this infinite number: even if you analyze each till kingdom come, because of their distinctness, you will never find the other in it; and if the other is nowhere present in it, how can their equality be found there either? The skeptical thrust of this question emerges from the fact that mathematical cognition has nothing beyond these representations on which to focus; yet, if they are truly completely distinct, there is no option but to look beyond them to discover their necessary relation. And what is there outside distinct mathematical representations to relate them other than experience? Yet, whereas experience can relate objectively distinct existents or matters of fact in contingent relations, it can never issue in the kind of objective, purely intellectual a priori necessary relations that mathematics is celebrated

138  Path to Kant: Psychologism and Conventionalism for being uniquely capable of yielding. The most experience can do is provide the kind of faux necessity Hume ascribed to internal experience of the purely subjective, affect-constituted associations effected by custom (which Kant branded “a bastard of the imagination”).11 But if any psychological representation of the necessity of objective mathematical relations is precluded, doesn’t that leave only the conventional kind? And doesn’t that reduce objective mathematical necessity to a mere artifact of language, with none but verbal validity, like conventionally constituted, socially grounded representations generally? Hume’s skeptical reasoning, as extended by Kant to mathematics and construed to encompass its conventionalist as well as its empiro-­ psychologistic dimension, may therefore be encapsulated as follows. (1) To deny the distinctness of the relata is to deny the objectivity of the relation. (2) However, to affirm the distinctness is to preclude the necessity of the relation. Consequently, (3) no relation can be both objective and necessary, and (4) this applies as much to mathematical necessity as to physical or any other ostensibly necessary relation of the distinct. Thus, (5) all putatively objective necessities without exception are purely semantic and logico-grammatical in character, i.e. merely verbally valid linguistic artifacts impossible apart from convention and its social infrastructure. The implication of Humean skepticism that no objective relations are necessary and no necessary relations objective seemed to Kant at once irresistible and unacceptable: In this way, the principles of Hume’s empiricism lead inescapably to skepticism even in respect of mathematics, and consequently in all scientific employment of reason. . . . Whether, amidst so horrible an overthrow as the one we have seen of the crowned heads of cognition [mathematics, science, and metaphysics], the common employment of reason will come through any better and not instead be caught up irrevocably in this destruction of all knowledge, so that a universal skepticism would have to follow from these principles (which of course would concern only the learned)—that is for each to judge.12 Kant, of course, judged that it would. So, to combat a skepticism that he himself had made universal, he thenceforth made it his principal order of business to prove the extra-linguistic, psychological possibility of objective necessary relations in the guise of what he termed synthetic a priori judgments. For with mathematics providing incontrovertible evidence of their actuality, the real mystery for him lay in their possibility.

What Humean Skepticism Does and Does Not Target To say of a psychologically impossible representation that the only representational reality left to it is as an artifact of language, possible only in

From Hume to Kant via Wittgenstein 139 and through social convention, is to say that the representation has the same legitimacy and validity—no more, no less—as meaningful discourse generally. Viewed through the lens of Kant’s extensions of Humean skepticism, however, isn’t this just to say that mathematics, science, metaphysics, and everything in common, everyday knowledge that would be rendered unrepresentable in the absence of linguistic conventions have the same—no more, no less—legitimacy as religious discourse, astrology, new age spiritualism, cosmogony, stone age mythology, science fiction, fantasy, and any other species of discourse that makes sense to at least some group, some subset of a population of speakers, now or at any point in the history of language? The answer is yes and no: yes so far as understanding is concerned, no with respect to knowledge (this distinction was introduced in Chapter 2). If indeed the representations of customary-association-transcending mathematical, scientific, metaphysical, and ordinary understanding are as psychologically impossible as Hume’s psychologism implies, then it seems that they have, and can have, none but the merely verbal validity that appertains to linguistic artifacts constituted entirely by semantic, logico-grammatical, and syntactic social conventions. But this is to take nothing away from any of them as knowledge, much less to deny the manifest superiority in practical consequence of science and mathematics in particular to all other forms of language-dependent human understanding, from religion and astrology to gossip columns and recipes. Indeed, Humean skeptics are just as ready as the most ardent epistemological maximalists to applaud and endorse the value of scientific and mathematical knowledge for the simple reason that human knowledge is not their target. Instead, their quarrel is with anyone who claims that these or any other species of knowledge incorporate, or in any way involve, representations whose meaning and validity transcend anything that is possible psychologically or conventionally. Their counterclaim is that representations that can be explicated neither psychologistically nor by linguistic convention are devoid of meaning, mere unintelligible nonsense masquerading as sense, notwithstanding any illusions to the contrary. In a word, the true target of Humean skepticism is platonism.

The Anti-Platonist Thrust of Hume’s Skeptical Challenge Hume’s skepticism regarding objective understanding was strongly opposed in his own day and continues to be in the post-Wittgensteinian present. The opposition principally takes the form of what is often termed platonism. Originally, this meant the ideas, or forms, that Plato regarded as fully objective, mind- and society-independent r­epresentations— the beautiful itself, the good itself, intelligence itself, power itself, life itself, etc.—which function as archetypes with respect to both

140  Path to Kant: Psychologism and Conventionalism particulars—beautiful, good, intelligent, etc. people and things—and mental representations—consciousness of the beautiful, good, etc. Insofar as the human psyche partakes of the preexisting framework supplied by these archetypal forms, its representations are held to be genuinely objective, and are otherwise merely subjective ephemera. Equally influential is the Aristotelian variant that naturalizes platonism: instead of particulars partaking in archetypal forms existing apart, forms are immanent to matter, i.e. internal to and inseparable from it. Forms transform what is otherwise indeterminate, undifferentiated matter into mental and physical particulars with a definite nature, or essence, while matter gives real existence to forms that are otherwise pure abstractions, altogether lacking in reality (i.e. in contrast to Plato’s forms, they are not even so much as possible existents in their own right). Aristotelianism was the dominant view until the early modern era, when the rediscovery and translation of Plato’s surviving writings resulted in something of a mix and match. For example, Descartes may be categorized as predominantly Platonic, Spinoza roughly equal portions Plato and Aristotle, and Leibniz more Aristotelian. Although Locke and Berkeley are often portrayed as Aristotelians, this tends to be exaggerated to the extent that the psychological character of their philosophies is downplayed or ignored. Nevertheless, Hume was the first major Western theorist of understanding to break free from platonistic influences completely and devise a purely psychologistic account of understanding that rejects representational objectivism in all its forms. Kant built on that legacy, a bird of the same psychologistic feather as Hume. But then the pendulum began to swing back. Hegel excised Kant’s psychologism and, especially in his Science of Logic (which will be considered in the conclusion), fused what remained into a novel blend of mainly Aristotle, Plato, and Spinoza. Since then, as philosophy has become more and more anti-psychologistic, platonism has enjoyed a revival. Today’s preeminent variant, at least among philosophers, stems mainly from Frege’s analysis of language, which treats linguistic (including mathematical) sense and meaning as objectively prior to and independent of both psyche and society, and so as both phylogeny- and consciousness-independent. Contemporary scientific platonism ranges across a spectrum remarkably similar to that of the seventeenth century. At one end there is platonistic transcendentalism, where objective representations exist phylogeny- and consciousness-independently on their own as an archetypal realm of laws of nature to which all mental and physical particulars ectypally conform. Physicist Max Tegmark’s view that physical reality is simply mathematics viewed through an imperfect empirical looking glass is an evident exemplar of such transcendentalism, but by no means the only one. At the other end is Aristotelian immanentism, where objective representations are given real existence by some kind of material substrate that they in turn mold into the mental and physical particulars of ordinary

From Hume to Kant via Wittgenstein 141 and scientific experience. A good example is evolutionary biologist Ernst Mayr’s praise of Aristotle for anticipating the embodied genotype, but any view of information as phylogeny- and consciousness-independently objective but requiring physical embodiment falls into this category.13 Hume’s skeptical challenge to platonism is propelled by his psychologism. It is fine to claim that objective representations independent of linguistic and other social conventions exist. But unless human understanding has some means of accessing them, they just as well might not exist so far as we are concerned, as inaccessible representations are necessarily nothing to us, and so can have nothing whatever to do with any of the representations actually used by philosophers, mathematicians, scientists, priests, therapists, beauticians, and everyone else. Yet, our access to them cannot be through conventional representation if, as platonists suppose, they are completely independent of human culture and lifeways and would in principle remain accessible to human psyches even if all trace of these were effaced from our experience. So, how else could we have access to them except psychologically, i.e. through some special mental faculty in the wholly private world of the individual, isolated consciousness? Psychological answers were easier to propose when creationism was still in vogue among scientists and philosophers. Malebranche, for example, supposed that, in addition to our natural psychological endowments, human understanding gains access to Platonic archetypes through a miraculous illumination vouchsafed by God. Descartes, Leibniz, and others supposed that ectypes of these archetypes are an innate endowment bestowed on human souls by a beneficent Creator. Since either way we are obliged to introduce the supernatural into human psychology, such notions retain little more appeal today for philosophers and scientists of mind than telepathy, clairvoyance, or prophecy. Contemporary platonists tend to duck the question of the psychological accessibility of their purportedly phylogeny- and consciousness-­ independently objective representations. Instead, they are apt to point to the incontestable objectivity and truth of human knowledge—as if that sufficed to refute a variety of skepticism targeted specifically at the supposed phylogeny and consciousness independence of human understanding and not at all at the veracity of human knowledge. The point of Humean skepticism is that if it is accepted that human psychology is in no fundamental respect different from non-human animal psychology, and that the non-psychological contents of human understanding are linguistic artifacts constituted by convention, culture, and distinctly human lifeways, then none of the contents can be supposed to transcend phylogeny- and consciousness-dependent human nature and artifice, and so none can be objective in anything like the platonist sense. More particularly, it is a skepticism that implies that exactly the same knowledge might be attainable by means of an understanding completely

142  Path to Kant: Psychologism and Conventionalism incommensurable with our own, equivalent solely in epistemic outcomes (the ability to build skyscrapers, perform organ transplants, etc.). The objectivity and truth of knowledge are therefore no argument for the objectivity and truth of human understanding if, as Hume, Wittgenstein, and others maintain, everything in the latter that goes beyond the natural (i.e. non-supernatural) endowment of the human-animal psyche is no less ineluctably human, albeit artificial rather than natural. Objective understanding as contemporary platonists conceive it consists of intelligible (e.g. mathematizable) information that exists wholly independently of both human nature and artifice, and indeed of the psychological and social phylogeny of creatures generally. To that extent, they are exactly like the forms of Plato and Aristotle: human-animal consciousness- and convention-independent representations that, when combined appropriately, yield the “eternal truths” that mathematicians discover and physicists asymptotically approach. Since such truths, if they really exist, transcend convention, mastering language cannot by itself suffice to access them. Ultimately, therefore, the skeptical challenge cannot be avoided: how is it possible for our serendipitously evolved, ephemerally existent mammalian understandings to access representations that exist wholly independently of us, both evolutionarily and culturally? When pressed for an alternative to convention, some platonists gesture in the direction of genetics, i.e. natural rather than supernatural innate representational endowments. But since it seems most unlikely that evolution could ever issue in psychological capacities that serve no other purpose than to produce the phylogeny- and consciousness-independent representations platonists deem indispensable to mathematics, science, and other highly sophisticated branches of knowledge, most seem content to leave the matter sufficiently vague to obfuscate the problem that emerges when psychologistic/conventional and platonist models of “objective” understanding are sharply contrasted. For so long as it is left vague, the unassailable epistemic warrant of conventional objectivity can surreptitiously bleed into platonism’s prime-matter molding forms and confer a patina of the same unassailability on the latter. That is why Hume, Wittgenstein, and others with similar views insist so strongly on highlighting the difference between linguistically based understanding grounded in human language games embedded in no less distinctively human forms of life, on the one hand, and platonism’s unsubstantiated, well-nigh mystical conception of an objective understanding that altogether transcends psychology and convention—a conception exemplified perfectly, according to the later Wittgenstein, by his own earlier Tractatus Logico-Philosophicus—on the other hand. Platonists must either yield to the Humean/Wittgensteinian skeptic or show how the human psyche, independently of both convention and outright supernatural intervention, is able to access the representations requisite to understand and know phylogeny- and consciousness-independent

From Hume to Kant via Wittgenstein 143 eternal truths mathematically, employ them empirically in the quest for scientific truth, and apply them in whatever other modes of cognition to which they may be deemed indispensable (eternal moral truths? religious truths? etc.). Innatism, again, is not an option. Aside from the absence of evidence, there remains the question of how the capacity for such representation can plausibly be supposed to have evolved from minds that lacked it. How, in particular, could it do so completely independently of creatures’ social development, much less have evolved in creatures with no social lifeways at all? One has to imagine that consciousness, starting from being minimally distinguishable from the perfect insensibility of something like Locke’s oyster, then evolving the neural correlates for the kind of consciousness that Berkeley and Hume explained by customary association, could suddenly blossom into the fully equipped understandings of a Gauss or Einstein, all completely independently of human sociality and the conventional representations made possible by it. To me, such a hugely improbable, phylogenetically non-adaptive saltation seems a bridge too far. It thus can still fairly be said that, with innatism excluded, platonists have no satisfactory solution to the accessibility problem, and so must cede the game to the Humean skeptic. Of course, the Humean skeptical challenge is only as good as the psychologism and conventionalism underpinning it. My focus in the remainder of the book will be on Kant’s attempt to show that Humean associationist psychology is a nonstarter unless supported by a transcendental psychology of pure (= strictly a priori) representational capacities of sense, imagination, and self-consciousness. The a priori psychologism he grounded on it purports to give the platonistic impulse its full objective due by showing that representations and truths may be brought to consciousness that transcend anything possible through convention or empirical psychology, yet without obliging one to have recourse to innate endowments discontinuous with human phylogeny, much less anything of a supernatural character. A purely subjective psychological “objectivity” that is nevertheless sufficiently robust to quell the urge to embrace any form of evolutionarily and psychologically unsustainable platonism, even in the cases of logic and mathematics: that is the promise of Kant’s response to Humean skepticism.

Notes 1. T 3.2.2 ¶10. Although Hume’s ostensible topic is property, his examples, including language, leave no doubt that the scope of his analysis extends far more widely: “I observe that it will be for my interest to leave another in the possession of his goods provided he will act in the same manner with regard to me. He is sensible of a like interest in the regulation of his conduct. When this common sense of interest is mutually expressed and is known to both, it produces a suitable resolution and behavior. And this may properly enough be called a convention or agreement between us, though without the

144  Path to Kant: Psychologism and Conventionalism interposition of a promise, since the actions of each of us have a reference to those of the other and are performed upon the supposition that something is to be performed on the other part. . . . Nor is the rule concerning the stability of possession the less derived from human conventions that it arises gradually and acquires force by a slow progression and by our repeated experience of the inconveniences of transgressing it. On the contrary, this experience assures us still more that the sense of interest has become common to all our fellows, and gives us a confidence of the future regularity of their conduct: And ’tis only on the expectation of this, that our moderation and abstinence are founded. In like manner are languages gradually established by human conventions without any promise.” 2. Psychological activity requisite for language acquisition and transmission includes keeping one’s eyes on the speaker, auditorily monitoring what one is saying, sharing a common context, and engaging in what psychologist Michael Tomasello terms “joint attention”: see especially Constructing a Language: A Usage-Based Theory of Language Acquisition. Cambridge, MA: Harvard University Press, 2003, and “Understanding and sharing intentions: The origins of cultural cognition,” in Michael Tomasello, Malinda Carpenter, Joseph Call, Tanya Behne, and Henrike Moll, Behavioral and Brain Sciences (2005), pp. 675–735. Other relevant scientific writings on the origin and transmission of language include Simon Kirby, “Culture and biology in the origins of linguistic structure,” Psychon Bull Rev (2017) 24: 118–37; W. Tecumseh Fitch, The Evolution of Language (Approaches to the Evolution of Language). Cambridge: Cambridge University Press, 2010; Kevin Laland, Darwin’s Unfinished Symphony: How Culture Made the Human Mind, Princeton: Princeton University Press, 2017; Derek Bickerton, More Than Nature Needs: Language, Mind, and Evolution. Cambridge, MA: Harvard University Press, 2014; Daniel Everett, How Language Began: The Story of Humanity’s Greatest Invention, New York: Liveright, 2017; and Guy Deutscher, The Unfolding of Language: An Evolutionary Tour of Mankind’s Greatest Invention. New York: Metropolitan Books, 2005. 3. Contrary to the common assumption, Hume did not equate linguistic meaning with ideas in the mind. Like Locke before him, he left no doubt that socially established conventions are the prime determinants of linguistic meaning: see HTC chapter 3 and KEUU chapter 8. The respective roles of consciousness and convention in linguistic meaning come through with particular clarity in a letter he penned to a cousin, cited and discussed in KEUU chapter 4-C. 4. Philosophical Investigations, II.xi 223, translated by G.E.M. Anscombe, Oxford: Blackwell Publishers Ltd., 1953. 5. Philosophical Investigations, I.18. 6. The only exceptions are maximally simplified “languages” like the ones consisting entirely of building site orders that Wittgenstein invented expressly for purposes of contrast. 7. See note §2 earlier in the chapter. 8. See chapter 1 note §11. 9. “That impious maxim of the ancient philosophy, Ex nihilo, nihilo fit [from nothing nothing comes], by which the creation of matter was excluded ceases to be a maxim according to this philosophy. Not only the will of the Supreme Being may create matter, but, for aught we know a priori, the will of any other being might create it, or any other cause that the most whimsical imagination can assign. . . . The falling of a pebble may, for aught we know, extinguish the sun, or the wish of a man control the planets in their orbits. . . . Anything may produce anything. Creation, annihilation, motion, reason, volition; and

From Hume to Kant via Wittgenstein 145 these may arise from one another or from any other object we can imagine” (E 12.3 ¶32n & ¶29 and T 1.3.15 ¶1; see also T 1.4.5 ¶¶ 30 & 32). 10. The laws of physics may make separate existence impossible, but if a cosmology like that favored by physicist Lee Smolin is correct, these laws can change, thus permitting the separately conceivable to exist separately as well. By contrast, no change in the laws of nature can permit a mountain to exist without a valley or vice versa. 11. PFM 258. 12. CPrR 52. 13. Jim Baggott, for example, defends Aristotelian immanentism in physics: “The scientific interpretation acknowledges that information is not much different from other physical quantities. But, as such, it is a secondary quality. . . . ‘Information is physical’ means that information must be embodied in a physical system of some kind and processing information therefore has physical consequences. Take the physical system away, and there can be no information. . . . The metaphysical interpretation suggests that information exists independently of the physical system, that it is a primary quality, the ultimate manifestation of an independent reality. ‘Information is physical’ then acknowledges that in our empirical reality of observation and measurement, information comes dressed in a clothing of physical properties. This is a bit like suggesting that heat or temperature are the ultimate reality, existing independently but projected into our empirical world of experience in terms of the motions of physical objects” (Farewell to Reality, London: Constable & Robinson Ltd., 2013, p. 258). David Deutsch takes a similarly Aristotelian immanentist view of mathematics.

Part II

Nature in Mind Through Kant’s I’s

The brain does not simply take the raw data that it receives through the senses and reproduce it faithfully. Instead, each sensory system first analyzes and deconstructs, then restructures the raw, incoming information according to its own built-in connections and rules—shades of Immanuel Kant! —Eric Kandel, In Search of Memory

Introduction to Part II and Summary of Chapters 6–10 Humean skepticism as buttressed by Wittgensteinian conventionalism can be condensed into the thesis that any representational content that empirical psychologism cannot explicate, convention can, and what convention cannot explicate, nothing can. What Kant termed critical ­philosophy—the a priori variant of psychologism to be examined in the succeeding chapters—can be similarly condensed: any representational content that neither empirical psychologism nor conventionalism can explicate, a priori psychologism can, and what the latter cannot explicate, nothing can. Since this excludes any view like platonism that accords the meanings of at least some of the notions employed in discourse an independence of both sensibly conditioned consciousness and human sociality, whatever philosophers, mathematicians, or anyone else supposes to have such meaning can be nothing more than unintelligible nonsense masquerading as sense, any illusions to the contrary notwithstanding. What is true of psychologism generally is maximally so of a priori psychologism: readers must bracket out not only all knowledge relevant to the notions at issue but all prior conceptual understanding as well. This, as noted in the introduction, is because the task of psychologism is to explicate meanings, with special emphasis on identifying psychological ingredients essential to notions that, in language, are free of any tincture of psychological content. There can therefore be no expectation that the psychological contents adduced as essential to the meaning of familiar notions will themselves be familiar, especially in the case where those contents are a priori, and so accessible neither via sense experience nor

148  Nature in Mind: Through Kant’s I’s via mastery of language. Thus, however antithetical to proper procedure it may seem, psychologism is an exception to the rule that authors’ claims be assessed on the basis of one’s prior understanding of the concepts at issue. Kant’s two-stage psychologistic explication of the notion of space is a case in point. No one’s preexisting notion of space is likely to include anything remotely like pure sensibility, and the same is true of the objective space that emerges from the determination of pure sensible space conformably to pure concepts of the understanding via transcendental synthesis. This is because neither of these purely psychological a priori representations of space contributes anything capable of conferring the least mathematical, scientific, or ordinary empirical meaning on the notion. Instead, according to Kant, these psychologistically explicated spaces together constitute the transcendental space from which all these others take their start—the presupposed pre-linguistic, non-platonist a priori inner core of meaning on which all these others are overlaid. Without it, space becomes representationally impossible. With it, mathematical, scientific, and ordinary notions of space become practicable even if their actual representation requires more than a priori psychologism alone is capable of explicating. Any attempt to say anything at all about Kantian transcendental space by drawing on familiar notions of space is therefore a fool’s errand. Chapter 6: The Kantian Cogito There needs to be a first principle that demarcates a priori psychologism from everything else and systematically integrates its components, or it would be little better than random groping. In particular, its viability depends on there being a single a priori consciousness such that all and only those representations its existence implies or entails fall within the purview of a priori psychologism, while everything else, even if a priori, lies beyond its scope. The principle Kant identified as the only possible basis for a priori psychologism is the Cartesian ‘I think.’ Chapter 6 traces the career of this notion from the point in the Second of Descartes’s Meditations on First Philosophy where Kant’s views coincide with Descartes’s to the precise point they diverge. Their unanimity concerns Descartes’s determination that the representation ‘I’ has logical universality and so needs to be represented intellectually or not at all (i.e. there can be no sensing or imagining it). They diverged in how they understood this result. For Descartes, its purely intellectual character underwrites the metaphysical truth that I am an intelligible rather than a corporeal entity, i.e. a spirit miraculously united by God with a human body and not a physical being in my own right. For Kant, by contrast, its truth is purely a priori logical: the only logical universality ever present to consciousness is the one inherent in the representation ‘I think,’ so that everything else

Nature in Mind: Through Kant’s I’s 149 universal in it is so only because it incorporates pure intellectual self-­ consciousness into its representation, including the totality of logical structures universals enable us to form—propositions, inferences, narratives, et al. And since this is just to say that something purely psychological— the universal self-consciousness ‘I think’—is essential to the explication of something purely logical—universality (generality, commonality)—the former must be acknowledged as the latter’s psychologistic explication. Chapter 7: The Logical I The logical portion of Kant’s psychologism may start with logical universality, but that is by no means the end. Incorporating the ‘I think’ into sensations, images, associations, and other representations lacking universality straightaway transforms them into universals capable of being combined to form logically structured representations, thus conferring on them the status of concepts in a genuinely logical sense. Yet, the ‘I think’ by itself does not suffice for the actual logical combination of concepts in propositions. It requires supplementation by innate logical forms of judgment capable of synthesizing any ‘I think’–generated concept with any other in a proposition, thus making possible a single a priori consciousness in which all possible such concepts are united—not literally, to be sure, but as a merely synthetic unity. Since whatever is included in this purely logical synthetic a priori unity of consciousness ipso facto falls within the scope of the ‘I think,’ logical forms thus extend the scope of universal self-consciousness to include not just all possible non-conceptual representations but all possible conceptual and propositional representations as well. Of all the representations psychologistically explicable through the logical ‘I,’ by far the most important in Kant’s philosophy are the pure concepts of the understanding that result by representing restrictions on the logical freedom inherent in forms of judgment. Though their only positive contents are the logical forms they restrict, the restrictions represented in pure concepts add a genuinely extra-logical necessity to universal self-consciousness—one that carries over to the synthetic a priori unity yielded by logical forms to convert it into a necessary synthetic a priori unity of consciousness (the first plank in Kant’s refutation of Humean skepticism). Of course, it is a long way from a purely logically defined necessity that relates merely to the placement of concepts in propositions and of simpler propositions in more complex ones to the genuinely objective necessities instanced by laws of nature, spatial and temporal relations, and the abstract necessities of mathematics. This is true particularly in the case of the necessary relations represented in the notions Kant used pure concepts of the understanding to explicate: the traditional categories of metaphysics. Understanding “categories” to be concepts of objects so fundamental that all others presuppose them while

150  Nature in Mind: Through Kant’s I’s they presuppose none, it seems impossible for categories to be explicated by manifestly non-objective representations of restrictions on the logical freedom of concepts and propositions. How can two such radically heterogeneous kinds of concept be so related that pure concepts of the understanding, despite being purely logical in content, suffice to psychologistically explicate metaphysical categories? Of all the problems Kant had to tackle on the way to writing the Critique of Pure Reason (some of which took the better part of a decade to resolve), this heterogeneity problem is undoubtedly the most formidable. And of all the challenges confronting his reader, none is more daunting than the solution recounted in the remaining chapters of Part II. Chapter 8: The Aesthetic I Recognizing that the logical I  does not suffice for the solution of the heterogeneity problem, Kant expanded his search to the consciousness on which the possibility of the ‘I think’ itself depends. Obviously, if the universal self-consciousness ‘I think’ is the a priori condition for logically structured thought, this condition’s own a priori condition cannot involve anything logical in its representation but must instead be purely sensible. Kant therefore devised a theory of pure sensibility—termed transcendental aesthetic—which has as its raison d’être the psychologistic explication of the a priori sensible consciousness that precedes the ‘I think’ and first makes it possible. The theory is many-sided and complex. Suffice it to say that comprehending it depends on scrupulously and fully carrying out the two abstractions that Kant required, first from sensation (particularly tactual and visual), and second from the understanding. What makes carrying out the second especially tricky is that pure understanding comes up for discussion in the Critique of Pure Reason only after transcendental aesthetic has ostensibly been dealt with. One must consequently return from the transcendental analytic of pure understanding to transcendental aesthetic in order to abstract everything that presupposes pure understanding from the purely sensible consciousness that pure understanding itself presupposes. Since Kant attributed objectivity to pure understanding and claimed that all non-transcendental objectivity—ordinary and scientific as well as mathematical—would be impossible without it, abstraction from pure understanding requires that all these be bracketed out of one’s notion of this purely sensible consciousness. The upshot is that transcendental aesthetic has to be understood purely as subjective psychology— nothing else, nothing more. As such, Kant’s theory of pure sensibility is best approached as a new solution to the problem that led Berkeley to devise his theory of vision: sensible spatiality. Instead of following the purely empirical path Berkeley trod, Kant blazed an a priori psychologistic trail leading, on the one

Nature in Mind: Through Kant’s I’s 151 hand, to the pure sensible consciousness presupposed by the ‘I think’ and, on the other, to strikingly new ways of thinking about the kinds of empirical psychology that humans share with other animals. The theory stakes out a purely psychological but nevertheless a priori middle ground between sensation and pure understanding occupied by a priori imagination. A priori imagination differs from the merely reproductive empirical variety by being not only non-intellectual but genuinely productive as well—i.e. it originates purely aesthetic representations not found in sensation yet in no way involving the ‘I think.’ In beings constituted like us, these are pure space and pure time. Thanks to the former, external senses like vision and touch are fused together to form a single external sense through which one and the same sense-divide transcending spatial appearance can be seen, touched, heard, etc. Thanks to the latter, our multifarious internal sensations are united in a single internal sense through which one and the same internal sense-divide transcending temporal appearance may be apprehended. Chapter 9: The Objective I Kant’s a priori psychologism cannot come properly into view unless one is careful to distinguish all of the following: purely aesthetic, merely subjective pure space and time from transcendentally objective space and time from mathematically objective (especially geometrical) space and time from empirically objective space and time (i.e. those of ordinary and scientific experience) from purely aesthetic, merely subjective empirical space and time. Having discussed the first in the previous chapter, Chapter  9 focuses on the second since it too falls under the aegis of a priori psychologism. Here, a major part of the task is simply to get clear what distinguishes it from the other species of objective space and time, and why Kant regarded it as their presupposition. Once this is done, the question devolves into the straightforwardly psychological query: how does the mind produce transcendentally objective space and time? The ingredients Kant required to answer it have already been adduced: forms of judgment/pure concepts of the understanding (from Chapter 7) + the purely aesthetic space and time of sensibility presupposed by the ‘I think’ (from Chapter 8). His challenge was to explain how two such radically heterogeneous kinds of conscious representation—the one 100% logical and 0% sensible, the other 100% sensible and 0% logical—can be combined at all, much less synthesized to yield objective transcendental space and time. The key to overcoming it is the capacity of the ‘I think’ to transform any sensible representation from a non-concept into a concept simply by being incorporated into it. For although pure concepts of the understanding apply only to concepts and propositions, not to sensible appearances, insofar as the contentless consciousness ‘I think’ is all that differentiates the two in the case of concepts and propositions

152  Nature in Mind: Through Kant’s I’s derived from the manifold of pure space and time, the determination of the former by pure concepts cannot fail to carry over to the appearances themselves. Thus, it is solely by means of this transcendental synthesis, as Kant termed it, that the appearances comprised within pure space and time are objectified. To help make transcendental synthesis clearer, I analogize aesthetic and transcendental space with the Facebook-like network that results when graph theory is used to illustrate what Einstein called “spooky action a distance.” Utilizing a discussion of the topic in a book on quantum theory for the general reader, I show that pure concepts of the understanding objectively differentiate and determine space very much in the way a network of otherwise disconnected nodes can be used to graph spatial relations. The upshot is that the role of pure concepts of the understanding in transcendental synthesis can be most easily and concisely understood as close transcendental kin to a graph theory. Chapter 10: The I of Nature Transcendental synthesis represents a partial solution to the heterogeneity problem of Chapter 7. Its complete solution, however, required Kant to show how transcendental objectivity can combine with physical reality in such a way that pure concepts of the understanding can finally merit his designation of them as categories—and, in the process, psychologistically explicate nature itself. The challenge he confronted was two-sided. First, if objectivity is possible only via productive imagination (pure intuition) and pure understanding (the ‘I think,’ logical forms), it is both pure and non-sensational. So, how can it have anything to do with the physical realities indicated pre-imaginatively and pre-intellectually by sensations? Second, and conversely, sensations present themselves in consciousness prior to and independently of both productive imagination and pure understanding. So, if objectivity has everything to do with productive imagination and pure understanding and nothing to do with sensations, how can the realities indicated by the latter possibly have the least objectivity? Insofar as the one seems as impossible to objectify as the other does to realize outside of pure imagination and pure thought, transcendental objectivity is in danger of being reduced to a mere fiction (“a phantom of the brain”), while the realities indicated by sensations risk collapsing into complete subjectivity (“less than a dream”). Kant’s solution again involves two purely psychological ingredients: a constituent of transcendental aesthetic so far left unmentioned that he termed the matter of appearances + the determination of that matter conformably to the pure concepts of the understanding fortified with spatial and temporal meaning by transcendental synthesis in a further synthesis he termed transcendental schematism. The resulting combination of objective transcendental space and time with the physical realities

Nature in Mind: Through Kant’s I’s 153 indicated by sensations gives physical reality to the former while conferring objectivity on the latter. At the same time, it serves as Kant’s psychologistic explication of nature itself. The explication cannot be equated with idealism, however, since it leaves the existence of physical things completely independent of consciousness, even if, by Kant’s psychologistic reckoning, everything about them—their spatiality, temporality, substantiality, causality, mathematical quantifiability, universal subordination to natural laws, and laws of nature themselves—is not. This means that in the absence of consciousness, physical existence remains in the form of what Kant termed things in themselves. Unrepresentable things in themselves are the inevitable result of a psychologistic explication of nature that, apart from existence, quite literally swallows up everything into the mind (that being the point of Part II’s title). After addressing a number of topics relating to mathematical physics, the chapter concludes by laying out, step by step, Kant’s refutation of the Humean skepticism discussed in Chapter 5.

6 The Kantian Cogito

There needs to be a first principle that both demarcates a priori psychologism from everything else, including the rest of a priori cognition, and systematically integrates its components, or it would be little better than random groping.1 The principle Kant identified as the only possible basis for a priori psychologism is the Cartesian ‘I think.’ The chapter traces the career of this notion from the point in the Second of Descartes’s Meditations on First Philosophy where Kant’s views coincide with Descartes’s to the precise point they diverge. The focus of the latter is Kant’s determination that the certainty expressed by cogito ergo sum has neither metaphysical nor epistemological but solely logical significance. The conditions that make the logical self-consciousness ‘I think’ possible as well as the representations made possible by it thus constitute the sole and entire subject matter of his a priori psychologism.

Kant’s Debt to Descartes Descartes’s Meditations on First Philosophy (1641) recount a thinker’s journey in the first person. Anyone can think along, no preparation necessary. The stage of the journey that mattered most to Kant starts in the Second Meditation when Descartes picks up a chunk of beeswax fresh from the hive and considers the nature of his representation of it. There are various sensible qualities—its look, feel, weight, smell, taste, the sound it makes when he strikes it. But he then recalls that when wax is melted over a fire, these features all change without the representation ceasing to be of the same wax it was originally. Since he can imagine doing the same to the beeswax he is currently holding, he can regard none of the features he senses in it as essential to his representation of it. Is changeability essential to the representation? Undoubtedly. But is its changeability defined by the changes he can imagine? Unlike the senses and memory, imagination is not confined to how the wax actually appears now or has appeared in the past. Nevertheless, in representing the wax changing in ways not previously experienced, the imagination is completely dependent on materials provided by the senses and retained

156  Nature in Mind: Through Kant’s I’s in memory. For example, even if I have never seen boiling wax evaporate, I  can imagine the beeswax evaporating when heated to a temperature hotter than any I have previously witnessed by drawing on my experience of what water does when it is heated past its boiling point. But this is merely imagining together things I have previously experienced separately, and so leaves me just as dependent on materials provided by sense and memory. Where such materials are lacking, there imagination falters. The changeability I  accord to the wax in my representation of it is not, however, restricted to the changes my experience presently equips me to imagine. Instead, it is open-ended and unlimited, constrained by neither past experience nor my limited sensory capacity to detect changes (e.g. ultraviolet changes in the wax’s appearance are invisible to me but not to all creatures). Indeed, even if I had at my disposal experience of everything that happens to the wax’s appearance under any and all conditions, through all possible ways of sensing wax (not just those specific to humans), imagination still would not be enough. For the representation of its changeability is so open-ended in scope that in even the most ignorant mind, a place is kept open for changes of a kind that might not even be sensible or imaginable at all, such as those that turn out not to be representable other than through the equations of quantum mechanics. Thus did Descartes conclude that the changeability essential to the representation not only of the beeswax in his hand but of particular material objects of every kind must have some source other than the senses and imagination. If not the senses or the imagination, then what is the source of the representation? The key to answering this, according to Descartes, is to recognize that representing the wax as able to remain the same through a spectrum of limitless change requires nothing more than representing the wax as what is common to the appearances one beholds, not only before and after melting but before and after any change whatsoever, whether sensible/imaginable or not. A representation that is common to things that otherwise differ, including things that differ in every sensible or imaginable respect, is a general representation, a universal. Today we know that the common factor in all changes beeswax may undergo is a particular molecular composition: as soon as that is changed, the common factor is absent, and it is no longer wax, no matter how wax-like it might appear to the senses. But even in earlier ages and among the scientifically illiterate in ours, it is understood, and included in the representation of the wax, that there is something, some “stuff,” common to even the most disparate appearances that makes them all individually the same “substance,” with the same uniquely waxen “nature,” insensibly differentiating it even from non-waxen things sensibly indistinguishable from it. Insofar as the general representation of wax is compatible with wax changing in ways we are not equipped to imagine, imagination-based would-be universals like Hume’s customary resemblance associations

The Kantian Cogito 157 (Chapter  4) cannot but fail to explain it. Descartes therefore looked beyond sense and imagination to a third, completely autonomous faculty of understanding to account for the universality of the representation of the wax and other ordinary material things. To be sure, many philosophers contested Descartes’s attribution of universals to the understanding. Hume, in particular, had another way to explain representations that outstrip the capacity of associative imagination: they have their source in language, not the psyche. We learn the words for wax and other common, everyday objects so early that we have no memory of doing so. Since the conventionally established meaning and grammar of such words includes precisely the kind of open-ended application that encompasses limitless changeability and commonness to things otherwise completely different, the fact that we master them so young can easily mislead us into supposing that a special psychology is required to produce universals that are in no way beholden to the social infrastructure and specifically human lifeways on which vernacular language depends. Thus, Descartes’s analysis of the wax runs straight into the Humean skeptical challenge discussed in Chapter 5: first, the demand to demonstrate the psychological reality of the association-transcending universality he attributed to the representation of the wax such that it could still be present in consciousness even if every trace of vernacular language and its social underpinning were erased from the mind; and second, the demand to show why, even if such a consciousness were possible, non-associative, convention-independent universals are needed, indeed indispensable, in order to be able to represent objects such as the wax at all. Descartes did not foresee Hume’s skeptical challenge, and it is difficult to see how he could have surmounted it if he had. To appreciate why Kant’s case is different, we first need to examine the implication Descartes drew from his analysis of the wax: “every consideration whatsoever that contributes to my perception of the wax, or of any other body, cannot but establish even more effectively the nature of my own mind.”2 This relates to the sequel to his famous cogito ergo sum (“I think therefore I am”): having determined that “I am, I exist, is necessarily true whenever it is put forward by me or conceived in my mind,” Descartes set out to understand “what this ‘I’ is.”3 The point of the analysis of the representation of the wax was to show that just as representations of particular material objects require a generality beyond the scope of imagination, so too does the representation of the ‘I.’ How so? To represent the wax as the occasion of a limitless number and variety of actual and possible representations is ipso facto to represent any of them as the representation I might now be having. To think this of myself, however, presupposes a representation of the ‘I’ far beyond anything that sense or imagination is capable of supplying. It requires me to think that, instead of the representations currently present in my consciousness, any of infinitely many

158  Nature in Mind: Through Kant’s I’s others might have been there in their stead and yet made no difference to my being able to think “I am, I exist” at this moment. In other words, in representing this ‘I,’ I represent numerically the same consciousness as able to accompany an infinite number and variety of possible representations, including not only those I have, will, or could ever sense and imagine, but also those that human senses and imagination do not permit me to represent at all, as well as those representations that are intrinsically insensible and unimaginable yet still thinkable, even if only with the aid of language (group theoretical constructs, string theory equations, etc.). Or, what is the same, I represent this ‘I’ as common to all the representations I could ever have, everything I could possibly think (or say), regardless of whether past experience, culture, or human mental faculties equip me actually to do so. Thus, in contrast to the representation of the wax, which excludes from the changeability that helps define it all changes that change wax into non-wax, the representation ‘I think’ is so maximally general as to exclude nothing representable whatsoever from its scope. Could Hume challenge this claim too for surreptitiously invoking linguistic convention? If consciousness is real independently of convention, as few dispute, then there is no reason to think self-consciousness cannot exist without it too. Hume certainly did, both in the form of the sustained affective disposition at the heart of his associationist account of personal identity and in the form of the self-same enduring consciousness in which successive perceptions are united that the self of association presupposes (Chapter 4). The question is whether the representation ‘I think’ genuinely constitutes a third species of psychologically real self-consciousness or instead is impossible without linguistic convention to underpin it. Descartes’s proof that the ‘I think’ is a case of genuine self-consciousness runs something like this. (1) The I that thinks, whether in language or purely psychologically, is ipso facto conscious (as the non-existent cannot be conscious, any notion that the I misapprehends itself as conscious when in fact it is non-existent is impossible and absurd in a sense stronger than any merely logical contradiction). (2) This apprehension would be unaltered if the thought it apprehends itself thinking were of something else; and since the thought could have been of absolutely anything else, be it true, false, necessary, impossible, empty, self-contradictory, or even nonsensical,4 the representation of the I is a true universal: able to accompany, and therefore common to, every other representation of which minds are capable, non-linguistic no less than linguistic. (3) Since this is just to say that no content is represented in the ‘I think’ at all, just a pure consciousness able to accompany/common to all representable contents, the ‘I exist’ cannot be supposed to be in any way dependent on language. For if the representation, in and of itself, has no content, then it cannot be suspected of having borrowed any, whether from language or anywhere else (e.g. the content ‘first-person pronoun’). (4) Thus, in contrast to the general notion of wax, there can be no doubt that the ‘I think’ is a purely

The Kantian Cogito 159 psychological representation wholly unbeholden to language, and, more particularly, a genuinely logically universal self-consciousness.5

Kant’s Divergence From Descartes Up to this point, Kant was in full agreement with Descartes. That ceased, however, when Descartes proceeded in the Third Meditation to contend that the representation ‘I exist’ implicitly (i.e. darkly) incorporates additional representational content without which the representation of its existence would not be possible. These include not only the representation of the I as a substance that, as conscious, has to be immaterial, but also the I’s finitude; and, on the ground that it is impossible to represent finite substance without also being able to represent infinite substance, Descartes further maintained that the representation of God is stamped on every I-substance in something like the way manufacturers stamp their logo on their products. The first problem with this reasoning is that, unlike pure self-consciousness, the notions ‘substance,’ ‘immaterial,’ ‘finite,’ ‘infinite,’ and ‘God’ are as vulnerable as ‘wax’ to the Humean challenge to prove they have psychological reality and are not inherently dependent on linguistic convention. As noted earlier, Descartes was just the sort of platonistic innatist at whom criticisms like Hume’s were specifically targeted. Whereas Kant acknowledged these criticisms and developed an alternative to platonism capable of establishing the psychological reality of ‘substance’ and other notions, neither Descartes nor, in my judgment, any subsequent platonist has ever devised an adequate response to the Humean challenge, especially as elaborated by the later Wittgenstein. The second problem Kant had with Descartes’s reasoning stems from the fact that no representational content, substantial or otherwise, is apprehended in the representation of the ‘I,’ be it in the ‘I think’ or the ‘I exist.’6 The ‘I’ is a universal insofar as it is consciousness common to all other representations, possible no less than actual, not a common content. Indeed, its very contentlessness is essential to its constituting a genuinely universal self-consciousness.7 By contrast, the representation of the wax as something changeable in more ways than we could ever imagine includes something unchangeable in its content that is common to all wax but absent from everything else. It may be represented as precisely as a chemical formula or as vaguely as the naive representation of it as some uniquely waxen “nature,” “stuff,” or “substance.” Yet, however imprecise one’s representation of the wax, it essentially includes either some distinguishing content over and above consciousness itself or, at the very least, a placeholder for such content. And the same is true of even the most general representations of substance, immateriality, and finitude, but truer still of the representation of an individual finite immaterial substance: the former have some actual, or at least presumed, content

160  Nature in Mind: Through Kant’s I’s adequate to distinguish each from the others and from all other general representations, while the latter must be thought of as so rich in content as to distinguish itself from all other individual things, possible no less than actual, howsoever similar otherwise. None of this is true of the ‘I.’ The fact that I might be thinking any representation instead of the one I think (cogito) now and still apprehend the same existence (sum) implies that no content I could ever think is, or can be, part of the apprehension of my (the I’s) existence—‘finite,’ ‘individual,’ ‘immaterial,’ and ‘substance’ included. For Kant, being a consciousness common to, and so able to accompany, every representation I could ever think, regardless of its content and however different it is from any other such content, is the very essence of the representation ‘I think.’ If any representations were possible to which the ‘I think’ was not common, they would ipso facto not be representations that I could ever think, and so would necessarily be nothing to me.8 Its genuinely unrestricted universality vis à vis all consciously representable contents thus implies that the ‘I’ representation itself cannot include even so much as a placeholder for a representable I-“nature” (I-“stuff,” I-“substance”), since that would prevent it from being common to the representations of both, e.g.,  beeswax and non-beeswax (i.e. everything that lacks the nature of beeswax, from nutmeg to the Pythagorean theorem). The ‘I’ is, and is only able to be, unrestrictedly universal precisely because it has no content of its own to contribute to conscious representation. The representation ‘I think’ is the otherwise completely empty thought of a consciousness, not a content, as common to all possible representations, and so is as much an informational blank slate representationally as it is vis à vis its neural correlate. The sole and entire representational contribution of the ‘I think’ is the purely logical generality (commonality, universality) this consciousness confers on otherwise non-general representations.9 And, pace Descartes, logical generality by itself is not enough to think anything in the least differentiated, implicitly (darkly) or otherwise, including even the most abstract general representations of substance, immateriality, and finitude, much less the apprehension of an existent individual finite immaterial substance. One of the contents absent from the representation ‘I think’ that is particularly important for present purposes is temporality. Temporality is basic to the two conceptions of self-consciousness we have encountered so far: self-consciousness as the pre-associative enduring unity of successive perceptions and self-consciousness as Hume supposed it to arise through associative imagination. In the case of Kant’s logically universal self-consciousness, by contrast, even if (unlike most physicists) we take the existence of consciousness as an unfolding succession of perceptions as given, time is no more part of the representation ‘I think’ than space is. In the spirit of Cartesian skepticism, one can even suppose that the succession of internal sense is entirely fictitious and that the conscious

The Kantian Cogito 161 mind is in reality completely atemporal without its making the least difference to the apprehension of the indubitable existence of the I. Indeed, not just time and space but circumstances and conditions of existence generally are not included, even darkly, in the purely logical representation ‘I think’ (cogito), with the consequence that the representation ‘I exist’ (sum) neither implies nor entails any of the following extralogical contents: that I am an immaterial substance capable of existing through time as Descartes supposed; that I  am a Humean associative fiction; that I endure even from one instant to the next; that I am a metaphysically puzzling “extra” that comes with having a functional human brain; that I am newly created at this instant along with newly created memories of a past that never happened as in Boltzmann-inspired science fiction; or that I have some other, as yet unthought of mode of existence. The only thing that cannot be absent from the representations ‘I think’ and ‘I exist,’ according to Kant, is logical universality: the contentless representation of a consciousness common to absolutely all contents representable as mine. One of the chief tenets of Kantian egology is the thesis that while logically universal self-consciousness does not entail any other form of selfconsciousness, all others entail it. For example, in the absence of any consciousness that could be represented as common to all representations, could Hume’s derivation of the self from a sustained affective disposition in associative imagination still be regarded as self-­consciousness? In the absence of the universal self-consciousness ‘I think,’ it is hard to see how any sustained affective disposition could qualify as a genuine instance of self-consciousness, be it sustained ringing in the ears, sustained backache, sustained anxiety, or the feeling pinpointed by Hume (which is not to deny that it may be the closest I-less creatures can come to true self-consciousness). So too internal sense of the goings on in our own minds: consciousness of the manifold present in one at any given instant as a manifold cannot be deemed an instance of self-­consciousness if universal self-consciousness is absent, for there would then be nothing to make the consciousness of the manifold at the next instant representable as numerically identical with the consciousness at the preceding. Similarly, a human-like creature or android that lacked universal self-­ consciousness but could be taught to use the English first-person pronoun ‘I’ and the verbs ‘think’ and ‘exist’ sufficiently well to pass an English speaker’s Turing test for Cartesian self-consciousness would not thereby qualify as genuinely self-conscious. To suppose otherwise would be to confound linguistic representations of the self with evidence of psychological reality, something that conventional stipulations alone are never capable of being (Chapter 5). Nor can the I-less be credited with moral self-­consciousness, social self-consciousness, or even physical self-­ consciousness. For where the ‘I think’ is absent, how can merely having volitions to act, social relations, or a body suffice for any notion of self

162  Nature in Mind: Through Kant’s I’s that extends to all representations, possible no less than actual? So far as Kant was concerned, not only is the representation ‘I think’ all that is needed to demonstrate the psychological reality of self-consciousness, it is the universal representation of self presupposed by all particular ones—­temporal, linguistic, moral, social, and bodily alike.

A Priori Psychologism and Pure Understanding The ‘I,’ for Kant, is the contentless representation of numerically the same consciousness as is able to accompany, and therefore common to, absolutely all representations that can ever be anything to me. Common to the totality of my consciousness, possible no less than actual; nothing omitted; complete representational universality. Again, this is only because it is merely a consciousness that is common. Although the representation ‘I think’ can supply any content given for thought with the added feature of logical universality (commonality, generality), it can never supply the content requisite to form a concept of itself for the simple reason it has none to give (“the simple and, in its own right, completely empty representation I”).10 Since, in the absence of any content for thought, no object can be represented, the ‘I think’ can never be represented as an object. This is not true in the case of everything else that tends to be regarded as subjective: I think my feelings, emotions, desires, and volitions, just as I (can) think every other (possible) representation of mine, making them objects in at least the attenuated sense in which “[a]ll representations, as representations, have their object, and can themselves in turn be objects of other representations”11 (among which one may include Humean facility and vivacity affects as well as the qualia of contemporary philosophers and scientists of mind). The natureless yet existent I  of the cogito, uniquely among representations, is radically subjective, whereas everything that has a representable nature (including the various guises of the empirical self) is objective, if only in this attenuated sense. Indeed, not only can the I never be represented as an object in any sense, we can never even get so far as trying: “since we must always avail ourselves of its representation in order to [think] anything about it,” we “can only revolve in a perpetual circle” if we attempt to represent this “bare consciousness . . . in isolation.”12 So far as the ‘I exist’ (sum) portion of the Kantian cogito is concerned, far from being “a representation that distinguishes a particular object,” it “represents nothing further than an unknowable, unrepresentable transcendental subject = X.”13 The ‘X’ in the formula, as noted previously, is not a variable or even a placeholder, since it is impossible ever to set a representable content in its place. But because the representation ‘I think’ is immediately the representation ‘I exist’ as well,14 the I is a subject not merely in some abstract sense, much less a merely grammatical one, but in full, unmitigated psychological reality. Pace Descartes, were we to

The Kantian Cogito 163 attempt to represent it, and thereby understand what the subject is, we would already have represented the I of the ‘I think’ in order to do so, and so merely revolved in a circle. If this were not the case and we could somehow represent the I itself, then the place indicated by the ‘X’ could be filled in, and the I would be a representable object, and so a content of consciousness like any other. But because the ‘I think’ is as much an informational blank slate representationally as it is with regard to its neural correlate, the ‘I exist’ can only ever be affirmed as an unrepresentable something = X, which is, in truth, no different from a question mark. It is in this sense that the ‘I think’ is subjective in the most radical sense: not a what of any kind or nature at all, but simply pure contentless consciousness. It is therefore only in a secondary, relative, purely epistemological sense of ‘subjective’—that of lacking ultimate cognitive worth—that sensations, emotions, desires, volitions, qualia, and other (attenuated) objects of representation that have no place in a final mathematical physics accounting of reality can properly be ranked as subjective. Kant designated the ‘I think’ a transcendental self-consciousness primarily on the ground that it is not an empirical representation. He maintained this, it is true, despite acknowledging that the proposition “I think therefore I am” is empirical. But to say that it is “empirical” simply means that “sensation . . . underlies the existential proposition ‘I exist,’ ” since without sensation “to supply the material for thought, the act ‘I think’ would not take place”15—something that is true of all the representations Kant regarded as a priori (e.g. pure space and time).16 What nevertheless makes this representation a priori rather than empirical is that the apprehension of the existence of the thinking subject in it is “an indeterminate empirical intuition, i.e. perception,” which “here means only something real that is given, given indeed only to thought in general . . . as something that in fact exists and in the proposition ‘I think’ is designated as such an existent.”17 It is the total lack of specificity and determinateness (“thought in general”) of the perception that here is crucial: since its contentlessness a fortiori excludes sensational content, the ‘I think’ does not single out any sensation(s) from the rest, actual or possible, as indicative of the I’s existence rather than something else’s—­something possible only a posteriori—in the way that one olfactory sensation is a posteriori indicative of smoke and another of carrion but none is indicative of the sound of a trumpet. Any sensation whatsoever—visual, tactual, olfactory, or any other (including those not in the human repertory)—suffices for the representation ‘I think’ to take place. Thus, on the ground that it makes sense to regard a representation as a posteriori, or empirical, only if forming it depends on consciousness of some particular sensation or collection of sensations in distinction from others, Kant did not hesitate to designate the ‘I think’ a priori self-consciousness, or, more usually, pure apperception.18 And what makes the ‘I think’ not just pure but transcendental apperception is that its a priori universality can be

164  Nature in Mind: Through Kant’s I’s conferred on representations that otherwise lack it (Chapter 7), including prototypically objective concepts like the transcendental categories of traditional metaphysics (Chapters 9–10).19 The purely intellectual character of universal self-consciousness should already be evident: unrestricted universality is the hallmark of conscious representation beyond the scope of the senses and imagination, and the ‘I think’ is universal in just this sense. Indeed, Kant singled it out precisely because it is the only intrinsically universal representation ever present in consciousness. The rest are all mélanges of sensations, memories, images, passions, emotions, desires, and volitions that never come closer to generality in their own right than I-less customary resemblance association can bring them (Chapter 4). Of course, if genuine universality goes beyond the senses and imagination, and the ‘I think’ is the only genuine universal in consciousness, then it should come as no surprise that Kant saw fit to define understanding expressly in terms of pure apperception, and, on that basis, designated the faculty responsible for it pure understanding and the science concerned to investigate it transcendental logic. In so doing, he became the first philosopher to define the understanding not in terms of the logical operations it performs—forming concepts to employ in propositions that can then be strung together for drawing inferences (reasoning)—but purely in terms of psychology, albeit a priori psychology. This is as ground-breaking as anything to be found in Kant’s philosophy. Instead of placing logic ahead of psychology, thereby precluding its psychologization, he set psychology ahead of logic, thereby effectively absorbing the latter into the former. Thus does apperception become “the highest point to which all employment of the understanding, even the whole of logic, and in accordance with it, transcendental philosophy, must be affixed.”20 In a word, Kant psychologized logic. The a priori psychologism that emerges from the foregoing considerations is a discipline whose one and only datum is the representation ‘I think.’ Everything in it converges upon or diverges from it, and the certainty as to what its components are and how they fit together derives entirely from the certainty of “I think therefore I am.” Conversely, what this datum fails to explain or justify affirming, a priori psychologism cannot explain or justly affirm, even if a priori (e.g. mathematical representations require mathematical definition and proof, not a priori psychological evidence).21 A priori psychologism thus defines a precisely delimited space for transcendental cognition that must never be transgressed. Finally a terminological remark. Kant did not use the expression ‘a priori psychology’ as I do in this book. Instead, he followed the standard philosophical practice in his day by using it to refer to the metaphysical psychology exemplified by Cartesian dualism. Had Kant applied it to his own transcendental philosophy he risked creating more confusion than clarity. Yet, just as the term ‘empiricism’ is a perfectly appropriate label to attach to Hume’s philosophy even if Hume himself seems never to

The Kantian Cogito 165 have used it, ‘a priori psychology,’ and more particularly ‘a priori psychologism,’ optimally captures everything in Kant’s philosophy that falls under the aegis of its supreme principle, pure apperception. Indeed, for us today, these terms are likely to be less confusing and more clarifying than the now archaic ‘transcendental philosophy’ (though I will occasionally use ‘transcendental psychologism’ as a synonym for ‘a priori psychologism’ to emphasize the cognitive dimension of the latter).

Notes 1. Kant insisted that an a priori philosophy like his must form a system in which (1) nothing essential is omitted or inessential included, (2) every link between components carries demonstrative certainty, and, as in an organism, (3) the parts are as inseparable from the whole as the whole is from its parts. See KAIM 258–9 and 380–2. 2. Meditations on First Philosophy, Second Meditation AT VII 33. 3. Second Meditation AT VII 25. 4. E.g. the thought that 2 + 2 = 4 is necessarily true; it is impossible for the thought that 2 + 2 = 56 to be true; any thought regarding trisected angles is empty because such things are impossible; the thought of a circular square is self-contradictory; and the thought that the clod of earth on my shovel is having a nightmare is nonsensical. 5. Kant utilized the expression ‘universal self-consciousness’ at CPR B132; see also A398 and AA 18 §5927. 6. “[T]his representation contains no content, and so no manifold” (CPR A381). 7. [O]ne cannot even say that it is a [general] concept but a mere consciousness that accompanies all concepts” (CPR A345/B404). 8. Or expressed in Kant’s inimitable manner: “The I think must be able to accompany all my representations, for otherwise something would be represented in me that could not be thought at all, which is as much as to say the representation would either be impossible or at least be nothing for me” (CPR B131–2). 9. The ‘I think’ is a “consciousness which makes [a representation] into a conceptus communis . . . [and] attaches to all common concepts as such” (CPR B133–4n). See also note §19 later in the chapter. 10. CPR A345/B404. Also: “The proposition I am simple  .  .  . means nothing more than that this representation, I, does not comprehend the least manifoldness within it and that it is an absolute (although merely logical) unity” (A354; similarly at B135 and A340/B398). The same is true of the sum portion of cogito ergo sum: “an understanding [in] whose pure apperception in the representation I am nothing manifold at all is given” (B138). The notions of the I mentioned earlier all involve a sensational, temporal, linguistic, moral, social, and/or physical manifold, and so have “natures” just as Descartes’s wax has its “waxen nature.” But there can be no nature proper to the simple, non-manifold I: it has none but logical unity, i.e. that of logically universal self-consciousness. 11. CPR A108. 12. CPR A345/B404. 13. CPR B146. 14. “[T]he supposed Cartesian inference, cogito ergo sum, is in fact tautological, since the cogito (sum cogitans) asserts actuality immediately” (CPR A354).

166  Nature in Mind: Through Kant’s I’s 15. CPR B422–3n. Kant was no panpsychist; consciousness, for him, must have originated in something like Locke’s oyster, distinguished from perfect insensibility, but descends no further. 16. See, for example, ID 406, Discovery 222, A452/B480n, AA 17 §4634, AA 18 §5049. A discussion can be found in KAIM chapter 3-A. 17. CPR B422–3n. 18. “For this inner perception is nothing further than the mere apperception I think  .  .  . [I]nner experience in general and its possibility, or perception in general and its relation to another perception, apart from any empirically given specifying distinction and determination among them, cannot be regarded as empirical cognition but must instead be regarded as cognition of the empirical in general and belongs to the investigation of the possibility of each experience which is certainly a transcendental investigation. The least object of perception (even, for example, pleasure or displeasure) added to the universal representation of self-consciousness would immediately transform [a priori] psychology into empirical” (CPR A343/B401). 19. Apperception “is the vehicle of all concepts in general, and so of transcendental concepts as well, and thus is always conceived along with these latter, and therefore is just as transcendental as they are, but can have no specific title because it serves only to present all things as belonging to consciousness” (A341/B399–400), so that “the mere apperception I think . . . makes possible even all transcendental concepts, in which it means that I think substance, cause, etc.” (A342–3/B400–1; also A348/B406). 20. CPR B133–4n. Also B131. The subordination of logic to psychology (pure self-consciousness) is almost universally ignored by commentators, but the fact that Kant affirmed it is undeniable, and its implications for how his philosophy needs to be understood cannot be exaggerated. Not least, it sets apperception at the ground of the categories so that the latter presuppose the former, not vice versa as generally supposed. See chapter 8 note §3. 21. In addition to mathematics, a priori cognition that lies beyond the kind for which the principle of Kant’s a priori psychologism suffices includes what he termed ‘metaphysics of nature,’ which may be thought of as a blend of a priori psychologism and Newton (a metaphysics of nature written in our day would combine a priori psychologism with relativity and quantum theory, just as one written in classical Greece or the Middle Ages would most likely mix it with Aristotelian physics). Interested readers should consult the discussion of a priori cognition in the “discipline of pure reason” in the eponymous section of the CPR Methodology; also KAIM 31-2, 394n7, 539, and 560-1.

7 The Logical I

Kant’s purely logical reinterpretation of the Cartesian cogito surmounts the first part of the Humean skeptical challenge: demonstrating the psychological reality of the “universal self-consciousness”1 ‘I think,’ and, with it, the psychological reality of a faculty of understanding (including pure understanding) over and above the senses and imagination. However, there is a second part to the challenge: showing that and how the ‘I think’ is needed for, indeed indispensable to, intelligent consciousness/conscious intelligence. To that end, the chapter starts by examining Kant’s psychologistic explication of language-independent universals, or concepts. Since the ‘I think’ does not suffice by itself for the actual logical combination of concepts in propositions, however, the focus shifts next to logical forms of judgment which make any ‘I think’–generated concept combinable with any other such concept in a single act of thought, or propositional representation, and furthermore permit the resulting propositions to be combined in complex propositions. The result is the unification of all possible concepts and propositions in a single consciousness a priori—not literally, to be sure, but only as a synthetic unity, which nevertheless suffices to confer on the ‘I think’ a truly universal representational scope, extending beyond sensations, mental images, associations, emotions, and other nonlogical representations to include concepts and propositions. By far the most important of the representations made psychologistically explicable by the combination of the universality of the ‘I think’ with the logical forms of judgment are pure concepts of the understanding. They represent restrictions on the logical freedom inherent in forms of judgment, and so provide a means of introducing extra-logical necessity into non-linguistic representation of precisely the sort Kant needed to refute Humean skepticism. To do this, however, he would first have to show that pure concepts of the understanding suffice to explicate the categories of traditional metaphysics, including cause and effect. The obstacle to so doing is that the categories are concepts of objects, whereas the pure concepts of the understanding have none but the purely formal logical meaning, devoid of all objective content, conferred on them by the logical forms and the ‘I think.’ How can the latter possibly explicate the

168  Nature in Mind: Through Kant’s I’s former given their radical heterogeneity in content? To answer this, however, Kant was obliged to trace the logical I back to the aesthetic I that first makes it possible—a topic requiring a separate chapter of its own.

Concepts in Mind The pure consciousness ‘I think,’ uniquely among purely psychological representations (i.e. representations in no way involving convention), is genuinely universal. But because it is entirely devoid of representable content, it depends on the presence of other representations, and ultimately on sensations, in order to take place (exist) at all. There is, of course, no reciprocal dependence: sensations may be present in minds from oyster to human in the absence of the ‘I think’ and pure understanding generally. The same independence vis à vis the ‘I’ applies to conscious representations dependent on sensations: memories, images, associations, and beliefs, together with the emotions, desires, and volitions that arise in response to sensation and other sensation-derived representations. But what sensations and all these representations lack is logical universality. Thus, without the ‘I think’ there is representational content without universality, and so no possibility of logically structured conscious thought, while in the absence of sensations there is logical form without representational content, and so no possibility of thought at all. Noting this fundamental divergence, it occurred to Kant that the two may not be contrary but rather complementary, so that, in receiving sensations and sensation-derived representational content to think, the ‘I think’ in turn confers the form of logical universality on that content. Thus does red sensation become a representation of red in general, the image of a unicorn a representation of unicorns in general, the customary association of fire with smoke a representation of the relation of fire to smoke in general, and so on for absolutely any conscious representation into which the universal self-consciousness ‘I think’ is incorporated. The underlying idea is the following. Being a priori, the ‘I think’ is dependent on no particular sensation or sensations as distinct from others. This indifference means that any set of sensations will do just as well any other, whether produced by human senses, other terrestrially or non-terrestrially actual senses, or senses that are possible (e.g.  by futuristic genetic engineering) but happen not to be actual. As a priori, it is equally indifferent to whether or not the empirical psychology operating with those sensations is associative like ours, and, in particular, is completely independent of the customary resemblance association that Berkeley and Hume deemed essential to non-linguistic general representation. Indeed, the universality of the ‘I think’ extends infinitely beyond the scope of any empirical psychological notion of generality whatsoever. For it not only includes any and all representations each of us individually has, or is capable of having, but also all possible sensations and sensation-based representations

The Logical I 169 without exception, whether humanly possible or not. It thus defines an a priori logical universe that quite literally encompasses all possible conscious representations: anything outside it is ipso facto not a possible conscious representation, while nothing inside it, no matter how different from everything else, fails to have the universal self-­consciousness ‘I think’ in common with all other conscious representations. It may seem curious to declare the ‘I think’ universal when everything in the real universe that is not a conscious representation is excluded from its scope. Kant’s response is that a consciousness that includes every possible representation within its scope cannot be supposed to omit anything except what is, quite literally, unrepresentable. Since the logical universe as logicians conceive it likewise cannot include anything except what is representable, anything excluded from the scope of the ‘I think’ also cannot be part of logicians’ logical universe. As this includes the real universe in every representable regard, and since what is unrepresentable is ipso facto nothing to us, any distinction between the logical universe of logicians and the universality of the ‘I think’ is necessarily a distinction without a difference. Hence, the scope of the universal self-consciousness ‘I think’ coincides perfectly with logicians’ logical universe. To be sure, the logical universe of language, as convention-dependent representation grounded in human sociality and lifeways, transcends any individual’s consciousness. Does this mean that linguistic representations are not accompanied by the ‘I think,’ with the implication that the scope of the ‘I think’ is not in fact truly universal because it omits that portion of the logical universe comprising the sayable? Actually, no: because we are necessarily oblivious of all we are never even darkly conscious of, and since we cannot talk about what we are completely oblivious to, the limits of the linguistic logical universe coincide with the limits of the logical universe of conscious representation as demarcated by the ‘I think.’ What cannot be thought of at all because it is consciously inaccessible is ipso facto unsayable. Indeed, insofar as language may be understood as a toolbox of techniques for communicating what otherwise is confined to the individual isolated consciousness, it is difficult to see how its logical universe could ever extend beyond the contents of actual conscious representations. Thus, Kant did not hesitate to conclude that the scope of the ‘I think’ (= all possible conscious representations) does indeed define the logical universe of all representations without exception, linguistic no less than purely psychological.2 This is Kant’s first and decisive psychologization of logic: explicating the logical universe by the universal self-consciousness ‘I think.’ To concede it is to grant that, in the final analysis, logic must be understood through pure apperception, starting with concepts, proceeding to propositions, and finally inferences. ‘Concept’ is Kant’s term for any non-universal conscious representation that becomes universal by relating it via the ‘I think’ to the logical universe of possible representations.

170  Nature in Mind: Through Kant’s I’s Because the term is nowadays used in such a wide range of significations, from purely psychological sortals of the kind Hume ascribed to nonhuman animals to language-bound concepts dependent on convention and uniquely human lifeways, I  will henceforth adopt the term ‘AUAconcepts’ to avoid confusion. Analytic unity of apperception (AUA) is Kant’s designation for the ‘I think’ specifically in its capacity as that consciousness whereby non-conceptual representations can alone acquire the form of universality requisite to transform them into concepts (“the analytic unity of consciousness which makes [a representation] into a conceptus communis  .  .  . attaches to all common concepts as such”).3 The form of AUA-concepts represents the logical universe itself, while their content represents that portion of the logical universe marked off by the sensation, association, or other representational content thought in it—a portion filled by all possible AUA-concepts in which that content is included, however much they may differ otherwise (i.e. in the remainder of their contents). For example, incorporating the representation ‘I think’ into a sensation of red immediately converts it into a universal that indifferently picks out all concepts in the logical universe, possible or actual, that share its content: cherries, red apples, blood, rubies, the Chinese flag, etc. Here the abstract generality of scope of the form of an AUA-concept is matched by an abstract generality of reference of its content. For the latter goes no further than dividing the logical universe in two: representations that do and do not share its content, however different otherwise: red vs. not red, wax vs. not wax, etc. In all other respects, it leaves them completely unspecified, not sufficing even to distinguish all red things from some red things from ten red things from a few, a pair, one, or any other way of singling out red things so as to reference them individually, as a group, or in their totality. Doing so, according to Kant, requires more logic than the ‘I think’ by itself is capable of supplying. Of course, arbitrarily adding AUA to non-universal representations will seldom result in useful concepts and would quickly gum up the mind’s works with useless ones. To avoid this, logic must dovetail not only with a priori psychology but with empirical psychology as well. To arrive at the most useful concepts, the understanding generally only incorporates the ‘I think’ into products of associative imagination, particularly associations encountered with sufficient frequency and constancy to instill a custom. Thus, if we encounter birds often enough to form a customary resemblance association, we need only attach AUA to its representation to add a new, empirically useful concept to our stock of AUA-concepts. And insofar as birds are associated, e.g., with eggs, equally useful AUAconcepts of eggs and their relation to birds can be added as well.4 What difference does adding AUA make to useful representational products of empirical psychology? For Kant, it is nothing less than the difference between blindness to associative relation and consciousness of it.

The Logical I 171 Humean imagination consists entirely of conscious comparisons actually made between actual sensations and their sorting into actual associative groupings. And while this process may serve to ingrain customs that thereafter lie in readiness to be triggered by any similarly associable sensations that may afterward present themselves (Humean “universals”), it cannot, pace Hume, yield true universals because custom leaves us entirely oblivious both to the logical universe of possible representations and to its bifurcation by the association. The process is entirely unintelligent, a blind response to feeling (of facility and vivacity) that can never let go of its affective leading strings without abruptly coming to a halt. Kantian AUA-concepts, by contrast, are genuinely intellectual. They enable us to consciously represent each and every associative combination as a grouping of denizens of the logical universe that are thereafter sortable not only by their sensible/imaginable properties but by their logical ones as well. For example, it makes it practicable to represent one associationderived concept (e.g. swans) as a species of another (large water fowl), so that every possible non-conceptual representation in the logical universe that instantiates the former ipso facto instantiates the latter. Here, both the species–genus relation and the relation of instance to a concept count as logical relations because they are possible only insofar as they involve (at least dark) conscious representations of the logical universe and its bifurcation. Thus, if Kant’s psychologization of the logical universe as the universal self-consciousness ‘I think’ (AUA) is correct, then creatures restricted to associative imagination but lacking AUA are ipso facto completely oblivious to these and all other logical features of their representations and the further logic-imbued sortings (e.g. sub-species to any iteration) they make possible.5 The radical heterogeneity between logical and non-logical psychological representation is fundamental to Kant’s psychologism. In the same way that psychological representations generally and conventiondependent linguistic representations cannot be considered different species of the same genus (Chapter 5), AUA-concepts are more than just a further species of psychological representation. Wherever the representational character of a consciousness depends on the bifurcation of the AUA-constituted logical universe, it can have nothing in common with the representational character of any that does not. The senses present colors, odors, sounds, and all the rest without the least regard for the logical universe. The memory recalls, the imagination fantasizes, and the affective faculties fill us with emotions and desires, all with complete indifference to the logical universe. AUA-concepts, by contrast, conceptualize wholly in and through its bifurcation, while everything else going on in consciousness contemporaneously is completely indifferent to this mode of representation. Their representational character, as logical, is therefore no less fundamentally distinct from the various species of nonlogical conscious representations than it is from linguistic representation.

172  Nature in Mind: Through Kant’s I’s Indeed, for Kant, the difference in psychological capacity between a mind capable of fashioning AUA-concepts from its representations and one lacking that capacity is as profound and consequential as the difference between creatures with even the barest minimum of thought (conscious memory, imagination) and those, like Locke’s oyster, capable of sensation only. For the addition of a logical dimension to psychological representation marks nothing less than the divide between intelligent and unintelligent minds. To be sure, creatures that associate purely affectively, without consciousness of the logical universe and its bifurcation, are capable of complex, highly sophisticated behavior. But can their behavior truly be described as intelligent if their consciousness is lacking even the darkest representations of AUA-concepts and their logical properties? If, as Kant contended, logically structured propositional thought is impossible without such representations, the answer has to be no. And since this is just to say that the ‘I think’ gives psychological reality to intelligence, the advent of universal self-consciousness, via its neural correlate, signifies nothing less than the emergence of intelligence itself as a natural phenomenon. Appreciation of this shows why it is mistaken to assume, as many do, that Kant restricted intelligence to linguistically competent humans. Any mind with the representation ‘I think’ ipso facto counts as intelligent, even pre-linguistic infants.6 Intelligence can be accorded to any creature, actual or possible, that is capable of pure apperception, however unlike humans otherwise, even including beings so asocial as to be devoid of anything remotely analogous to language or socially grounded symbolic communication of any kind. This is not to deny that Kant regarded all non-human animals known to him as incapable of apperception and therefore unintelligent. But that does not mean he would have persisted in that view had he known what we know today. Surely Homo neanderthalensis, H. heidelbergensis, and possibly H. erectus as well were as “intelligent” in Kant’s sense as today’s human infants. What about the antecedents of Homo erectus back to the split with chimpanzees? Hominoids, catarrhines, haplorhines, and their evolutionary antecedents? Cetaceans, proboscideans, corvids, psittacines, and possibly even certain non-avian therapods such as Troodons? The only point about which Kant was certain is that intelligence requires the representation ‘I think,’ so that if any of these creatures have the latter, they ipso facto have the former as well. And, in the first instance, this simply means they have, independently of language, the ability to represent the products of their senses and associative imagination as denizens of the logical universe.7

Logical Form Psychologized Ironically, the advent of AUA-concepts threatens to undermine the universality of the very universal self-consciousness that makes them

The Logical I 173 possible. Any non-universal representation—sensation, memory, image in thought, associative relation, etc.—can be combined in a single consciousness with any other non-universal representation simply by perceiving or imagining them together. But once converted into universals by the addition of analytic unity of apperception, the non-logical modes of combination that feature in sense perception and imagination cease to apply. This is because the combination of distinct universals requires a way of combining representations of different ways of dividing up the logical universe in a single consciousness, which neither sense nor imagination (associative included) is capable of doing. Hence, without a special, uniquely intellectual ability to combine distinct AUA-concepts specifically according to their logical character and consciously represent that combination, no single consciousness, including the ‘I think,’ could be represented as able to accompany, and so as common to different AUA-concepts, much less all possible ones. Yet, for even one representation to fall outside the scope of the ‘I think’ means not only that the latter forthwith ceases to constitute a universal self-consciousness perfectly coincident with the logical universe, but that “something would be represented in me that could not be thought at all, which is as much as to say the representation would either be impossible or at least be nothing for me.”8 So, for any and all AUA-concepts I create, or could create, to become anything at all for me, my ‘I think,’ I must be able to logically combine them in a single consciousness. Since pure understanding is the capacity from which AUA-concepts originate, Kant looked to it as well for a purely logical means of combining AUA-concepts in a single consciousness. However, to properly understand the logical forms of judgment he identified as meeting this requirement, we first need to distinguish two levels of generality in his psychologism. At the lower level there are the particular logical forms whereby the pure understandings of humans and similarly constituted creatures combine AUA-concepts in a single consciousness. They are the logical forms Kant drew from Aristotle: subject-predicate form, groundconsequent form, plus ten others.9 Since, on Kant’s analysis, they are forms specifically for combining AUA-concepts in one consciousness, their meaning is primordially psychologistic even before it is logical or linguistic. This means that they must be considered first and foremost in their role of preserving the universality of the ‘I think’ by enabling it to be extended to include AUA-concepts whose defining logical character would otherwise “be nothing for me.” Kant accordingly made sure to suggest, without affirming or denying, that subject-predicate and the other Aristotelian logical forms are phylogeny-relative, i.e. that other creatures may extend the scope of the ‘I think’ to AUA-concepts by means of altogether different logical forms than Aristotle’s.10 Thus, his psychological philosophy ascends to a level of generality where the particular logical forms do not matter and only their role in egocentric a

174  Nature in Mind: Through Kant’s I’s priori psychologism is considered—logic as a purely psychological affair of the a priori unity of consciousness. The question of what kind of logical form is requisite to restore universality of scope to the ‘I think’ by extending it to AUA-concepts becomes, at the more general level, a choice between a logical relation that literally combines all possible AUA-concepts in a single consciousness and one that instead issues merely in what Kant termed a synthetic unity. A literal unity of all possible AUA-concepts in a single consciousness is manifestly impossible, not only because no finite mind is capable of actualizing all possible (i.e. the infinite totality of) AUA-concepts in consciousness at a single instant, but also because logical forms like Aristotle’s have the capacity to combine only finite numbers of AUA-concepts. In the Aristotelian case, it is mainly two by two: a form to combine two AUAconcepts in a subject-predicate proposition, which another form can then combine with another subject-predicate proposition to form a groundconsequent proposition.11 Clearly, it is impossible to combine all possible AUA-concepts in a single consciousness in this manner. In a merely synthetic unity of all possible AUA-concepts, by contrast, two-by-two is quite sufficient. This is because the consciousness of AUAconcepts is here purely formal, so that we are not even required to represent any particular AUA-concept, much less all of them. Instead, we are conscious only of the logical positions in which any of the infinite totality of possible AUA-concepts can be placed, so that, when filled, a single act of thought unifying its constituent AUA-concepts ipso facto results.12 In beings constituted like us, the a priori synthetic unity of AUA-concepts in a single consciousness is achieved by the Aristotelian logical form of categorical judgment. This form has two positions for AUA-concepts, subject and predicate, which it logically relates in assertoric modality both qualitatively and quantitatively. For example, a categorical relation between the AUA-concepts ‘water fowl’ and ‘edible’ asserts that some or all of the part of the logical universe specified by ‘edible’ is predicated affirmatively or negatively of the part of the logical universe specified by ‘water fowl.’ But, again, it is not the particular forms that matter at the most general level of Kant’s a priori psychologism. It is instead that a logical form of judgment is the formal representation of a synthesis that can logically unite any one of the infinite totality of possible AUAconcepts with any other in a single act of thought—a consciousness that can therefore be represented as able to accompany, and so as logically common, all possible AUA-concepts. Thus, although merely a synthetic unity of AUA-concepts, the logical form of categorical judgment, or any equipollent form, restores unrestricted scope to the universal self-­ consciousness representation ‘I think’ by incorporating AUA-concepts into its unity alongside non-logical representations, and thereby succeeds in making such concepts, possible no less than actual, cease any longer to be “nothing for me.”

The Logical I 175 A logical form is purely formal for Kant in that its representation takes no account whatsoever of the contents thought in the resulting judgment (since this includes sensational contents, formal representations are ipso facto pure as well). The logical form of AUA-concepts is the relation of their contents, whatever they may be, to the logical universe that results when the universality of the self-consciousness ‘I think’ is incorporated into their otherwise non-logical representation. The logical form of judgment is the means to realize the potential for different representations of this universe to be combined in a single consciousness. Of course, as purely formal, the addition of logical form is only necessary, not sufficient to form a genuine proposition, as logical form by itself cannot guarantee that a proposition will result. In particular, even if a would-be proposition is logically well formed, it would still fail to be a genuine proposition if, for other than strictly formal logical reasons, it cannot be thought without generating not just falsehood but one or another species of impossibility, e.g. “2 + 2 = 77,” “water is not always H2O,” or “Elms love to ski.” It thus cannot be stressed too strongly that Kant’s transcendental (i.e. a priori psychologistic) concern with Aristotelian logical forms extended no further than the purely formal logical synthetic unity of AUA-concepts whereby they preserve the universality of the ‘I think.’ Yet, the very logical form that solves the problem of unifying AUAconcepts in a single consciousness creates a new problem of unifying the resulting propositions. The explanation of why this is so starts similarly: if the propositions formed from logical synthesis of AUA-concepts could not themselves be synthetically combined in a single consciousness, then no consciousness would be able to accompany, and so be logically common to them, including the ‘I think,’ making all propositionally formed consciousnesses “nothing for me.” A form like categorical judgment is, however, unsuited to this task since it is specifically adapted to have its two positions, subject and predicate, filled by AUA-concepts, not assertoric, categorically related propositions; nor can anything save AUA-concepts be logically quantified or qualified. Filling them with the propositions that result from combinations of AUA-concepts is no more possible than filling the subject and predicate positions of categorical propositions with representations that do not have AUA attached and so do not include relation to the logical universe in their representation. Thus, the only way distinct propositions can be combined in a single consciousness is by means of a specially adapted logical form whose positions can be filled by propositions composed of categorically related AUA-concepts and nothing else. Crucially, this form too must be a form of judgment, for in order to be possible purely psychologistically (non-linguistically), as a synthetic unity of consciousness, its result has to be representable in a single act of thought, or proposition (Chapter  5). The difficulty with this is that all propositions as such are already single acts of thought, so that a

176  Nature in Mind: Through Kant’s I’s compound of two propositions would consequently not be one thought but three: the two component propositions plus the compound, which would somehow have to be simultaneously one thought and three. Since what is required to reestablish the universality of the ‘I think’ is a way of synthetically uniting all possible propositions via a single act of thought that can then be represented as a consciousness common (logically universal) to all propositions as such, this simply will not do. The solution lies in a logical form that suspends the act of thinking in the component propositions so that there is in fact only one proposition: the compound thought. In creatures constituted like us, logical suspension is effected by replacing the logical form of assertoric modality of stand-alone categorical propositions with the logical form of problematic modality, so that the only thing asserted is the relation of the two propositions, not the related propositions themselves. The principal form this takes in creatures like us is the logical form of hypothetical judgment, which assertorically relates problematic propositions as ground to consequence.13 Since, in this way, any of the infinite totality of possible propositions combining distinct AUA-concepts can be combined with any other in a single consciousness, the logical form of hypothetical judgment or any equipollent form thus serves to expand the I-demarcated logical universe to include propositions formed from AUA-concepts as well as those concepts themselves.14 As with the logical forms of categorical judgment, quality, quantity, and assertoric modality, the unity established by the logical forms of hypothetical and problematic judgment is not a real unity, actually combining all possible propositions simultaneously in a single consciousness, but a purely formal synthetic unity whereby any proposition in the infinite totality of propositions possible from AUA-concepts can be combined with any other by placing them in the ground and consequent positions. Nor can we suppose that these logical forms are valid ­phylogeny-independently, and so the only ones capable of establishing an all-encompassing synthetic unity of the propositions possible from AUAconcepts. The claim is merely that they are sufficient therefor, not that they are necessary. For again what is fundamental are not the particular logical forms involved but the a priori psychologistic implication that some such logical synthesis suffices to formally unite all possible propositions relating AUA-concepts in a single synthetic unity of consciousness. Only if such a consciousness exists can the ‘I think’ be represented as able to accompany, and so as logically common to everything capable of being represented in a single act of representation, propositionally formed consciousnesses included. The synthetic unity of propositional thought effected by logical forms like that of hypothetical judgment thus serves to expand the logical universe to include all representations without exception, propositions no less than AUA-concepts and the non-­ logical representations that supply their content (sensations, associations,

The Logical I 177 etc.), and so restores unrestricted universality to the representation ‘I think.’ Moreover, because it brings the regress of logical forms to an end, it restores it definitively.15

Propositions in Mind Kant’s psychologization of logic by way of his interpretation of the ‘I think’ as a logically universal self-consciousness is perhaps the first serious attempt to square propositional thought with human-animal psychology and, insofar as it purports to do so without ever presupposing linguistic convention, may very well stand alone even today. For this reason, it is helpful to view it through an evolutionary prism. It begins with a creature lacking an ‘I’ but nevertheless capable of the kind of highly sophisticated, behaviorally efficacious conscious mentation that Berkeley and Hume devised their associationist psychology to explain. Thanks to some fortuitous mutation or other alteration in its genome, its progeny included creatures capable of the representation ‘I think.’ Having no evident selective advantage by itself, this neural capacity presumably could have established itself in the population only as a spandrel piggy-backing on some genetically connected trait that earned its evolutionary keep (to be discussed in the next chapter). Given sufficient time for further opportune genomic alterations to occur, their progeny’s brains would be capable not only of generating AUA-concepts but also of deploying logical forms to combine them in propositional representations that they could then connect via more complex propositions and inferences, all without benefit of language. Since the behaviors this logical synthetic unity of consciousness makes possible include everything that sets species like ours apart intellectually, such creatures would presumably have enjoyed advantages in the competition for survival over less logically well-endowed individuals and populations. Nevertheless, some will find the claim deeply problematic, even selfrefuting, that anything as abstract, intellectual, and evidently non-­ subjective as logical universality and logical forms of judgment can possess the same psychological reality as sensations, memories, dreams, emotions, and everything else capable of manifesting itself in the purely private, isolated, uniquely subjective realm of an individual creature’s consciousness. Yet, one must be careful not to set the bar too high here. Are sensations, emotions, desires, and any of the other representations that supply logical forms with content any less mysterious? The position I  have adopted in this book and that seems to me the default position everyone today is obliged to take is that psychological (conscious) reality is entirely a function of evolved neural correlates. Since it is a complete mystery why and how those neural phenomena rather than other neural, organic or non-organic natural phenomena are able to yield consciousness of any kind, why should there be more of a problem where their

178  Nature in Mind: Through Kant’s I’s yielding logical forms is concerned than sensations, emotions, dreams, or anything else? Conversely, all conscious representations are equally informational blank slates regarding their neural correlates and everything neural and extra-neural involved in their correlates’ causation. So, if logical forms have none of the features that informationally connect them to neural counterparts, is the same any less true of sensations and the rest? The connection between the subjective psychological reality of consciousness and the objective physical reality of neurophysiology is a complete mystery in both directions, today and quite possibly for some time to come. Thus, Kantian logical forms of judgment pose no special mystery but instead are best regarded as simply an additional species of phylogeny-dependent neuro-psychological reality in addition to sensations, emotions, dreams, and the rest. Certainly, there is a great deal of philosophical tradition standing in the way of accepting that propositional thought, being identical in existence with its neural correlate, can be as much a part of the neural portion of the realm of nature as sensation, memory, imagination, and the rest of conscious representation. How can purely formal, a priori logic be part of physical reality? Even if it is supposed to be innate, as a Fregean or a Chomskyan might maintain, that can at most explain the origin of logical representation in us, not its origin tout court. As regards the latter, some place logical form in the same purely intelligible “platonic” realm in which mathematics is often supposed to reside, whereas those who restrict reality to physical reality often situate logic in the same sphere of social convention as language. Some do not even distinguish logic from language at all, so that asking whether logic is part of nature is no different from asking the same of Semitic syntax. What all agree in opposing is the notion that propositional logic is neither language nor mathematics, neither conventional nor platonistic, but fundamentally psychological. How, then, is Kant’s psychologization of logic to be defended? Kant would be the first to admit that logic is not part of empirical psychology. His psychologistic explanation of logic relies exclusively on a single datum: the purely formal, a priori representation ‘I think.’ This pure universal self-consciousness is, on his analysis, indistinguishable from logic’s most fundamental representation: a universality sufficient to define the logical universe of concepts logically combinable into propositions. If he is right, then, instead of the many absurdities identifiable in empirically psychologized propositional logic à la Locke or William James, the a priori psychologistic character of logic becomes obvious and undeniable. For if the logical universe is ineluctably part of a priori psychology, how can the logical forms adapted to it by evolution not be so as well? Admittedly, Aristotelian logic is today commonly regarded as a quaint relic of an embarrassingly primitive analysis of propositional thought that has long since been superseded by mathematical logic, philosophical

The Logical I 179 linguistic analysis, and linguistics proper. It is seen as especially mortifying in Kant’s case because, on the eve of the century in which mathematical logic finally took off, he declared Aristotelian logic to be a completed science that could never be added to or subtracted from. What possible response could Kant have? Actually, a quite persuasive one, at least if his analysis of the logical significance of the universal self-consciousness ‘I think’ is taken seriously. In the first place, Kant would have been the first to concede that propositional analysts not only can but should ignore the psychologistic explications required by some of logic’s most fundamental representations. For propositional analysis has as its task to represent the logical structure of propositions and their relations without regard to how they originate, whether in the psyche as conscious representations, in language as conventional contrivances, or in some version of platonic heaven. However, to say that propositional analysis can ignore the psychological nature and origin of fundamental logical representations does not make it competent to pronounce against it. Precisely because it cannot explain logic’s most fundamental representations but must instead take them for granted, propositional analysis can no more be regarded as a source of evidence against than for Kant’s a priori psychologism. And the same is true of everything else that, in his view, owes its most fundamental, otherwise inexplicable representations to a priori psychology, including mathematics, science, ethics, and aesthetics: they not only need not but cannot take their psychological nature and origin into account. Kant also would not have hesitated to admit that the number and variety of propositional forms in mathematics, science, and ordinary thought are potentially limitless. However, his question was whether any of these forms can properly be regarded as logical. Kant answered no for the same reason he claimed that Aristotelian logic is a completed science to which nothing can be added or subtracted for all future time. For, on his analysis, the divide between logic and mathematics in particular can be drawn neither by propositional analytic nor by mathematical means, much less linguistic ones, but is instead entirely a function of the different a priori psychologistic inputs each requires. Indeed, so far as he was concerned, his most original contribution was the thesis that pure understanding as defined by the ‘I think’ presupposes a pure sensibility,16 and one of its most important consequences is that mathematics, in contrast to logic properly so called, must incorporate contents that originate in this faculty and is impossible otherwise. Nor is that all: mathematics also presupposes the determination of pure sensibility by what Kant termed pure concepts of the understanding. I will examine these concepts in the next sections and their determination of pure sensibility in later chapters. Here it suffices to remark that Kant distinguished logic from mathematics and everything else as being free of all dependence not only on pure concepts of the understanding

180  Nature in Mind: Through Kant’s I’s but on any and all content provided by sensible representations, pure no less than empirical, making logic the science of the understanding in its own right, independently of its relation to all other faculties. By contrast, mathematics, on his analysis, presupposes both the manifold of pure sensibility and its determination conformably to pure concepts of the understanding—a process he termed transcendental synthesis (Chapter  9). Moreover, he made quite clear that “mathematics” here means not just geometry but arithmetic, algebra, and calculus as well, i.e. the state of the art in his own day,17 and he left no doubt as to his readiness to extend his thesis not only to all future mathematics but to all mathematics without exception, whether humanly possible or not. And this obliges one to ask the following: would Kant have classified today’s mathematical logic, inclusive of everything from truth-functional logic to set theory, Gödel, and beyond, as logic or mathematics? For if the answer is mathematics, so that mathematical logic would have to be supposed no less dependent on pure sensibility than simple counting or Euclidean geometry, then Kant can no longer be condemned for giving hostages to fortune when he maintained that logic as bequeathed by Aristotle is a completed science.

Propositional Thought Without Language Kant’s distinction between the logical form and non-logical content of conscious representations may seem vulnerable to a kind of critique best set forth by Wittgenstein. In the early stages of his career, from 1911 through the Tractatus Logico-Philosophicus (1921), Wittgenstein sought to eliminate traditional philosophy in favor of mathematical logic, which he traced to a foundation in the formal properties of language. At that point, he conceived of language in Frege’s platonist rather than a conventionalist manner. This changed, however, when Wittgenstein came to the realization that language not only has no need of incorporating the purely formal features on which he had previously supposed it to depend, but is quite incapable of incorporating them. Take negation, for example. Logical negation was supposed to be completely indifferent to the content negated, so that if a thing is not red, then the logical scope of what it may be instead is left completely unrestricted: it might as well be painful or equal to the square root of forty-nine as a color other than red. What Wittgenstein came to appreciate is that the denial that something is red makes sense, and so counts as in good logical order, only if it is a color other than red. The implication is that the logic of negation is not as indifferent to context and content as he had formerly supposed. At first Wittgenstein tried to rescue the formal conception of logic by distinguishing different logical senses of ‘not’: one for the “logical space” of color, another for the “logical space” of arithmetic, and so on for the logical spaces defined by all other contexts and subject matters.18 In the

The Logical I 181 end, however, he was compelled to acknowledge that negation is just too context- and content-specific even for that. For example, if I say that the early morning sky is not blue, it then only makes sense to say that it is gray or (at dawn) red, but not any other color (green, purple) or pattern of colors (piebald, polka-dot, etc.). Realizing that similar considerations apply to everything he formerly regarded as formal logical in character, Wittgenstein concluded that there is no such thing as a formal logic of propositional representation at all, and rather than confuse matters by continuing to speak of the “logic of language,” he preferentially spoke instead of the grammar of concepts employed as instruments in language games adapted to human lifeways. Although Wittgenstein never singled out Kant’s conception of logic for criticism, he treated the entire post-Cartesian fascination with the ‘I think’ as a case of philosophers being deluded by the seeming logical possibility of preceding anything one says or thinks by the words ‘I think.’ To recognize the supposed insight this affords into the foundations of logic not only as barren but as nonsensical, just imagine what would result if it really were universalized by actually setting the words ‘I think’ before everything one says or thinks. Far from anything of philosophical moment, it would merely be a tic, no different from, nor more consequential than, prefacing every proposition with “Hey!” or intoning all propositions as rhetorical questions. For Wittgenstein, the remarkable career of the representation ‘I think’ is yet another instance of the baleful tendency of philosophers to be misled by grammar into subliming the logic of language. If Kant’s purposes required that the universality of the ‘I think,’ the logical forms for combining AUA-concepts, and the resulting propositions be realized linguistically, much less play an essential role in language, then his conception of logic would fall squarely within Wittgenstein’s crosshairs. But they do not. Indeed, Kant would probably be the first to grant that Aristotelian logical forms of non-linguistic propositional thought are far too coarse-grained to capture the logic of virtually anything in language. From the standpoint of the a priori psychologist, however, the goal is not for the logic of conscious propositional thought to capture language but the reverse; and, to that end, a logic like Kant’s, primitive enough to be captured by any language whatsoever, is precisely what is required. On the principle that convention cannot empower us to express anything to which we are oblivious (i.e. not even darkly conscious of), language must be preceded by non-linguistic, purely conscious propositional thought, with a logic specially adapted to actuate it, from which non-psychological, convention-based propositional discourse can then take its start. This condition is satisfied by the ‘I think,’ the universality of which is not that of an ordinary representational content, much less a formulaic tic, but a contentless consciousness representable purely formally as logically common to all representations, possible no less than actual.

182  Nature in Mind: Through Kant’s I’s The logical forms founded on it merely serve the psychologistic end of combining representations to which the universal self-consciousness ‘I think’ has been attached—i.e. AUA-concepts—in a single act of propositional thought, which representations the ‘I think’ can also accompany and so incorporate into its logical universe. Thus, despite sufficing for a purely formal synthetic unity that confers unrestricted representational universality on the ‘I think,’ Aristotelian forms of propositional thought remain primitive enough that any language humans are likely ever to concoct will still be able to capture the logic of non-linguistic, purely psychological propositional representation. What then of such notions as subject, predicate, copula, all/some, affirmative/negative, and assertoric that come with the logical form of categorical judgment as a single logical package? Or such notions as ground, consequent, and problematic that come with the form of hypothetical judgment? Are they not, one and all, essentially linguistic? Yet, to insist on answering affirmatively, one has to disregard the extremely restricted sense conferred on them by Kant’s a priori psychologism. Their scope therein is confined entirely to the logical universe constituted by the ‘I think,’ its bifurcation by the contents of AUA-concepts, and its extension to include the logic of propositions formed from AUA-concepts and complex propositions formed from these. There is no further context or content—sensible or any other, pure or empirical—to differentiate propositional representation into the distinct, delimited logical spaces of post-Tractatus Wittgenstein, much less the language games of later Wittgenstein. For example, Kantian logical quantity (all/some) differs from more robust notions of quantity in that Kant’s logical universe does not consist of objects or anything distinguishable individually, and so does not allow for the representation of any kind of enumerable multiplicity. There is nothing countable in it, no amounts of anything, nothing increasable or decreasable, no possibility of addition, subtraction, or any other arithmetic operation, nothing amenable to being set in 1–1 correspondence or any of the other modes of ordering that become conceivable on mathematical and mathematical-logical concepts of quantity. None of the quantitative notions that might plausibly be supposed to specify a logical space or a language game can get a purchase given only the foundation furnished by the logical universe constituted by Kant’s logical ‘I.’ There are only the two properly logical ways of partitioning it quantitatively: all or some.19 Since this is true regardless of the content of the representations concerned, there seems to be no reason not to regard them as properly logical forms that apply universally to all AUA-concepts. And if this is deemed to be too primitive a mode of representation to qualify as language, then Kant would again be the first to agree, for AUA-logic does not need to capture language, only to be captured by it. The case is no different where logical relation is concerned. Relation by the categorical copula as subject and predicate is completely indifferent to

The Logical I 183 the content of the AUA-concepts concerned, nor does their determination in this relation alter or affect their content in any way. It thus contrasts with language, where, in general, only nouns, noun-phrases, and pronouns may be subjects and only verbs, verb phrases, adjectives, and adjectival phrases may be predicates. Even then, when the subject is a thing only attributes suited to things can be predicated of it without nonsense resulting. For example, if “some dogs” is in the subject position, “have brown fur” makes sense as a predicate but not “have ‘-tion’-endings.” Of course, if “some nominalizations” is the subject, “have ‘-tion’-endings” makes sense, while “have brown fur” does not. By contrast, the logical form of categorical judgment is completely indifferent to the context and content of the AUA-concepts that occupy the logical positions of subject and predicate, leaving the judger free both to predicate anything of anything else and to reverse the order so that what was subject becomes the former predicate’s predicate. Yet, while their grammatical indeterminacy may make them useless for linguistic purposes, and may not even result in a genuine proposition at all (e.g. “some nominalizations have brown fur”), they remain available to perform roles in purely psychological representation, up to and including transcendental synthesis. The logical form of hypothetical judgment is likewise completely indifferent to context and content, permitting any categorical judgment to be represented as ground or consequent of any other and leaving the judger free to reverse their order. All that matters—all that ever matters for Kant’s a priori psychologism—is the synthetic unity of consciousness these forms ensure, not whether they are employable linguistically or in the least useful empirically. And since the same indifference and reversibility, directed to the same quintessentially psychologistic end—a priori synthetic unity of the manifold of representation in one consciousness— apply to the remaining logical determinations he derived from Aristotle, Kantian logical forms thus present no target for the kind of considerations that led Wittgenstein to reject the form/content distinction in the logic of language. The mistake one should take most care to avoid is assuming that in order for conscious propositional representation to be the foundation for linguistic propositional representation, it must itself be a language. It should be obvious, and no doubt was to Kant, that a species of representation confined to a meager twelve content- and context-indifferent logical forms, devoid of all properly linguistic—semantic, grammatical, and syntactic—determinateness, and applicable exclusively to nonlogical, purely psychological contents that are essentially the same in non-human animals entirely lacking the capacity for logically structured propositional thought, cannot be regarded as a language, even a private one. Nevertheless, there are some who seem reflexively inclined to equate conscious propositional representation with the kind of wordless private language so devastatingly critiqued by Wittgenstein. His argument, at

184  Nature in Mind: Through Kant’s I’s its core, is that a private language would necessarily lack the fixity only publicly established conventions can confer, that without this it could have no set rules, hence no grammar, and so whatever it was could not be language.20 If we assume, as I think we should, that Kant would agree, what does that leave for the conscious propositional representation he supposed to exist independently of language to be and to do? Far from being language, such representation is in the first instance merely a modest evolutionary enhancement of the kind of mentation that is possible pre-logically, which, for Kant, is best exemplified by Humean customary association. Its principal benefit is to permit a creature to relate its representations even when such feelings as facility and vivacity are lacking and/or in the absence of frequent, constant experience of their co-occurrence. It could, for example, experiment with different categorizations than those association had yielded to determine if others might better serve its purposes; it could hypothesize; and it could contrive categorizations more general than are possible purely associatively (as in Humean “philosophical relations”). But such representations, albeit propositional, would be little, if at all, closer to genuine language than the representations of a creature wholly confined to customary association.21 Why then did Kant suppose that anything so primitive as AUA-logic opens the way to truly intelligent propositional representation, including the most sophisticated kinds of mathematical and scientific cognition? The answer starts with his determination that the categories of traditional metaphysics originate as pure concepts of the understanding that derive their content entirely from logical forms of judgment.

Platonism’s Feet of Clay Platonism continues to appeal to mathematicians, scientists, and philosophers who share the intuition that if truths exist independently of minds, so too must the representational content of the propositions (e.g. equations) needed to think them. The difficulty is reconciling this with the fact that no one who has failed to master the myriad grammatical and other conventions constitutive of vernacular language has ever been able to learn, much less practice or teach, mathematics, science, or any comparably rich cognitive pursuit. For if the representational content of language is essentially a function of conventions inextricably embedded in uniquely human sociality and lifeways, how can the same not be true of the representational content of the silent, spoken, and written cogitation indispensable to the pursuit of these disciplines? It is not just that such esoteric propositions cannot exist in a vacuum. It is that they cannot be understood at all in a linguistic vacuum. And if they cannot be understood apart from linguistic conventions and all the social infrastructure requisite to support them, how can the truths they express, as they alone enable us to express them, be conceived to exist independently of them

The Logical I 185 either? Indeed, can the propositions requisite to assert these truths’ supposed language-independence themselves be formed in a linguistic vacuum? If not, then everything the platonist supposes to be objective and eternal in human representation turns out to be as inextricably bound up with, and impossible apart from, human social convention as a curtsy or a thumbs-up. The platonist can respond by arguing either that language is not in fact everywhere convention-dependent or that mathematics, science, and/or other cognitive pursuits are not everywhere language-dependent. Since no one claims that English or any other vernacular language exists independently of convention, the platonist who opts for the former alternative is committed to positing an unspeakable, incommunicable “eternal language” that confers on conventional signs convention-independent senses that objectively specify a unique convention-independent meaning for each. Philosophies of language in the tradition of Frege and Wittgenstein’s Tractatus may be characterized as platonist in this sense, and the same might be said of Chomskyan linguistics, at least in its earlier phases. The problem with this view becomes apparent when one asks why it should be called “language” at all. Nothing seems to limit the scope of its power to conferring grammatical sense and meaning on English and other vernacular signs. Why not on ideas in the mind as well? Indeed, why not objects in nature? This after all is precisely what Plato’s metaphysics and its Aristotelian variant come to: every representation, conscious or verbal, and every thing, artifact or natural, owes its being (“essence”) to being an ectype of some eternal archetype or a matter with some inherent essential form. The slope from eternal language into platonistic metaphysics is evidently all too slippery. But even if this objection is waved, there is still the point emphasized by Wittgenstein and others that the supposition of an eternal incommunicable language as essential in order for ordinary language to function in even the simplest everyday ways seems entirely gratuitous. As Berkeley put it in a similar connection, the notion that “a couple of children cannot prate together of their sugar plums and rattles and the rest of their little trinkets” until they have set aside all the things of this world to gain access to an ineffable realm of eternal senses and meanings “will be found a hard task for that tender age.”22 English and other vernaculars, like other convention-based human activities, seem perfectly capable of taking care of themselves without the aid of such sublimities. The other option open to platonists is to concede that all language is convention-dependent and insist instead on the language-­independence of mathematics, science, and/or other advanced cognitive pursuits, if not in toto then in certain essential respects. Candidates include abstract categories such as causality, substance, reality, negation, necessity, and existence, as well as spacetime and number; mathematical concepts, propositions, and proofs; and the concepts, models, and theories of

186  Nature in Mind: Through Kant’s I’s physics. But quite apart from the difficulty of demonstrating their ­language-independent provenance, there is the challenge of explaining how the mind is able to access these platonically real representations given that they cannot be acquired merely by mastering the grammatical and other conventions constitutive of vernacular language (Chapter 5). Since colors, sounds, and other sensational qualities incidental to consciousness of these representations are no more part of their intrinsic content than the sound and look of language are essential to the conventions that determine its representational content, access to them also cannot be had through the senses. Platonists tend to characterize the access they suppose themselves to have to platonically real representations as “intuition,” by which they mean a mode of representation fully as passive and immediate as sensation but different in kind. Since the representations at issue are paradigmatically intellectual, most therefore treat the conscious apprehension of such representations as instances of intellectual intuition. The question is how a faculty of intellectual intuition could ever have evolved in creatures that lacked it, and how, even if it did, it could have established itself in a population of hunter-gatherers scratching a living scores of millennia ago in places like the Serengeti. Hume, who saw no reason to suppose that there is anything extra-linguistically present in modern human minds that is different in kind from the contents of the minds of other animal species, mocked “this artifice” of “spiritual and refined perceptions” that “fall not under the conception of the fancy, but must be comprehended by a pure and intellectual view of which the superior faculties of the soul are alone capable.”23 His problem with the thesis is evident even in a notion as humble as that of a triangle. If the representation one has truly is of a triangle, then, according to the platonist, it must already have all the properties that mathematicians could ever prove of these figures. This means that platonists like Descartes were committed to maintaining that everything Riemann and his successors would discover about a geometrical object like a triangle was already there, albeit darkly, in Descartes’s consciousness, Euclid’s, and that of everyone else who has ever had the concept (e.g. the slave boy of Plato’s Meno). Since the number of mathematical properties discoverable in such objects is limitless, perhaps infinite, the representations apprehended in pure intellectual intuition would indeed be of a most superior and refined character! It is difficult to see how platonists could account for the physical possibility of neural correlates able to sustain intellectual intuitions infinitely rich in content. But even setting that objection aside, their advent in our phylogeny does not seem explicable otherwise than as a uniquely improbable saltation, and so the closest thing in nature to a miracle. Not surprisingly, philosopher-mathematicians like Descartes were quite forthright in attributing them to God, differing only on whether intellectual intuition

The Logical I 187 works by our being vouchsafed direct access to the eternal ideas in God’s own intellect (e.g. Malebranche) or by having access to innate ectypes of divine ideas implanted by our Creator (Descartes, Leibniz). For how else than by a kind of miracle could the capacity to intuit platonically real representations have arisen in our hominid lineage, meshed with consciousness as it existed previously, and done so in such a way as to so boost its possessors’ fitness that they prevailed in the lottery of natural selection during the Quaternary, Neogene, or still earlier? Even if there were no alternative explanations, the sheer difficulty of reconciling intellectual intuition with human phylogeny should be enough to give one pause. But there are alternatives to platonism, not least the Wittgensteinian view that mathematics, science, and other sophisticated cognitive pursuits are nothing more than ingenious, uniquely efficacious refinements of existing linguistic conventions and nothing apart therefrom.24 So, faced with the utter biological implausibility of intellectual intuition, as well as the appealing but unsupported and unprovable metaphysical hypothesis of platonically real representations, platonism seems unsalvageable.

The Kantian Alternative: Pure Concepts of the Understanding There may be other alternatives to platonism, but the only one I know of that is truly independent of platonism, empirical psychologism, and conventionalism alike is Kant’s a priori psychologism. Like platonists, Kant denied that mathematical, scientific, and other cognitive representations are wholly dependent on linguistic convention, social infrastructure, and/or human lifeways. Nor is there anything in his position incompatible with the existence of purely intellectual representations able to transcend both Humean associationist and Wittgensteinian linguistic-­ conventionalist explications. But with platonically real intellectual representations excluded, what kind remain? Here it is worth noting that if Kant had taken a single step in his theory of understanding beyond the ‘I think,’ AUA-concepts, and logical forms of judgment explicable purely psychologistically, he would ipso facto have been embracing a form of platonism and made his a priori psychology heir to all the objections to which platonism is. The key to Kant’s entire system, the thing that keeps it from becoming platonism, is the fact that the pure concepts of the understanding affirmed in it contain no content other than the logical forms of judgments. And since this is just to say that they do not take intellection across the boundary separating a priori psychologism from platonistic metaphysics, Kant’s system is indeed as distinct from platonism as it is from empirical psychologism and conventionalism. This is best appreciated by examples. In intelligent beings constituted like us, there is, according to Kant, a logical form of categorical judgment

188  Nature in Mind: Through Kant’s I’s that relates one AUA-concept as subject to another as predicate. Since the content and context indifference of the form means that it does not dictate for any concept which logical position it will occupy, it leaves us free to relate any concept to any other both as subject to predicate and as predicate to subject. However, it also leaves us free to annul this freedom by arbitrarily regarding any concept’s logical position as fixed and unalterable. The logical form of categorical judgment thus becomes the source of two further concepts, one of something that is determinately always and only subject, never predicate, and the other of something that is determinately always and only predicate in relation to subjects so determined (in the case of the latter, this means that the concept can still occupy the position of logical subject in relation to other concepts provided the latter have not previously been determined as always and only subject). These two concepts cannot, however, be understood as themselves logical. Logic, as Kant conceived it, leaves us free to predicate the subject of one proposition of another concept in a new proposition, predicate this subject of something else in yet another proposition, and so on, without ever reaching a final subject25 that can be predicated of nothing else. Consequently, the concept of a final subject—something that is determinately always and only subject—is an extralogical concept even though its sole and entire content derives from the logical form of categorical judgment (directly, no intellectual intuition required). Similarly, the concept of a final predicate as something that is always and only predicate in relation to concepts previously determined as final subjects counts as extra-logical despite having no content other than that furnished by the logical form. Since the logical form of categorical judgment is the source of the sole and entire content of these two concepts, they are no less purely intellectual than the logical form from which they are derived. Thus, final subject and final predicate constitute genuine pure concepts of the understanding. And since their explication is purely a priori psychologistic, with no admixture of platonism, conventionalism, or empiricism, Kant saw in them the key to a new, genuinely critical means of explicating the traditional metaphysical category of the relation of substance and accident. A second example is the logical form of hypothetical judgment in which one categorical proposition is related to another as ground to consequent, with the logical modality of each thought as problematic instead of assertoric (only the whole is assertoric). Since the form does not dictate which logical position a proposition will occupy, it leaves us free to relate any proposition to any other both as ground to consequent and as consequent to ground. But it also leaves us free to arbitrarily annul this freedom by regarding any propositions so related as fixed and unalterable in their relation. To do this, however, we must extra-logically represent not only the ordered relation as fixed but the modality of its components as well, so that they are extra-logically conceived as possible only in that

The Logical I 189 relation and order and impossible otherwise. The latter we can do simply by annulling our logical freedom to regard any proposition as problematic in relation to any other and restricting ourselves to regarding it as problematic in one, and only one, hypothetical relation and order. In this way, we form two extra-logical pure concepts of the understanding— a relational concept Kant used to explicate the traditional metaphysical category of the relation of cause and effect and a modal concept he used to explicate the category of possibility-impossibility—that (1) draw all their content from the hypothetical logical relation and problematic logical modality whereby (2) the proposition in the ground position is represented as restricted to being the ground of the proposition in the consequent position (and no other), and (3) the proposition in the consequent position is represented as restricted to being the consequent of the proposition in the ground position and no other. But although the pure concepts of cause and effect and possibilityimpossibility nullify the logical freedom to undo a ground-consequent relation by either reversing their order or relating them in the same way to other propositions, they leave us free to relate them hypothetically to other propositions so long as they are in a different logical position. Using a single arrow for the purely logical, freely undoable relation, P1 → P2, and a double arrow for the extra-logical concept of the fixed, unalterable relation, P1 ⇒ P2, this is just to say that while the latter prevents us from representing P1 ⇒ P3 and P3 ⇒ P2, it still permits us to represent P1 → P3, P3 → P2, P2 → P1 (but not P2 ⇒ P1), P0 → P2 (but not P0 ⇒ P2), as well as P0 ⇒ P1 and P2 ⇒ P3, and so also permits the representation of finite series of propositions like P0 ⇒ P1 ⇒ P2 ⇒ P3, as well as limitless series like. . . ⇒ P0 ⇒ P1 ⇒ P2 ⇒ P3 ⇒ P4 ⇒ . . .26 Since these extra-logical concepts have no content other than the logical relation of hypothetical judgment and problematic logical modality, both cause and effect and possibility-impossibility therefore qualify as purely intellectual concepts. To actualize the possibilities represented in causal relations, we need to do no more than change the modality of the grounding proposition from determination by the pure concept of possibility-impossibility derived from the logical form of problematic modality to determination by the no less purely intellectual concept of existence-nonexistence derived from the logical form of assertoric modality.27 In accordance with the meaning of the pure concept of a cause and effect relation, this will immediately change the modality of the consequent from possibility to actuality as well; and if the consequent is itself the immediate ground of a further proposition, its actualization will actualize that proposition, and so too for any others subsequent to it in a series. Given that the sole and entire content of these concepts—cause and effect, possibility-impossibility, and existence-nonexistence—derives from logical forms of judgments, they require no special faculty of intellectual intuition to access them and so can no more be conceived as consciousness-independent, platonically

190  Nature in Mind: Through Kant’s I’s real representations than the ground-consequent logical relation or problematic and assertoric logical modality themselves—i.e. all alike are as fully explicable psychologistically as the ‘I think’ itself. And the same is true of the entire set of Kantian pure concepts of the understanding: all derive from logical forms of judgment in much the same way, without ever needing to transgress the boundary separating a priori psychologism from platonism. Although my descriptions of the origin of pure concepts of the understanding in the previous examples may seem somewhat complicated, the process is actually sufficiently simple and direct to allow one to suppose that a creature endowed with logical forms of judgment would almost surely arrive at the pure concepts they make possible. For to employ these forms is to be immediately, if darkly, conscious of the degrees of freedom each permits, and so too of the freedom to annul them. By exercising the latter freedom one ineluctably arrives at an equally dark consciousness of pure concepts of the understanding that depends neither on language nor on supersensible access to any platonic realm. Together with the logical forms, they constitute the sole and entire conceptual production of pure understanding. Anything further requires their supplementation by representations arising from other, non-intellectual sources. Otherwise, their only use is to provide new ways to organize the deliverances of associative imagination: a further type of categorization to experiment with, a further type of hypothetical to put to the test, etc. Albeit slight, this might be selective advantage enough to establish an innate tendency to form and use pure concepts of the understanding in creatures possessing the capacity to produce their neural correlates.28 In any case, by contrast with platonism, one is not obliged to posit anything remotely like a miracle to suppose that pure concepts of the understanding could arise naturally and spread in any population of creatures capable of forming the representation ‘I think,’ attaching it to non-conceptual representations to form AUA-concepts, and joining AUA-concepts in propositions by means of logical forms of judgment. Here it might be objected that Kant’s explication of pure concepts of the understanding in terms of logical forms of judgment is itself a version of platonism. After all, Kant’s theory ascribes an innate origin to these forms, and so may be supposed to tell us only how they come to originate in us, not how logical forms themselves originate. Being purely intellectual, with no contribution from sense experience or convention, what is there then to anchor their origin in the subject and prevent one from supposing that they exist consciousness-independently in themselves, be it in platonic heaven or elsewhere? The primary Kantian response is that the logical forms of judgment are essentially bound up with the representation ‘I think’ and the logical universe it makes possible, and so are impossible not just consciousness-independently but apperception-­ independently. But even setting that aside, platonists can take no

The Logical I 191 consolation from the innateness of logical forms of judgment. Platonism is concerned exclusively with representations of objects, be those objects mathematical (numbers, sets), physical (bosons, fermions), chemical (atoms, molecules), geological (minerals, tectonic plates), astronomical (galaxy clusters, the Big Bang universe), moral (virtues, duties), religious (salvation, grace), musical (tones, melodies), or anything else. Pure concepts of the understanding, by contrast, are bereft of objective content, including even the most attenuated, “purely subjective” sort (qualia, feelings).29 They are nothing more than context- and content-indifferent forms whose sole and entire function is to synthesize non-logical psychological representations according to the universality conferred on them by the ‘I think,’ understood as the representation not of a common content but merely of a common consciousness. Being both 100% psychological and 100% non-objective, pure concepts of the understanding are therefore the very antithesis of platonic forms.30

The Gulf Between Pure Concepts and Categories: The Heterogeneity Problem Despite unfounded assumptions to the contrary, ‘pure concepts of the understanding’ are not, for Kant, synonymous with ‘categories.’ The former are in no sense concepts of things or their doings. The sole and entire content of pure concepts of the understanding is logical, and the only representational role they are capable of performing is restricting the logical behavior of AUA-concepts in propositions (e.g. as final subjects rather than subjects than can also serve as predicates) and the logical behavior of these propositions in more complex propositions (e.g. as possible in only one grounding relation to only one other proposition rather than freely combinable with any other). Since everything else in consciousness—­sensation, emotion, desire, memory, imagery, associative relations, etc.—is devoid of logical features and has nothing whatsoever to do with propositions, their logical behavior, and the logical behavior of their AUA-conceptual components, the very idea of combining pure concepts of the understanding with non-intellectual representations therefore seems a nonstarter. The two are simply too heterogeneous. How Kant reckoned with this radical heterogeneity will emerge in subsequent chapters. Here it suffices to recognize that unless and until this is done, it risks making a mockery of his claim that pure concepts of the understanding are categories. Categories have been defined in various ways, but for our purposes they can be thought of as concepts of objects so fundamental that they presuppose no other such concepts while all other concepts of objects presuppose them (mathematical objects such as second derivatives and Galois groups not excepted). Most lists of categories include the relations of substance and accident and cause and effect. As indicated earlier,

192  Nature in Mind: Through Kant’s I’s Kant’s claim was that these categories turn out, on analysis, to be none other than pure concepts of the understanding: substances are final subjects, accidents final predicates, and causality nothing other than fixed unalterable ground-consequent relations. Everything else that tends to be connected with them is inessential, superficial sensory and emotional overlay like the painful smack felt from a blow, the resistance felt when trying to move something heavy, the shocked reaction to a sudden bang, and so on. For if we take away pure concepts of the understanding from the representation of such phenomena, nothing subsistent or efficacious is left to represent, only affects lacking all objective significance. By contrast, if we take away the affects but retain pure concepts of the understanding, we have no difficulty representing what exists as a subsistent final subject with an accidental final predicate that has its necessitating ground in another such accident. The same reasoning applies with respect to discourse regarding substances and causes: take away pure concepts of the understanding and all that remains are conventions of word use that have no more intrinsic objective significance than conventions of highway use or office etiquette. But take away language and its social infrastructure and retain pure concepts of the understanding, there is again no difficulty representing what exists objectively as subsistent final subjects with accidental final predicates causally connected via fixed unalterable ground-consequent relations. And so too for the remaining pure concepts of the understanding found in creatures constituted like us: according to Kant, they provide a means to explicate in purely a priori psychologistic terms the intelligible core of the traditional categories of unity, plurality, totality, reality, negation, limitation, interactive community, possibilityimpossibility, existence-nonexistence, and necessity-contingency, while stripping them of all inessential affective and linguistic baggage. By psychologistically explicating the categories as pure concepts of the understanding, Kant sought not only to free them of all tincture of platonism and linguistic conventionalism, but, above all, to preserve them against the kind of empirical psychologistic reductionism that turns concepts of objects into “nothing but bastards of the imagination which, impregnated by experience, has brought certain representations under the law of association.”31 Hume’s treatment of cause and effect is again the locus classicus. As we saw in Chapter 5, Hume argued that a causal relation in the categorial sense of an objectively necessary connection between distinct existents is quite literally impossible to represent in a single act of thought. This is because distinctness entails that it is possible for either of the relata to exist and the other not, and so contradicts the implication of the category that it is impossible for the one to exist and the other not. Objective necessity is instead representable only via convention—linguistic practices embedded within a social infrastruc­ ture that are fully independent of the conscious goings on in any individual’s psyche. In psychological representation, by contrast, the place of

The Logical I 193 objective necessity is occupied by the ersatz “necessity” of the facility and vivacity affects of maximal intensity incidental to customary contiguity associations (Chapter 4). What Hume seems never to have anticipated, however, is anything like the pure concepts of the understanding of Kantian a priori psychology. The pure concept of a fixed unalterable ground-consequent relation makes it possible to represent in a single act of thought precisely the kind of extra-logical relation Hume deemed impossible independently of language: a necessary connection between items presupposed as distinct. Similarly, the pure concept of a fixed unalterable subject-predicate relation introduces another sort of extra-logical necessity of connection between distinct representations of the kind Hume held to be logically and psychologically impossible in a single act of thought. Or, in Kantian terms, the irreversibility imposed on propositions by pure concepts of the understanding transforms the synthetic unities of consciousness constituted by the freely reversible—hence non-necessary—relations thought through logical forms of judgment into genuinely necessary synthetic unities. And so too the remaining pure concepts of the understanding: insofar as each involves the elimination of the degrees of freedom of which we are conscious in the corresponding logical form, it extra-­logically necessitates one logical possibility to the exclusion of all the others, thereby effecting additional necessary synthetic unities of consciousness, all likewise wholly independent of language, empirical inputs, and mystical platonistic illuminations. There is only one problem. Hume’s denial that a necessary connection between the distinct is logically or psychologically possible outside of language specifically concerns necessary connections between distinct existents. Yet, as Kant was well aware, pure concepts of the understanding are purely formal determinations that apply not to existent objects, actions, and movements but merely to the logical behavior of AUAconcepts in propositions and of these propositions in more complex propositions. The necessary connections represented in them are not real relations between things but merely formal relations between concepts and propositions, and it was never Hume’s intention to claim anything at all regarding the logical behavior of the latter. The burden of proof was therefore Kant’s. He needed to show how pure concepts of the understanding, with none but the purely formal logical meaning conferred on them by the logical forms and the ‘I think,’ and so devoid of all objective content, can psychologistically explicate any concepts of objects at all, much less the most fundamental of all concepts of objects—the categories of traditional metaphysics. Since the a priori psychology of purely intellectual representations alone—the ‘I think’ + logical forms—clearly does not suffice to overcome the radical heterogeneity in content between pure concepts and categories, Kant saw no option but to extend a priori psychologism into an

194  Nature in Mind: Through Kant’s I’s entirely new domain by investigating the conditions under which the ‘I think’ itself first becomes possible.32 In particular, he had to explain what must already be in place in sensibility, ahead of all intellectual representation, in order to bridge the gap between a mind capable at most of association and one capable of pure self-consciousness and the logically structured propositional representations it makes possible. Only when this was done could the path be cleared for Kant to solve the heterogeneity problem.33

Notes 1. CPR B132. 2. Some logicians use the notion of a universe of discourse to characterize what I am terming ‘logical universe.’ But if this presumes that all logically structured thought is ipso facto linguistic, or at least linguistically expressible, then it is too narrow, for on Kant’s account, creatures without language but possessed of pure self-consciousness might still have logically structured thought. To accommodate both non-linguistic intelligent minds and the possibility that language might be as specific to humans as orchestral concerts, the logical universe must be conceived as maximally wide in scope, and so, potentially at least, more universal and abstract than even the universes of mathematics, physics, and all other disciplines limited to specific kinds of content and/or types of representation specific to minds constituted like ours. Even the universe of post-Fregean mathematical logic is not sufficiently general and abstract to count as ‘logical’ in Kant’s sense insofar as it involves denumerable objects, one-to-one correspondence, or anything else implicative of individuality howsoever abstract, since, on his theory, objective individuality requires sensibility no less than understanding (and, where the denumerable is concerned, a sensibility and understanding constituted like ours): see chapters 8 and 9; also KAIM chapters 6-E, 10, and 14-E. 3. CPR B133–4n. Also: “we have concepts only through the unity of consciousness” (AA 18 §5650 [1785–8]), i.e. “the apperception that necessarily underlies understanding and all thought” (PFM 318). See chapter 6 note §19 for other relevant citations. 4. In his logical writings, Kant speaks of making concepts by exercising the psychological operations of comparison, reflection, and abstraction (e.g. L 94–5). These, however, relate not to the logical form of concepts as universals but to the contents thought in them, and more particularly serve for the fabrication of useful concepts. See KAIM chapter 9-B. 5. Kant also held that forming AUA-concepts from associations serves to make one conscious of the rule of the association. Since this concerns the cognitive rather than the logical dimension of empirical psychology, I will discuss it in the course of my examination of Kant’s account of cognitive experience in chapter 10. 6. “That man can have the I in his representation raises him infinitely above all other living things on earth. Because of this he is a person . . . even when he cannot yet speak the I, since he still has it in thought: just as all languages, when they discourse in the first person, must still think it even if they do not express this I-ness through a particular word. For this capacity (namely, to think) is the understanding” (Anthropology, p. 127). 7. Kant considered the possibility of thought a primary philosophical question. See KAIM introduction to part III and chapter 12 for discussion of the

The Logical I 195 investigation to which he consigned it: the subjective transcendental deduction of the categories. 8. CPR B131–2. 9. “[T]he form of the judgment consists in the determination of the way in which distinct representations as such belong to one consciousness” (L102), while judgment itself is “the representation of the unity of consciousness of distinct representations” (L101). Kant’s list of Aristotelian logical forms of judgment was organized into four groups of three: the universal particular and singular quantities of propositions, the affirmative negative and infinite qualities of propositions, the categorical hypothetical and disjunctive relations of propositions, and the problematic assertoric and apodeictic modalities of propositions. See CPR A70/B95 and KAIM chapter 10. 10. See CPR B145–6 and A230/B283. Unfortunately, neglecting or discounting Kant’s openness to the possible phylogeny dependence of logical forms of judgment is all too common among Kant interpreters. 11. I have opted to use ‘proposition’ or ‘propositional thought’ (in distinction from ‘propositional discourse’) in preference to Kant’s somewhat archaic use of ‘judgment.’ The only exception is Kant’s term for logical forms—‘logical form of judgment’—because it relates to a specific component of his a priori psychologism. See also chapter 1 note §4. 12. “Logical form is to the intellectual representation of things precisely what space and time are for the appearances of a thing: namely, they contain the positions for ordering them” (AA 17 §4629 [1771–3]). 13. It has the additional logical significance that a change in the modality of the copula of the proposition in the ground position from problematic to assertoric ipso facto changes that of the proposition in the consequent position to assertoric as well (modus ponens in inference form). Similarly, changing the modality of the proposition in the consequent position to assertoric while simultaneously changing its logical quality from affirmative to negative (or vice versa) ipso facto changes both the modality and quality of the grounding proposition (modus tollens). 14. There is another logical form that combines propositions, the logical form of disjunctive judgment. Since it is only of secondary importance to the logical I, it will simplify matters here to omit it. Interested readers should consult KAIM chapter 10, where I also argue that De Morgan’s laws and other truthfunctional equivalences almost certainly do not apply to Kant’s logical forms on the ground that, by his criteria of the logical, truth-functionality counts as a mathematical construction rather than a merely logical synthesis. 15. The regress of logical forms stops because hypothetical propositions can themselves become the grounds and consequents of other hypothetical propositions. That leaves only a single act of thought since, when this occurs, the modality of the hypothetical propositions in the ground and/or consequent positions is ipso facto changed from assertoric to problematic, so that the only hypothetical proposition that is assertoric is the one in which they are embedded, i.e. the outermost hypothetical proposition. Thus, the logical form of hypothetical judgment really does suffice to synthetically unite all consciousness. Of course, as noted earlier, logical forms cannot prevent nonsense from resulting if the contents combined through them give an absurd result for other than formal logical reasons. 16. A notion that “never occurred to anyone” (PFM 375n)—and still seems not to. 17. Defined above all by Euler but also the Bernoullis, Lagrange, d’Alembert, Clairaut, and Kant’s friend Lambert. 18. See Philosophical Remarks of 1930–1 (Edited by Rush Rhees, translated by Raymond Hargreaves and Roger White, Oxford: Basil Blackwell, 1975) and

196  Nature in Mind: Through Kant’s I’s the 1929 article “Some Remarks on Logical Form” (in Philosophical Occasions: 1912–51. Edited by James Klagge and Alfred Nordman, Indianapolis: Hackett Publishing Company Inc., 1993). 19. The reason the logical form of singular judgments must be omitted here is that there is nothing to distinguish it as purely logically from the logical form of universal judgment: see KAIM chapter 10-B. 20. E.g. Philosophical Investigations, I-265: “Let us imagine a table (something like a dictionary) that exists only in our imagination. A  dictionary can be used to justify the translation of a word X by a word Y. But are we also to call it a justification if such a table is to be looked up only in the imagination? . . . [J]ustification consists in appealing to something independent.— ‘But surely I can appeal from one memory to another. For example, I don’t know if I  have remembered the time of departure of a train right and to check it I call to mind how a page of the time-table looked. Isn’t it the same here?’—No; for this process has got to produce a memory which is actually correct. If the mental image of the time-table could not itself be tested for correctness, how could it confirm the correctness of the first memory? (As if someone were to buy several copies of the morning paper to assure himself that what is said was true.) Looking up a table in the imagination is no more looking up a table than the image of the result of an imagined experiment is the result of an experiment.” Wherever nature fails to supply constraints (fixed rules) but constraints are needed, conventions fill the void, e.g.  the conventions and associated social infrastructure that make time-tables possible. Wittgenstein’s critique of private language is concentrated in Philosophical Investigations, I §§244–315. 21. The kind of psychologism against which Wittgenstein’s arguments are most directly effective, and against which they were most obviously directed, is typified by William James’s empirical psychological analysis of language and its logic. 22. Principles of Human Knowledge, Intro. §14. 23. T 1.3.1 ¶7. 24. See especially Remarks on the Foundations of Mathematics. Translated by G.E.M. Anscombe, Cambridge, MA: MIT Press, 1967. 25. I have adapted the expression from Kant’s use of ‘letztes Subjekt’ in reference to substances at CPR A205/B250. 26. For the sake of perspicuity, I  use “P0 ⇒ P1 ⇒ P2 ⇒ P3” to abbreviate “(P0 ⇒ P1), (P1 ⇒ P2), and (P2 ⇒ P3).” 27. Assertoric judgment is the default modality of propositions: the mere psychological act of forming a categorical or hypothetical relation in thought is ipso facto to assert it. The logical leeway this leaves, and which the pure concept of existence-nonexistence eliminates, is the freedom to make or not make (including unmake) propositions from given logical elements (AUAconcepts in the case of categorical relations and propositions formed from AUA-concepts in the case of hypothetical relations). Accordingly, when the pure concept of existence-nonexistence annuls this freedom, the modality of the proposition is either irreversibly fixed as assertoric (existence) or irreversibly fixed as not assertoric (nonexistence). See KAIM chapter 10-D for detailed analysis. 28. In chapters  9–10, in the wake of transcendental synthesis, it will become clear that and how their adaptive potential is in fact huge. 29. Recall the discussion of attenuated objects in chapter 6 and the purely epistemological sense in which alone they count as subjective. 30. In contrast to logical forms of judgment, pure concepts of the understanding do not even count as innate. For only via experience of operating with logical

The Logical I 197 forms to synthesize propositions can one discover the possibility of restricting the freedom each provides, and thereupon contrive to form AUA-­concepts by incorporating the ‘I think’ into representations of those restrictions. Since nothing empirical (convention included) is involved in a consciousness limited to the formal logical dimension of proposition formation, there is nothing to compromise either the purity or the purely intellectual character of the resulting AUA-concepts. Accordingly, Kant saw fit to borrow a notion from natural law to characterize pure concepts of the understanding as “original acquisitions” (On a Discovery, pp. 221–3). 31. PFM 258. 32. This is the question Kant addressed in what he termed the subjective transcendental deduction of the categories: see CPR xvi-xvii and analysis in KAIM introduction to part III and chapter 12. 33. Even many otherwise careful Kant scholars construe ‘pure concept of the understanding’ as synonymous with ‘category’, and so overlook Kant’s need to prove identity of meaning in the case of two such manifestly heterogeneous kinds of concept. Only when it is recognized that this need exists, and moreover that only psychologistic explication can meet it, does it become apparent that pure concepts of the understanding do not definitively earn their designation as categories until their capacity to underwrite principles of pure understanding becomes realizable via transcendental schemata and their combination with empirical synthesis in transcendental judgments (chapter 10).

8 The Aesthetic I

Transcendental Aesthetic If the ‘I think’ (cogito) is understood as a universal self-consciousness common to every possible representation, then the ‘I exist’ (sum) entails the existence of a prior consciousness that contains every possible representation individually within it. For how else could one represent one’s consciousness (as opposed to any consciously represented content) as common to any representation if that representation were nowhere to be found within it, i.e. if we were not even darkly conscious of it? Since a representation not even darkly present in my consciousness is ipso facto nothing for me, it is impossible for me to represent my consciousness as common to that representation. Yet, if I  can represent a consciousness as common only to that which is already, ahead of that representation, present within it, then the same is no less true of the ‘I think.’ For how else could the contentless consciousness ‘I think’ be represented as a consciousness common to all possible representations if the totality of representations, possible no less than actual, were not already, ahead of its representation, present in it? It therefore seems undeniable that the universality Kant credited to the ‘I think’ is possible only given a prior consciousness in which the totality of possible representations is immediately contained, if only synthetically—a consciousness that is individual in the same purely formal sense the ‘I think’ is universal—its inverse, as it were. This recognition led Kant to regard the two species of consciousness— consciousness as an individual many-in-one structure (many represented as contained in one) and consciousness as a universal one-in-many structure (one represented as common to many)—as essentially complementary.1 In Kantian parlance, the relation of the universality distinctive of the ‘I think’ to a pure, purely formal individuality of consciousness able to encompass the infinite totality of possible representations is analytic.2 This means that the supposition that a creature could represent its consciousness as logically universal, and so coincident with the logical universe, in the absence of such individuality is quite literally a ­contradiction—just as, conversely, it is a contradiction to suppose that

The Aesthetic I 199 a creature could have its consciousness be such an individual and yet be incapable of forming the representation ‘I think.’ This analytic relationship is the cornerstone of Kant’s account of self-consciousness and goes to the heart of the supreme principle of his a priori psychologism. But what does it mean? Kant’s thesis might at first seem similar to Leibniz’s panpsychic conception of minds as individual substances, or “monads,” that contain the infinite totality of possible representations immediately within them, differing only by the degree of conscious access each has to them. Thus, in any monad, consciousness of one and the same representation may be more or less dark, with no two monads coinciding perfectly in their level of conscious access to all representations, thereby individuating monads as different perspectives on one and the same infinite, all-encompassing representational universe. At the base of Leibniz’s hierarchy are monads that have solely dark consciousness; these are the “minds” his panpsychism ascribes to particles of inanimate matter. Above them are monads with some degree of light amidst the darkness (organisms up to plants), next those with significantly more and better access consciousness (animals), then those with access sufficient for the kind of self-consciousness exemplified by Descartes’s cogito (persons), after which come those in which the balance of access over dark consciousness far exceeds ours (angels), and finally, at the apex, a monad with no darkness in its consciousness of the representational universe whatsoever (God). While Leibniz’s monadology may strike us as outlandish, the case in its favor is more compelling than might initially be supposed. It should be understood to rest on two premises, neither of which is especially easy to refute: that the ‘I think’ (cogito) and its existence (sum) are indubitably certain, and that the existence of the ‘I think’ analytically entails the existence of a consciousness that individually encompasses the infinite totality of possible representations. Put these together and the existence of a representationally all-encompassing individual consciousness cannot be less certain than the existence of the ‘I think’ itself. So, if, like Leibniz, one knows of no alternative to substance metaphysics whereby to explain the possibility of the ‘I think,’ then the conclusion becomes inescapable that the ‘I’ is a self-subsistent infinite individual consciousness that contains the infinite totality of possible representations immediately within it. And, in that case, what other than differing distributions of brightness and darkness in the consciousness of those representations remains to distinguish one soul substance from another? The challenge facing an anti-platonist like Kant was to find a nonmetaphysical, purely psychologistic explication of the two, analytically related sides of self-consciousness, individual and universal. The first thing to note is that he could not have had recourse to substance metaphysics even if he wanted to. Whereas Descartes, Locke, Leibniz, and others viewed the human mind as simply one among innumerably

200  Nature in Mind: Through Kant’s I’s many other existents to be explained by means of substance, causality, and other metaphysical categories, Kant’s post-Humean psychologistic explication of the categories via pure concepts of the understanding makes that impossible. For if, as Kant maintained, the pure concepts presuppose logical functions of judgment that in turn presuppose AUAconcepts and the logical universe made possible by the ‘I think,’ then the categories have to be explicated by means of the ‘I’ and not vice versa. To do otherwise is to proceed in a circle: “the subject of the categories can obtain no concept of itself as an object of the categories from the fact that it thinks, for in order to think these categories its pure self-consciousness, which was to be explained, is presupposed as their ground.”3 But if substance, causality, et al. are unusable for representing the ‘I think’ because they presuppose it, they can be no less useless when it comes to explaining the individual consciousness that the ‘I think’ itself presupposes. Yet, with metaphysical categories excluded, how can that consciousness be explained at all—the monad-like individual consciousness whose existence the I think’s own existence (sum) entails? To ascertain the kind of answer required, we need to bear in mind that, for Kant, the ‘I think’ is the source of the most fundamental logical representation of all: universality. As that which is alone responsible for converting non-concepts into concepts, this universal self-consciousness is a constitutive logical condition of all concepts as such (Chapter  6). Moreover, in demarcating a logical universe divisible according to the contents of these concepts, it adds something to their representation that cannot be brought to consciousness otherwise than in propositions, so that the ‘I think’ constitutes a logical condition of propositional thought as well (Chapter 7). But if the possibility of both conceptual and propositional consciousness—all logically structured thought as such—devolves directly upon the ‘I think,’ then its inverse—the pure, formal individuality of consciousness that precedes and makes possible the ‘I think’ itself— cannot incorporate anything in the least universal into its representation, including AUA-concepts, logical forms of judgment, and a fortiori the categories. Accordingly, Kant classified it under the same general rubric he did all pre-logical representation, from sensations to associations in imagination: sensibility. Kant’s term for the theory of sensible representation was aesthetic, which he used in the original Greek sense of αἰσθητικός (aìsthetikós), ‘of perception by the senses.’ Since the individuality of consciousness presupposed by the universality of the ‘I think’ is purely formal, and so pure, he treated it separately from the empirical theory of sensible consciousness under the rubric transcendental aesthetic, which is simply the pre-logical, purely sensible portion of a priori psychologism. Kant was well aware that its topic—the pure sensibility he supposed to exist in addition to the empirical kind—was entirely novel (which it remains today). Nevertheless, far from downplaying it, he highlighted it as the defining element of his

The Aesthetic I 201 philosophy, not only as that which makes all the rest possible, but as the feature that differentiates it most fundamentally from all other philosophies. Since making the ‘I think’ possible is merely a consequence of Kant’s novel aesthetic doctrine, however, I shall first consider it in its own right, as the centerpiece of a new, radically original theory of sensibility, and only then, in this light, return to the question of its relation to the ‘I think.’

Apprehending Sense-Divide Transcending Appearances in Intuition Kant’s theory of sensibility is best approached from the standpoint of Berkeley’s theory of vision since both have effectively the same starting point. Berkeley, it will be recalled, challenged the common assumption that vision and touch (broadly construed) both give direct access to space. For how can that be the case if sight gives only sensations of color and light but none of feeling, and touch gives only feeling and not the least sensation of color or light? If the two species of sensational consciousness have no inputs in common, how can they both yield sensations of space (spatiality, the spatially determinate)? Berkeley concluded they could not, settled on touch as the unique source of genuinely spatial sensation, regarded sight and the other senses as spatial only in a secondary, merely illusory sense, and so attributed both the multi-sensory external sense and its sense-divide transcending objects to associative imagination rather than to sense proper. Kant took a different tack, denying that sensation is ever spatial, or capable of giving us the least notion of space, tactual not excepted. One reason is that tactual sensation is merely an accident of sensational phylogeny. Had evolution proceeded differently, or if we were the creation of futuristic genetic (or another kind of) engineering, we might well have had a sense of bodily spatiality in which tactual sensations did not figure at all. Berkeley would balk at allowing it to be termed “spatial,” but only because he equated spatiality to feeling. Kant did not, and saw no way of avoiding the conclusion that a space accessible through (at least!) two such heterogeneous modes of sensation as touch and sight cannot itself be sensation. This does not mean our non-sensational consciousness of sense-divide transcending space is qualitatively identical with bodily spatiality itself, or physical spatiality generally. As psychologically real, the space present in consciousness must, like tactual sensations, have its own neural correlates, and so is as much a total informational blank slate where its neural correlates, their causes, and the body itself are concerned as they are. Indeed, Kant’s view is perfectly compatible with there being creatures with neural correlates for a non-spatial non-sensational representation that performs the same role in the representation of sensedivide transcending objects that space does in us. What matters at the

202  Nature in Mind: Through Kant’s I’s more general level of a priori psychologism (Chapter 7) is not space as such but the purely psychological role it performs in enabling heterogeneous sensations of different senses, be they human or not, to all be exhibited in consciousness via a single, sense-divide transcending manifold of homogeneous non-sensational representations, be they spatial or something else. Supplemented by empirically discovered correlations between sensations, the spaces that non-sensationally exhibit varying sensations of the same sense at different times and different senses at the same time can be associated and, in some cases, identified. For example, if a visual sensation becomes associated with a tactual sensation, the non-sensational homogeneous spaces exhibiting these heterogeneous sensations may be set in immediate juxtaposition, or even identified as one and the same sense-divide transcending space accessed via two different senses. That this is impossible to do with the sensations themselves creates a problem only for purely sensationalist theories like Berkeley’s. For whereas yellow and blue are homogeneous to the extent they can be apprehended in immediate juxtaposition, the feeling of an itchy scalp is too heterogeneous to be juxtaposed with either; and whereas the feelings of an itchy scalp and a feverish forehead can be apprehended in immediate juxtaposition, yellow is too heterogeneous to be juxtaposed with either. And while empirically discovered correlations can indeed lead us to associate visual with tactual sensations, they can never completely blind us to the heterogeneity of the sensations themselves. Berkeley conceded these points, but argued that association exerts so strong an influence on our imaginations as to create an illusion of visuotactual homogeneity. This, however, faces the problem that even if the association is the strongest imaginable, it can no more disguise their radical heterogeneity than the almost equally close association in the minds of native speakers of English between the word ‘dog’ and actual canines disguises theirs. So, even if there were anything like an “illusion” of visuo-tactual homogeneity, it can have nothing to do with the closeness of their association. Kant’s view was that far from it being an illusion, we have consciousness of a fully real homogeneous manifold of non-sensational spaces. The same non-sensational capacity whereby homogeneous spaces exhibit heterogeneous tactual and visual sensations a priori can then proceed to relate or identify those spaces whenever observed correlations between the sensations warrant it a posteriori. It can do the same with sounds and smells: after exhibiting them spatially a priori and observing correlations with visual and/or tactual sensation a posteriori, their spaces can be related or identified as well (wholly or in part, depending on their congruence, overlap, or containment). In this way, the heterogeneous data of the external senses, be they the human complement or another, can all be used to apprehend one and the same non-sensational homogeneous

The Aesthetic I 203 manifold of spaces in intuition, as Kant termed it, including individually the same space—a space that, as accessible through multiple senses (e.g. seen, touched, and heard), constitutes a genuine (i.e. non-illusory) sense-divide transcending object, or appearance, e.g. the corncob I taste, smell, feel, hear, and see as I eat it. These non-sensational sense-divide transcending objects do not have to be spaces in every sensible being, but whatever they are, it is essential that they be usable to exhibit heterogeneous sensations homogeneously, in precisely the manner necessary to fuse multiple external senses (kinds of sensation) into a single external sense (intuition) related to one and the same manifold of sense-divide transcending objects (appearances). Thus, far from being themselves appearances as Hume and nearly all other early modern philosophers supposed, sensations, for Kant, are never more than the medium through which appearances—be they spaces or something analogous—are apprehended in intuition.4 Kant granted that there may well be creatures that lack the capacity to produce spaces or anything analogous whereby to exhibit their heterogeneous sensations homogeneously. But in creatures able to do so, he held that sensibility must be understood as a dual process in which a non-sensational space is apprehended in tandem with every sensation. Apprehension is sensible for the same reason sensation is: none of the appearances we are conscious of in it have logical form, much less logical structure, and are generated prior to and independently of the representation ‘I think.’ So far as introspection is concerned, it is also as automatic a process as having sensations, occurring, like sensation, in dark consciousness with no need for discernment or attention, much less association or relation of any kind. Indeed, it cannot be emphasized too strongly that Kantian non-sensational exhibition in apprehension is directed at the most primitive level of sensible mentation—termed synopsis—before sensations are reproduced in images, discerned, compared, correlated, related, ordered, or organized in any way, much less intellectualized via AUA-concepts and propositions. Since this makes apprehension the ­second-most primitive level of sensible consciousness after synopsis, it is crucial never to forget that the non-sensational spaces apprehended in intuition, just like the sensations they exhibit, are “met with in the mind scattered and single in themselves.”5 Intuitions are presumably able to pull their evolutionary weight (via their neural correlates) through their role in facilitating consciousness of sense-divide transcending objects. We do not have to systematically master a complex associative visual language of tactual spatiality in order to represent non-tactual sensational proxies of tactual objects, “spatialized” by a kind of illusion. Instead, Kantian transcendental aesthetic posits that we directly apprehend genuinely sense-divide transcending spaces, through all our senses, by the simple act of exhibiting them nonsensationally in intuition, with no need for recollection, discernment,

204  Nature in Mind: Through Kant’s I’s attention, comparison, or any of the other complex mental activities posited by Berkeley’s associationist solution to the problem of representing sense-divide transcending objects through a single, unified external sense. Only afterwards does experience play its role in the organization of the spaces that exhibit sensations in intuition, enabling us to distinguish spatially significant from spatially insignificant sensations, correlate or identify them across sensory fields, and thereby apprehend sense-divide transcending objects empirically in intuition—the blue-wet expanse, the noisy-smelly confine, the squishy-sweet surface, etc. The phylogenetic relativity of everything that enters into Kant’s transcendental aesthetic cannot be emphasized too strongly. It is perfectly compatible with there being creatures that have completely different senses from ours that still exhibit their no less heterogeneous sensations through the same external intuition we do: non-sensational space. Their sensation of bodily spatiality, for example, would be non-tactual but just as fit as tactual sensation to be exhibited through spaces that could then be represented in immediate juxtaposition with the homogeneous spaces that exhibit the sensations of their other senses. But, again, this does not mean that intuition of the external world is necessarily spatial: Kant’s theory is equally compatible with there being creatures with the same senses we have but with a different non-sensational representation = X, entirely unknown to us, whereby to exhibit them in external intuition. Rather than tactual bodily spatiality, such creatures would have tactual bodily X-iality, which could then be represented as homogeneous with the X’s that exhibit visual, auditory, olfactory, and gustatory sensations. To be sure, Kant did not affirm the possibility of some such X. He merely insisted that it cannot be precluded, which, in contemporary terms, is as much as to say that non-sensational sensation exhibitors like X may exist, if not terrestrially then extra-terrestrially, if not in nature then via futuristic genetic (or other) engineering. Thus, while space may be non-sensational on Kant’s conception, there is no reason to regard its neural correlates as any less dependent on phylogeny than the neural correlates of sensations, albeit at a less primitive stage of the evolution of consciousness.

Imagination Reconceived as Productive: Pure Formal Intuition Kant ascribed the appearances generated to exhibit sensation to the imagination and its synthesis of apprehension in intuition. Previous philosophers assumed that sense perceptions constitute a single, homogeneous manifold of consciousness—a kind of “meta” sensory field uniting all sensations howsoever heterogeneous (the “common sense” of Aristotelian and Medieval psychology)—prior to and independently of all exercise of imagination. They thereby tacitly assigned the synthesis of

The Aesthetic I 205 apprehension to sense alongside, and in addition to, what Kant termed “the synopsis of the manifold a priori” whereby sensations are given but without any consciousness to unite them as a single manifold, synchronically or diachronically.6 Kant was therefore nothing if not correct when he averred that “[n]o psychologist has ever so much as even thought that the imagination might be a necessary ingredient of perception itself.”7 Without providing evidence, much less proof, his predecessors simply took for granted that the imagination is limited to replicating previously apprehended sense perceptions in thought, rearranging them to form new combinations that could be sensed but happen not to have been, and, in its most important role, associating images in thought to sense perceptions in accordance with innate psychological principles à la Hume. So conceived, it is a purely reproductive faculty like memory, not one that is originative of new, non-sensational representational content.8 Kant’s thesis that imagination is essential to sense perceptions themselves thus broke with precedent in two ways: by making it active already at the level of their apprehension in a single, purely sensible consciousness as a manifold, ahead of all reproduction and association, and by deeming it productive of the, e.g., spatial homogeneity of apprehended appearances themselves. Kant took imagination to be productive in the same way he did the understanding: it yields forms for synthesizing representations that otherwise could not be combined in a single act of conscious representation. In the case of the logical forms of judgments, the inputs are AUAconcepts, and propositions are the representations that synthetically combine them—thereby also combining, albeit at one remove, the sensations and other representations that supply AUA-concepts with their content. In the case of purely aesthetic forms like space, the inputs are sensation-exhibiting appearances, and intuitions are the representations that synthetically combine them—and so too, at one remove, the sensations these appearances exhibit.9 There, however, the resemblance ends. Propositional form is directed exclusively toward the logical universality contributed to its constituent representations by the ‘I think’ whereby alone they become denizens of a properly logical universe. Since universality is a one-in-many (one represented as common to many) structure, propositional combination does nothing to determine whether the contents of the AUA-concepts it combines have, in addition, the many-inone (many represented as contained in one) structure distinctive of the individuality of intuition, and works equally well one way or the other. With intuitional form, it is just the reverse. Intuition relates exclusively to the homogeneity of non-sensational, purely aesthetic appearances that enables them to be set immediately together, many within one, whether by juxtaposition or some X-taposition, and without regard to the heterogeneity of the sensations they exhibit. It does nothing to determine whether the appearances it individually contains have, in addition, the

206  Nature in Mind: Through Kant’s I’s one-common-to-many logical universality of the ‘I think’ incorporated into their representation, and works equally well one way or the other. The formal individuality of intuitions means that any appearance can be set immediately together with any other in an intuition that can itself, in turn, be set immediately together with another in a still more comprehensive intuition, and so on without end. Conversely, if the exhibition of sensation requires that an appearance be differentiated within itself, we need only apprehend it in an intuition that represents it as a composite synthesized from intuitions that may each in turn be represented as composite, and so on as many times as is required to exhibit in imagination everything present to consciousness in sensation. Thus, even if we had senses capable of representing the greatest and slightest sensational differences in response to stimuli at every scale and dimension (including, e.g., the most telescopically powerful and microscopically acute sight possible), our productive imagination would still have the capacity to exhibit every one of them via non-sensational appearances all contained immediately together as the manifold of a single, individual intuitive consciousness. What need do we have for a productive imagination of such vast synthesizing capacity? Take, for example, visual sensations and their exhibition by spatial appearances. Human visual sensation is composed of differences of color down to the minutest shades of the same color, glossiness, brightness, and no doubt other qualities for which names are lacking. Some visual differences we are able to discern and focus attention on, while others are too subtle, complex, or otherwise unsuited for access consciousness (e.g. the individual grains of sand of the beach I am jogging on). But whether confined to dark consciousness or not, only some of these will correlate systematically with tactual and other nonvisual sensational differences, and so contribute material for representations of sense-divide transcending spatial objects. The problem is that the imagination’s spatial exhibition system cannot determine a priori, ahead of actual experience, which visual differences do and which do not have non-visual correlates, and, of those that do, to which particular non-visual sensational differences they correlate and how (tactual? olfactory? sensations unknown to humans?). To wait on experience to distinguish signal from noise, however, would largely defeat the purpose of non-sensational exhibition, which is to facilitate spatial association and identification between even the most heterogeneous sensations (there is a further problem relating to the temporal side of correlation that will be considered later). For this reason, productive imagination’s spatial exhibition system must produce non-sensational appearances for all sensational differences indiscriminately, without regard to their role (if any) in the representation of sense-divide transcending objects, or even whether they are discernible and capable of being attended to. If, as with human gustatory sensations, the kind of fine differentiation needed for correlation to non-gustatory sensations is wanting, along

The Aesthetic I 207 with the ability to individually discern and attend to such differences as may be present, then the spaces that exhibit them can be disregarded. If, as with differences of human auditory sensation (by contrast with the auditory systems of bats and cetaceans), correlations are few and discernment and selective attention limited, the multitude of leftover spaces can also be disregarded. But if, as with differences of human tactual and visual sensation, correlations are fine-grained and systematic, nearly all of the spaces produced to exhibit them can be retained as inputs for aesthetic organization in empirical associative imagination and logical organization in propositional understanding. For, on Kant’s theory of pure sensibility, what is essential is that wherever correlations exist and representations of sense-divide transcending objects are possible, there must be spaces already apprehended a priori in intuition available to be reproduced, discerned, attended to, compared, correlated, associated, conceptualized, and judged a posteriori.10 To achieve such a synthesis of apprehension, the form productive imagination uses for exhibiting sensations, whether space or some other, “must lie ready for the sensations a priori in the mind.”11 This means that it is possible to become conscious of the form and its potential to exhibit sensations in appearances entirely a priori in what Kant termed a pure, or formal, intuition.12 It will be recalled that a representation counts as a priori for Kant if having it requires no awareness of the particular quality of any sensation (red, loud, acrid, etc.) or whether any particular sensation ever actually exists. Formal intuition is exactly such a consciousness: it represents an appearance-synthesizing form that, lying ready in the mind, is the same for all sensations, regardless of both their particular qualitative content and which do or do not actually come to exist in one. Since the synthesis represented in it can generate appearances (e.g.  spaces) within and without one another without end, formal intuition is sufficient to exhibit any and all sensations, human and non-human. This makes formal intuition an individual consciousness capable of accommodating within it the infinite totality of possible sensations, however heterogeneous, by means of homogeneous appearance proxies. More precisely, formal intuition is a representation of the form that enables distinct non-sensational appearances to be synthesized that admit of being set immediately together (e.g.  by juxtaposition) within the same intuitive consciousness to any iteration—appearances within and without appearances without end. So, rather than a literal unity, formal intuition is merely a synthetic unity of the manifold of appearances and, indirectly, of the sensations they exhibit, in the same way that logical forms of judgment are synthetic unities of AUA-concepts and the judgments formed from them (Chapter 7). The only difference, albeit a crucial one, is that the synthetic unity effected by logical forms serves to confer universality on the ‘I think,’ whereas the synthetic unity effected by aesthetic forms like the pure formal intuition of space is the original

208  Nature in Mind: Through Kant’s I’s unity of sensibility itself: an individual consciousness capable of containing the totality of possible sensations within it, even if only indirectly, at one remove, via the non-sensational appearances that can be synthesized to exhibit them.13

Aesthetic Space I cannot emphasize too strongly how important it is not to attribute more to formal intuition of space than transcendental aesthetic can support. In particular, Kantian formal intuition must not be conflated with objective space, much less Euclidean space, as unfortunately almost always happens. In a priori psychologism, purely aesthetic formal intuitions like space account simply and solely for the synthetic unity of sensibility—the individual consciousness in which the totality of possible sensations is indirectly represented—that precedes and makes possible the universality of the ‘I think.’ The universal self-consciousness ‘I think’ in turn makes possible AUA-concepts that, together with logical forms of judgment, yield propositional representations. These logical forms are in turn the source of the pure concepts of the understanding that, for reasons that will emerge in subsequent chapters, Kant equated with categories, or fundamental concepts of objects. On his view, then, there can be no objectivity apart from the pure concepts of the understanding, and that includes objective spaces, be their representation transcendental, mathematical, or empirical. Since aesthetic formal intuition must already be present in the mind before the ‘I think’ is even possible, much less pure concepts of the understanding, the purely aesthetic formal intuition of space must be carefully distinguished from all representations of objective space. In particular, in contrast to objective transcendental, mathematical, and empirical representations of space, formally intuited space is altogether lacking in fixed, determinate order and relation, and so devoid of everything that can serve to objectively individuate one spatial appearance from any other, much less every other, possible no less than actual. For example, in order for the space of conscious representation to qualify as Euclidean, not only pure sensibility’s formal intuition of space is required but also all the AUA-concepts (of points, lines, etc.) requisite to form the geometrical propositions—definitions, axioms, postulates— without which it is impossible to represent space as Euclidean. Even setting aside their dependence on language, concepts such as these are impossible, according to Kant, apart from pure concepts of the understanding, and so too apart from both the logical forms of judgment from which these transcendental concepts derive and the ‘I think’ on which AUA-concepts generally depend. Since none of these are possible unless and until aesthetic formal intuition is already in place, none can be supposed to enter immediately or essentially into the latter. Thus, the formal intuition of space is not only neither Euclidean nor non-Euclidean

The Aesthetic I 209 but completely indeterminate as regards number, limit, distance, metric, part–whole relationships, and everything else that makes space suitable for properly mathematical representation or objective representation of any kind.14 What is objective space for Kant, fundamentally and essentially? To answer, we must forget all we know about space from mathematics, physics, and experience generally, and focus exclusively on its transcendental (i.e. a priori psychologistic) meaning. For our purposes, objective transcendental space may be thought of as an all-encompassing infinite network of possible spaces, each of which is uniquely differentiated from and completely determined in relation to every one of the infinitely many other spaces, at every possible scale and in all possible (even infinite) dimensions. To say that each is “uniquely differentiated” and “completely determined”15 means that any space we represent is determinately at such-and-such remove from every other, being either adjacent to it in some dimension or, if not, then with such-and-such other spaces intervening on any particular route from the one into the other. Where such determination is complete (i.e. nothing left undifferentiated or indeterminate), each space has a unique set of relations to other spaces that individuates it with respect to each and every one of the infinitely many others fully objectively, i.e. entirely independently of which sensations (if any) it happens to exhibit. By contrast, in purely aesthetic formally intuited space everything is completely undifferentiated and indeterminate. Its spaces have no order or relation, no points, lines, or limits of any kind, no distances, directions, dimensions, or objective differentiation or determination of any kind. Nor can they derive these from the sensations they exhibit since, on Kant’s theory of sensibility, not even tactual sensations are in any intrinsic sense spatial at all (they are so only at second hand, via their neural correlates, regarding which they are complete informational blank slates). This is not to deny that sensations can be used to mark out spatial differences subjectively, as when one distinguishes the green field from the blue sky, the warm water from the cold air, or the noisy house from its silent surroundings. But this falls far short of the kind of objective individuation that would allow represented spaces to retain their identity regardless of which sensations they exhibit or whether they exhibit any all. Accordingly, formally intuited space, considered in its own right, is nothing but structureless, featureless juxtaposition within and without juxtaposition, all the way up and all the way down. It is purely sensible, aesthetic, and imaginary, whereas the transcendental space in which each space is objectively individuated with respect to every other is the result of the determination of formally intuited space by the understanding and its pure concepts. Given that these concepts can determine nothing except the logical behavior of the components of propositions (Chapter  7), transcendental space is possible only insofar as the manifold of

210  Nature in Mind: Through Kant’s I’s formally intuited space is first converted into AUA-concepts that are then combined into propositions that can in turn be determined conformably with pure concepts of the understanding via transcendental synthesis (Chapter  9). Thus, in contrast to the purely aesthetic formal space of productive imagination, which is devoid of all logical determinateness and propositional structure, objective transcendental space, for Kant, is nothing more or less than a propositional network.

Time and Internal Sense Kant distinguished sensations into two kinds, external and internal.16 The former differ from the latter not by virtue of being spatially e­ xternal—a feature restricted to non-sensational appearances—but merely insofar as their presence in consciousness requires no conscious mental activity of any kind, be it recollecting, imagining, discerning, or any other. Why term them ‘external’ in that case? Because their presence is completely independent of the representational activity of the conscious subject, their existence counts as subjectively unconditioned. To that extent, each such sensation is indicative of an existence external to even the darkest consciousness, whether represented in the guise of something in the brain, something beyond the brain, or the something  =  X that Kant termed, alternatively, thing in itself or transcendental object, regarding which we lack the capacity to form even the most basic representation (Chapter 10). It just so happens that in beings whose sensibility is constituted like ours, the formal intuition whereby heterogeneous external sensations are exhibited in intuition as homogeneous non-sensational appearances is space. But this should not obscure the fact that, in Kant’s theory of sensibility, external sensations, tactual included, have and can have no intrinsic spatiality. Fortunately, there is no similar ambiguity in the case of internal sensations. Our “several internal senses”17 furnish sensational consciousness of our own mental activity and passions—perceiving, recalling, discerning, attending, imagining, emoting, desiring, resolving, et al. Here there can be no danger of confusing internality with spatial internality. Internal sensations are merely the various ways in which the mind is constituted to affect itself through its own conscious operations, and so are “internal” only in the sense of being indicative of no existence other than the conscious mind itself. To be sure, like all sensations, they involve neural correlates. But the brain’s organs (neural correlates) of internal sensation register only what the brain itself originates in its capacity as the organ of conscious representation generally. If we lacked these sensations, we would therefore be entirely oblivious to everything going on in our minds, and so would also lack the capacity to remember, associate, conceive, or otherwise represent these goings on (only self-awareness in the guise of what Kant termed spontaneity would remain).18

The Aesthetic I 211 Internal sensations of the several internal senses should not be supposed to be any less heterogeneous than external sensations of the external senses: internal sensational consciousness of a perceiving is nothing like internal sensation of a remembering, that of a remembering nothing like that of a fantasizing, fantasizing nothing like believing, believing nothing like emoting, emoting nothing like desiring, desiring nothing like resolving, resolving nothing like forming a concept or propositional representation, and so on for however many non-overlapping internal sensory fields exist in a creature’s internal sensibility overall. Nor is their heterogeneity limited to quality. Insofar as intersubstitutability fails to hold, internal sense fields may be regarded as non-overlapping in just the way the external fields of vision and touch are non-overlapping. For just as a yellow external sensation between a red and a blue can be displaced by sensation of another color but not by tactual feeling, smell, or any other non-visual external sensation, the emotion of anger can be displaced in internal sense by a different emotion (relief, embarrassment, regret, etc.) but no more by a believing, desiring, resolving, or other non-emotive internal sensation than by an external sensation of color or smell. The admission of multiple internal sensory fields, however, raises the same problem that arises with multiple external ones, viz. how, notwithstanding their heterogeneity, all can converge on one and the same sense-divide transcending internal object. It is the problem of the unity of internal sensibility. To be sure, there is already something paradoxical about the notion that we can be conscious of our own varied conscious activities only insofar as we are passively affected by them in manifold ways in internal sensibility, with each internal sensation having its own dedicated neural correlate. But without a synopsis of sense capable of individually registering and differentiating the various conscious doings taking place at any instant—apprehending, remembering, imagining, discerning-attending, comparing-relating-associating, conceiving-judging-inferring, emoting, desiring, resolving, etc.—there would be no way to apprehend them (even if only darkly) as a manifold in internal intuition, much less reproduce the apprehended manifolds of different instants in order to represent the present one as their temporal successor. In short, Kant, like Hume and others before him, saw no alternative but to conclude that humans are examples of a kind of mind that is naturally as sensitive internally as it is externally (with touch—broadly construed to comprise sensation beneath as well as in the skin—classed by Kant as a component of the multi-sensory external sense that is nevertheless interior to the body, i.e. the sensus interior of the sensus externus as distinct from the properly internal sensations of the sensus internus).19 Kant nevertheless broke with his predecessors’ theories of internal sensibility by extending to internal sense his psychologistic explication of external sense in terms of pure formal intuition. Time and

212  Nature in Mind: Through Kant’s I’s time-consciousness are the focus of his divergence. Hume, for example, built his account from the premise that consciousness exists only in the present instant, so that once that instant has passed, it is “impossible to recall.”20 This is, of course, not to deny that it can be remembered. Memory, however, is not direct perception of the past in the way sensation constitutes direct perception of the present, but rather something thought in present consciousness that represents something else, distinct from it, that was present in a now past, impossible-to-recall consciousness. Since this implies that time-consciousness’s own existence, like that of every other consciousness, is confined to the present, the succession of consciousnesses that constitutes the life of the mind can never itself be witnessed; and since what cannot be witnessed in internal sense also cannot be recollected in memory or conjured up in fantasy, the succession of consciousnesses turns out, strictly speaking, not to be a possible conscious representation at all. So, to explain how this (or any) succession can nevertheless be represented instantaneously, Hume saw no option but to suppose that internal sense (1) provides us with an internal sensation, or perception, of all the representations in the consciousness of that instant as a manifold, (2) a perception that, at the same instant, can be represented together with a manifold of recollections of previous such perceptions, (3) all under the form of a succession, (4) the components of which can then be temporally ordered and related by associative means.21 Not being based on anything directly witnessable in the real succession of consciousnesses, the temporal “manner in which different perceptions make their appearance”22 may be supposed to have the same kind of relation to real physical succession that tactual bodily spatiality has to real bodily spatiality: just as there is no saying that real physical bodily space (whose ultimate description is today given by particle physics) is anything like—in any way homogeneous with—the bodily spatiality of tactual sensational consciousness, the internally sensed temporality (“manner of appearance”) that enables us to instantaneously represent the succession of consciousnesses that constitutes the life of the mind may be completely heterogeneous with their real physical  temporality (whose ultimate description is today given by relativity theory). A  different phylogeny or futuristic genetic (or other) engineering might very well have resulted in internal perceptions incommensurable in manner of appearance with ours that correlate no less systematically with the real physical temporal succession of consciousnesses (i.e. of their neural correlates). Moreover, just as perceptions are informationally blank slates regarding their neural correlates generally, perceptions of succession are informationally blank regarding their neural correlates’ temporal reality, in relation to both their predecessors and successors in the brain and their temporal relations to physical reality generally. The implication is that the real (unwitnessable) successively unfolding life of the mind may be nothing like our always instantaneous, informationally blank internal

The Aesthetic I 213 subjective temporal experience of it. Indeed, they may be as completely unlike as physicists suppose when they equate physically real temporality with the fourth dimension of the spacetime of a “block universe” that betrays not the least trace of an ever-vanishing present or any of the other features distinctive of internally sensed subjective time. While there is no evidence that Hume went quite this far, Kant most certainly did. He recognized that if the time of internal sense experience is something dependent on the peculiar, phylogeny-dependent constitution of the self-affecting conscious subject, “the subject in which the representation of time originally has its ground cannot determine its own existence in time by means of that representation.”23 Indeed, he even went so far as to maintain that if one could represent one’s consciousness through a differently constituted sensibility or without sensibility of any kind intervening, “the representation of time . . . would not occur at all.”24 Kant took this view because he subjected time—just as he did space, causality, number, logical universality, and much else besides—to the touchstone of a priori psychologism in order to determine whether anything indispensable to its representation fails to be consciousnessindependent. What differentiates his psychologistic analysis of time from Hume’s and that of every other empirical psychologist is that the consciousness on which he determined time to depend is pure, not empirical. This not only serves to free its representation from everything specific to the sensation-dependent subjective time of internal experience, but also clears the way for its integration into physically real time as a constitutive transcendental ingredient of the latter’s representation. For that to happen, however, Kant first needed to show that, in the first instance, time is to internal sense and sense-divide transcending internal objects exactly what pure formally intuited space is to external sense and sense-divide transcending external objects: a purely aesthetic form lying ready a priori in the mind whereby heterogeneous internal sensations can be exhibited through homogeneous nonsensational temporal appearances.

Subjective and Objective Time and Spacetime Even today psychologists tend to take for granted the unifying many-inone representational structure of sensible consciousness, starting with a consciousness in which all the sensations currently existing in sensibility are contained. Kant, by contrast, wondered how even this minimal, instantaneous unity of sensibility is possible. The sensations it comprises are all different from one another in the sense that each might in principle still exist if any, or even all, the others did not. Each is thus a veritable conscious mind unto itself, capable of existing as the lone pearl of sensation in a creature like Locke’s oyster or decrepit old man. Kant saw no problem in the notion that sensations can multiply. But how can what multiplies unite to become a manifold of sensations contained

214  Nature in Mind: Through Kant’s I’s immediately together in the unity of a single, instantaneous consciousness? That was his question. One answer is that different sensations from all the senses fuse together to form multi-sensory panoramas much as colors in a pointillistic painting fuse to form a visual scene. Yet, notions like “fusion” merely redescribe sense-divide transcending, manifold-containing consciousness without explaining it. So too fusion at the level of their neural correlates: given that sensations are complete informational blank slates regarding their neural correlates and vice versa, any (e.g.  information theoretic) unity present in their neurophysiological architecture can no more be supposed to explicate the unity of sensible consciousness than the latter unity can explicate any unity proper to neural networks. In Kant’s view, sensation by itself (the synopsis of sense) is insufficient to explain unity of sensibility, synchronically no less than diachronically. There is no super-sense like the “common sense” of Aristotle and the Medievals that takes into its unity all sensations, internal no less than external, in the way the visual field takes different colors into its unity. Instead, the unity of sensibility, particularly across distinct sensory fields, must be achieved synthetically, in imagination, through the production of a second set of homogeneous non-sensational appearances—apprehension in intuition—in tandem with heterogeneous sensations. The form the unity of external sensibility takes in creatures constituted like us is space, which makes it possible to produce a species of appearances that can be combined by juxtaposition—one outside another—in a single, individual, purely aesthetic intuitive consciousness. But what about internal sensations: which form, what appearances, effect the unity of internal sensibility? According to Kant, the form this unity takes in humans and similarly constituted creatures is time, which makes possible the production of a species of appearances that can be combined by succession—one after another. Synchronically, this means even the most heterogeneous internal sensations are exhibited homogeneously in imagination by a single non-sensational temporal appearance which, when represented together with recollected or imagined temporal appearances, is apprehended in intuition diachronically as part of a succession of times. Internal sensations themselves, by contrast, no more exist in time than they do in space. The temporality of internal experience—a throbbing pain, a blinking light, a syncopated beat, intensifying anger, an interrupted train of thought, etc.—appertains to sensations only at second hand, a vicarious result of their indelible union with the non-­sensational temporal appearances synthesized in productive imagination to exhibit them. Given a relevantly different phylogeny or futuristic genetic (or other) engineering, they would be exhibited not by temporal appearances but via some other formal intuition totally unknown to us = Y, and instead of having subjective temporal experience via internal sensations those same sensations would yield subjective Y-emporal experience. For

The Aesthetic I 215 on Kant’s theory of internal sensibility and aesthetic time, it is only in imagination that the succession of consciousnesses that constitutes the life of the mind exists, nowhere else. Sensations themselves, internal no less than external, are altogether devoid of intuitable form—temporal, spatial, or any other. Once Kant’s distinction between temporal or spatial appearances and non-spatial, non-temporal sensations is grasped, we can appreciate why he insisted that being imaginary is not the same as being fictitious: “because time is the essential form of our sensibility and the receptivity of its intuitions whereby in general objects are given to us, and because the universal condition of sensibility must at the same time necessarily be the a priori condition of the possibility of all objects of the senses as appearances and so agree with it, it is not fictitiously invented by the imagination but underlies all its compositions and creations.”25 It is not a claim that physically real objects must temporally agree with the subjective time apprehended a posteriori in internal experience, but only that they must agree with pure time apprehended a priori in formal intuition. This, as will emerge in due course, is because the latter is a necessary ingredient of the objective transcendental time that, according to Kant, precedes and makes possible all objective empirical and mathematical representations of time. But quite apart from its contribution to the representation of physically real objective time, it is the function pure time performs of conferring a unity on internal sensibility it would otherwise lack (given synopsis alone) that makes time or any equipollent formal intuition so much more than a fiction. For all that matters at the most general level of Kant’s a priori psychologism is (1) that some internal sense-divide transcending object be representable through all the data of our “several internal senses,” (2) that this can happen only if internal sensibility constitutes a single consciousness comprehending all and only these data, and (3) that such a unity of consciousness is possible only insofar as imagination produces a pure formal intuition in and through which all possible internal sensations, howsoever heterogeneous, can be exhibited via imagination-generated homogeneous (temporal, Y-emporal, etc.) appearance proxies. So, even while conceding that purely aesthetic time of productive imagination is as imaginary as the wildest fantasy of reproductive imagination, Kant was on firm ground in insisting that anything so fundamental to sense perception itself (apprehension) is the very antithesis of fiction. Now contrast Kant’s a priori psychologistic explication of the time of purely aesthetic time-consciousness with Hume’s a posteriori explication of time as the “manner” in which different perceptions “make their appearance.” According to the latter, I have to observe kerosene-soaked rags igniting to perceive the manner in which a sudden profusion of black smoke makes its appearance; I have to perceive the lion’s sudden appearance before me before I  can inwardly experience the manner in which

216  Nature in Mind: Through Kant’s I’s terror wells up within me; I have to observe the blue sky make its appearance as I exit the building to detect the manner in which it relates to the no less equally empirical consciousness I had before I exited; and so on. Being ineluctably empirical, time cannot be supposed to be independent of sensational consciousness in either existence or individuation. As regards the former, this is just to say that if all consciousness a posteriori of particular sensations were eliminated, then all trace of time would disappear along with it—just as, on Berkeley’s account of space, spatiality would forthwith cease to exist if every trace of tactual sensation were expunged. As regards the latter, it means that if all consciousness of sensations were eliminated, there would be nothing to differentiate the time of igniting rags from the time black smoke first appeared, the time of sighting the lion from the time terror was first felt, etc.; in short, individual times would become completely undifferentiable and so lose all order and relation. Yet, to grant that the time of internal timeconsciousness is essentially dependent on consciousness a posteriori of the manner in which sensations, emotions, and thoughts make their appearance in human consciousness leaves no option but to concede that time is as ineluctably subjective as they are. Such subjectivity, however, renders time completely useless when it comes to explicating objective, physically real time, which has to be conceived as independent of sensations in both existence and individuation. If temporality is to be accorded any objectivity at all, it therefore has to derive it from convention, platonic heaven, or somewhere else. For if none can be derived from the psychological representation of time, there can be no escaping the conclusion that, in any final ontological accounting of physical reality, Humean psychological time is just as much a fiction as the most errant dream. No such conclusion, however, need be drawn when the time of purely aesthetic time-consciousness is supposed to depend not on empirical but on pure consciousness, as Kant did. In the first place, being pure, the formal intuition of time is in no way beholden to a posteriori consciousness of particular sensations, and so can be represented as existing completely independently of the sensations exhibited by the temporal appearances contained within it. That leaves only the question of the differentiation, relation, and ordering of times within it, i.e. their sensation-independent individuation: can this ever take place wholly independently of particular perceptions of such phenomena as sirens, melodies, and speech, fading, blinking, and steady lights, sudden or gradually waxing and waning feelings, or trains of thoughts being continued, ended, or interrupted—i.e. independently of all a posteriori consciousness? What here is vital to recognize is that the answer does not have to be no if the time of internal time-consciousness is explicated as a purely formal intuition of pure sensibility. For if not sufficient, this intuition at least provides an essential ingredient for the individuation of times in pure consciousness.

The Aesthetic I 217 How? This is where pure concepts of the understanding first emerge as constitutive elements of Kant’s psychologism. Being no less pure than formally intuited time, their application to the totality of times possible within the latter becomes practicable completely a priori. Accordingly, Kant’s task became showing how their application to the formal intuition yields consciousness of a genuinely objective transcendental time wherein each time is uniquely differentiated and completely determined in respect to every other, possible no less than actual (Chapter 9). Since this differentiation and determination suffices for their complete objective ­individuation—i.e. an individuality independent of all a posteriori consciousness of particular sensations—it makes transcendental time the very antithesis of Hume’s subjective, sensation-dependent, wholly fictitious time. It also makes it something Kant could argue to be essential to all empirical and mathematical representations of time, including the representation of nature itself—the objective, physical reality of time (Chapter  10). Thus, Kant’s psychologism is already of itself, without any need for conventionalist (e.g. mathematical) supplementation, in full concordance with contemporary physicists’ denial of the least objective reality to the time of subjective internal experience, and so too with their complete exclusion of it from any final ontological accounting of physical reality. Moreover, because the objective, physically real time Kant incorporated into his psychologism is determined conformably only with transcendental concepts but is otherwise conceptually completely indeterminate, there is nothing to prevent it from accommodating relativistic spacetime or any other non-transcendental, scientific conception of the physical reality of time that may supersede it.26

Aesthetic Time Still, we must be careful not to get too far ahead of ourselves, or we risk conflating purely aesthetic formal space and time with the objective transcendental space and time of transcendental synthesis, and so compound the risk of conflating the latter with properly mathematical and empirical space and time. As regards the former, their purely aesthetic character rules out any contribution from understanding, including the ‘I think,’ logical forms of judgment, pure concepts of the understanding, and concepts generally (mathematical, scientific, and ordinary); while their character as pure precludes any contribution that requires consciousness of particular sensations. This means that both the nature and origin of purely aesthetic formal space and time must be understood entirely in terms of a pre-logical, non-objective, radically subjective,27 yet non-fictive faculty of productive imagination. If indeed this purely productive function of imagination exists as Kantian a priori psychologism describes it, it would presumably have had to evolve separately from its merely reproductive employments in memories,

218  Nature in Mind: Through Kant’s I’s associations, dreams, and fantasizing generally. The one common feature of the forms it produces is the infinitely iterable many-in-one structure (individuality) that would have become possible only with the genetic and/or epigenetic alterations responsible for the advent of the requisite neural correlates. This means that creatures with reproductive imagination need not have productive as well. Such creatures could still have a unified internal sense but only at the price of having a single internal sense that constitutes a field in something like the way visual or tactual sensations do. Nevertheless, as purely sensational, the unity such a field might confer on creatures’ internal sensible consciousness could not be a temporal unity of consciousness for the same reason Kant held that all consciousness of space and spatiality would be absent from the visual and tactual fields of creatures that lacked a formal intuition with which to synthesize spatial appearances: time and the temporal, like space and the spatial, is always intuition, never sensation. As with apprehension of spatial appearances in external intuition, the apprehension of temporal appearances in internal intuition is confined to dark consciousness. In the case of the former, this could be attributed to the mind’s inability to anticipate which sensations do and do not correlate across sense divides, and, for those that do, to which particular sensations of which senses they correlate. It is therefore required that external appearances be synthesized to exhibit every external sensational difference without exception, however nuanced, whether or not it ultimately turns out to provide material for the representation of sense-divide transcending external objects. Since individually discerning and directing attention to these countless sensation-appearance couples would be a vast waste of what are presumably neuro-energetically costly modes of mentation, and may be practicably impossible anyway, synthesis of apprehension in external intuition must be presumed to take place mainly, if not entirely, in dark consciousness. The case of apprehension in internal intuition is different in that there are few if any systematic correlations among internal sensations. The reason that discernment and attention are never required here is instead that every internal sensation, possible or actual, turns out to relate to one and the same sense-divide transcending internal object: the empirical self (discussed in Chapter  10). This is not to say that there are no non-internal representations that we regard as self-related as well: every external sensation present in my consciousness is my visual, tactual, etc. sensation; so too my body, my passport, and everything else that defines who I am in relation to the physical and social worlds I inhabit. Nevertheless, insofar as sensible access to any external object, one’s own body included, requires external sensations, and everything that depends on external senses to access is ipso facto external to oneself, the self proper is accessible exclusively through the internal senses. There is accordingly no fitness benefit to be derived from expending neural energy on

The Aesthetic I 219 discernment and attention to sift through couples of internal sensations and the temporal appearances that exhibit them to determine which are relevant to empirical self-cognition and how. In any event, creatures like ourselves rarely have need or occasion to direct attention inwardly since “the representations of the external senses constitute the actual material with which we occupy our mind.”28 Thus, sensible consciousness, encompassing both external and internal varieties and appearances no less than sensations, constitutes an “immeasurably great” field of darkness on which discernment and attention operate only in the most selective, circumscribed ways—“like a great map of our mind on which only a few places are illuminated.”29

Hume’s Quandary Revisited Once purely aesthetic subjective time is distinguished from objective transcendental time as Kant proposed, a quite possibly fatal flaw in associationisms like Hume’s becomes apparent. Customary association depends on the ability to experience the occurrences of representations, their co-occurrences, and their sequences as (more or less) frequent and constant. However, even so much as dark conscious experience of these is impossible without a representation of all perceptions as events occurring in a unified, ever-changing, irreversible linear time in which each moment is objectively individuated from all the rest (i.e. uniquely differentiated and completely determined as both determinately before or after every other and so much before or after it). But if Kant’s account of internal sensibility is correct, consciousness of the time of perceptions as events in objective time is impossible purely subjectively and aesthetically via imagination but requires, in addition, the conversion of the formal intuition of time into objective transcendental time by means of transcendental synthesis founded on pure concepts of the understanding. Since this means that consciousness of more or less frequent, constant ­co-occurrence and sequence that customary association premises presupposes the full panoply of a priori cognitive psychology, Hume’s assumption that he could make use of customary association to account for objective space, time, causality, and identity becomes susceptible to the charge of presupposing precisely what it purports to explain.30 More fundamentally still, Hume failed to ask how, custom aside, association itself is possible. We saw the insoluble problem this creates for empirical psychologism at the end of Chapter 4 in connection with the inescapable quandary in which he found himself when it came to explaining how successive perceptions come to be united in consciousness. Association fails because it involves consciousness of transitions of thought, which is possible only insofar as one and the same consciousness endures through the transition, thus implying the pre-associative existence of a numerically identical temporally persistent consciousness. Hume viewed

220  Nature in Mind: Through Kant’s I’s it as first and foremost a problem with his associationist account of the self because the system of customary causal associations it premises presupposes consciousness of myriad associative transitions, and the more transitions we are supposed to be successively conscious of, the longer the pre-associative numerical identity of consciousness must be conceived to endure. Since he could see no way to understand this pre-associative idea of an enduring consciousness except in terms of the traditional metaphysical notion of a soul-substance, and since the unrenounceable principles of his skepticism precluded metaphysical explanation of any kind, he had to “confess this difficulty is too hard for my understanding.”31 Hume would have been better advised to frame his quandary in terms of time-consciousness than personal identity. For the key to finding a way out is the recognition that even if sensations have a temporal appearance, it does not follow that their temporality is itself sensational. Kant’s theory of internal sensibility shows how the temporality of internal sensations can consist entirely in imagined non-sensational temporal appearances, with sensations themselves otherwise falling out of the temporal picture altogether—just as on his account of external sensibility space disappears even from tactual sensation and resolves itself entirely into a synthetic unity of non-sensational spatial appearances. This means that, on the one hand, the synthetic unity of temporal appearances in timeconsciousness poses no quandary because all succession of appearances is preceded by, made possible by, and contained within the individuality of formally intuited aesthetic time, with the consequence that their unity in one consciousness has the same psychological explanation they do—productive imagination’s synthesis of apprehension in intuition— and so offers no occasion for recourse to metaphysical principles. On the other hand, because internal sensations in synopsis (as distinct from apprehension) have no unity of consciousness of their own, be it temporal or any other, but instead exist pre-imaginatively only as atemporal, aspatial isolated individuals, each occupying its place within one or another radically heterogeneous internal sensory field, the metaphysics of substance has no succession of sensations—and a fortiori no unity of their succession in consciousness—to explain, and so becomes irrelevant. Thus, Kant’s extension of the scope of imagination beyond association to include the synthesis of apprehension responsible both for generating the successive appearances themselves and for the synthetic unity of internal sensibility that comprehends them enabled him to bypass Hume’s quandary entirely.

The Original Synthetic Unity of Apperception Aesthetic time suffices for one thing and one thing only: the unity of internal sensibility. Thanks to it, internal sensations exhibited by the same temporal appearance are apprehended together in internal intuition as

The Aesthetic I 221 simultaneous, while those exhibited by different temporal appearances are apprehended as in (undifferentiated, indeterminate) succession. Not even the minutest internal sensational difference, whether discernible or not, is left unexhibited by a temporal appearance belonging to this unity, nor is any internal sensation possible that could not be so exhibited and incorporated into it. It may be a unity that internal sensations have only at second hand, via the non-sensational temporal appearances that exhibit them. It may be that the infinitely iterable many-in-one structure of internal intuition is merely the synthetic unity of the manifold of temporal appearances. But productive imagination’s synthesis of apprehension in internal intuition nonetheless suffices to make the totality of possible internal representations, appearances and sensations alike, representable a priori as the manifold of a single intuitive consciousness. Thus does the formal intuition of time yield unity of internal sensibility. Yet, Kant’s interest in the production of unity of sensibility did not end with its synthesis through this or that formal intuition. As noted early in the chapter, the universality of the ‘I think’ differs from that of the AUA-concepts it makes possible insofar as what is universal is not a representational content of any kind (fresh beeswax, red, fear, etc.) but a consciousness common to all other representations, possible or actual, dark or accessible via conscious discernment and attention. However, a consciousness cannot be represented as common to any representation that is not already contained within it. A consciousness common to all possible representations must therefore take its start from a consciousness that contains all possible representations individually within it, i.e. one and the same consciousness with that totality as its manifold. What this individual consciousness is, however, given that it can be nothing like a Leibnizian monad, was initially unclear. Only now are we in a position to see that it is none other than the unity of sensibility, external no less than internal, effected by aesthetic time. Why does time’s unity of sensibility include the exhibition of external sensations while space’s unity excludes that of internal? External sensations are prior to and independent of the self-affecting actions and passions of the mind to which our internal senses are sensitive: discerning, directing attention, comparing, recalling, fantasizing, believing, grieving, desiring, resolving, representing propositionally, etc. To be sure, the neural correlates of internal sensations can—indeed must—be represented in space through external sensations. But because internal sensational consciousnesses themselves are one and all informational blank slates regarding their neural correlates, it makes no difference whether the latter are in themselves spatial, temporal, or something else completely unknown to us = X. Internal sensations have their own formal intuition through which the apprehension of the appearances that exhibit them occurs, and it is not space but time. And since this is just to say that “not all that is in time is at the same time in space,”32 it follows that aesthetic

222  Nature in Mind: Through Kant’s I’s space is capable of effecting only the unity of external sensibility but not that of sensibility in its entirety. Aesthetic time has no such limitation. As the formal intuition that precedes, makes possible, and contains within it the temporal appearances that exhibit the sensations of all our “several internal senses,” aesthetic time includes the temporal appearances that exhibit the data of the internal sense sensitive to the acts of productive imagination in its synthesis of apprehension in external intuition—i.e. to the pure and empirical syntheses of apprehension in external intuition responsible for producing the pure formal intuition of space and spatial appearances. By incorporating internal sensations of all acts of external apprehension into the unity of internal sensibility, the formal intuition of time thereby indirectly incorporates into its succession formally intuited space and all the spatial appearances contained within it; and insofar as the latter exist in tandem with the external sensations they exhibit, those sensations too fall within the unity of internal sensibility effected by aesthetic time, albeit at two removes. In this way, the formal intuition of time makes possible not only the a priori synthetic unity of the infinite totality of possible internal intuitions and sensations but also, indirectly, that of the infinite totality of possible external intuitions and sensations, with the consequence that “all that is in space is in time.”33 Thus, the unity of sensibility formed by folding the spatial into the temporal is none other than the all-encompassing individual consciousness presupposed by the universal self-consciousness ‘I think,’ and so the very unity Kant termed the original synthetic unity of apperception (SUA) and elevated above everything else in a priori psychologism—including the ‘I think’ (analytic unity of apperception [AUA])—as its supreme principle.34 It is imperative, however, not to equate the synthetic unity of sensibility with either the unity of formally intuited time as such or an amalgam of the unities of formally intuited space and time. Because the precise nature of formal intuition is dependent on the particular constitution of a creature’s sensibility, the possibility cannot be precluded that the very same internal sensations we have might be synthetically unifiable in an all-encompassing individual consciousness (SUA) by means of formal intuitions other than space and time. Accordingly, what is essential for Kant’s psychologism is not the temporality of internal intuition, the spatiality of external intuition, or the X-iality and Y-emporality of any alternatives that might evolve naturally or be genetically (or otherwise) engineered, but rather the infinitely iterable many-in-one structure common to all formal intuitions as such. For no matter how pure sensibility is constituted, so long as it enables a creature to synthesize a consciousness that individually comprehends within it, at however many removes, the infinite totality of possible external and internal sensations completely a priori, it will afterwards necessarily be able to represent, equally a priori, an ‘I think’ able to accompany, and so logically common to, all possible

The Aesthetic I 223 representations (it is in this sense that the ‘I think’ of AUA, viewed from an evolutionary perspective, is a spandrel of SUA). To be sure, it is a merely synthetic a priori unity of the manifold in one consciousness in which sensations are contained only indirectly, not a literal unity of the absolute totality of representational contents of the kind envisaged by Leibniz. Yet, to the non-metaphysical, a priori psychologistic end of explicating the individual consciousness that makes the universal selfconsciousness ‘I think’ possible, that is all Kant required. The upshot is that the synthetic a priori unity of sensibility effected via formal intuitions needs to be understood as not just the presupposition but the inverse of the universal self-consciousness ‘I think,’ with the consequence that a complete notion of pure self-consciousness, or unity of apperception, must include a purely aesthetic ‘I’ (SUA) no less than a purely logical ‘I’ (AUA): [O]nly by my being able to combine a manifold of given representations in one consciousness is it possible for me to represent the identity of the consciousness in these representations itself. That is, the analytic unity of apperception is possible only under the presupposition of some synthetic unity. Accordingly, the thought that those representations given in intuition one and all belong to me is as much as to say that I unite them, or at least can unite them, in one self-consciousness.35 The individuality of SUA—the aesthetic ‘I’—and the universality of AUA— the logical ‘I’—are two sides of a single coin. A consciousness that can be represented as able to accompany, and so as logically common to (i.e. identical in), every representation that can ever be mine entails a consciousness that individually unites all possible representations within it, even if only as a synthetic unity. The unlimited universal scope of the former is impossible without the all-encompassing individuality of  the latter. The true transcendental (a priori psychologistic) concept of pure self-­consciousness is therefore one that includes both synthetic and analytic unity of apperception equally and inseparably. It is both analytically, so that the one is as logically impossible without the other as mountains without valleys.36 Unlike mountains and valleys, however, AUA is preceded and made possible by SUA but not vice versa. Pure ­ sensibility—­ apperception as the aesthetic ‘I’—precedes pure ­understanding—apperception as the logical ‘I.’ The synthetic unity of the first has to be in place before the analytic unity of the second can be represented, i.e. before the representation ‘I think’ can take place. That is why Kant held that the original, purely sensible synthetic unity of apperception is “the highest point to which all employment of the understanding, even the whole of logic, and in accordance with it, transcendental philosophy, must be affixed—indeed this capacity is the understanding itself.”37

224  Nature in Mind: Through Kant’s I’s How does the purely aesthetic I  of the synthetic unity of apperception constituted by sensible formal intuitions alone (unity of sensibility) relate to the logical synthetic unities of apperception made possible by forms of judgments considered in Chapter 7? Since the latter presuppose the ‘I think’ and so can play no role in originally making it possible, only the synthetic unity of apperception generated from formal intuitions deserves to be deemed original in the strict Kantian sense of being presupposed by all other synthetic unities while itself presupposing none. At the same time, without the logical forms of judgment, the logical properties implicit in AUA-concepts would remain unrepresentable in consciousness. Since this would prevent all AUA-concepts from being combined in a single consciousness that could then be represented as common to all, the universality of the ‘I think’ could not extend to them, and they would quite literally be “nothing for me.”38 This means that an ‘I’ whose universal scope encompasses not only all non-logically structured aesthetic representations but all possible propositions formable from AUA-concepts presupposes the synthetic unity of apperception originally constituted by logical forms. So, relatively to all synthetic unities that in any way involve propositional thought (including objective transcendental space and time), the synthetic unity of apperception constituted by the logical forms of judgment must be regarded as original as well. And since, as we shall now see, Kantian a priori psychologism resolves all objective representations, from the sense-divide transcending objects encountered in ordinary experience to the laws of physics and the most arcane constructions of mathematics, into propositionally based synthetic unities, the (relatively) original synthetic unity of apperception constituted by the logical forms proves to be no less psychologistically fundamental to intelligent mind than the unqualifiedly original synthetic unity constituted by purely aesthetic formal intuitions.

Notes 1. Both the contrast and the complementarity are clearly evident at CPR B136n (cited in note §34 later in the chapter). 2. Kant’s general distinction between analytic and synthetic propositions has frequently been skewered by analytic philosophers who assume it to have been intended to apply to language as well as to a priori psychologism. Yet, in Kant’s own view, while the distinction is “indispensable in regard to the critique of human understanding”—for which read ‘a priori psychologism’— he deemed it “of no considerable use in any other regard” (PFM 270). The analytic relationship in which the ‘I think’ is involved is specified in the texts cited in note §36, but there is still much to consider before their import can be appreciated. 3. CPR B422. Also: “Apperception is itself the ground of the categories” (CPR A401; see A346/B404 as well). Here it is worth remarking that one reason Kant commentators seem to me almost invariably to go wrong is that they overlook or disregard his commitment to the thesis that apperception grounds the categories and instead ascribe to him its converse. The most

The Aesthetic I 225 obvious, but by no means the only, problem with this practice is that the converse is viciously circular: given that the categories are concepts, how can they ground the ‘I think’ if, as Kant clearly held (B133–4n), the latter is a universal self-consciousness that is a constitutive logical condition of all concepts as such, transcendental concepts like the categories not excepted (expressly: see A341/B99–400 and A348/B406, both cited in chapter 6 note §19)? On my reading, by contrast, only the objective I discuss in chapters 9 and 10 presupposes pure concepts of the understanding/categories, whereas the logical I of chapter 7 and the aesthetic I of the present chapter are presupposed by them. For more detailed discussion, see KAIM preface, chapters 5 and 9, and conclusion-A. 4. Because Kant’s is a purely sensible concept of intuition, with no conceptual or propositional component, it is imperative never to confound his somewhat idiosyncratic use of the term with the intellectualist sense in which other philosophers and mathematicians tend to employ it. 5. CPR A120. 6. See CPR A94 and A97. Another way to put the point is that by ascribing apprehension to imagination rather than sense, Kant was the first not to conflate it with synopsis. 7. CPR A120n. 8. The facility and vivacity affects that Hume attributed to the imagination do not constitute exceptions because they have to do with not the contents we are conscious of but merely the manner in which we are conscious of them (chapter 4). In Kant’s aesthetic theory, they fall under the rubric of internal sensation (see later in the chapter). 9. “Logical form is to the intellectual representation of things precisely what space and time are for the appearances of a thing: namely, they contain the positions for ordering them” (AA 17 §4629 [1771–3]). By occupying positions in space, appearances are set in juxtaposition, and by occupying positions in time, are set in succession. Time will be considered later in the chapter. 10. The neural correlates responsible for apprehension might be supposed to be selected in a manner akin to Gerald Edelman’s notion of neuronal group selection, where a plethora of neural connections in the infant brain are extensively pared away when not reinforced by use, allowing for a high degree of functional plasticity. 11. CPR A20/34. 12. My purposes in this book are best served by following Kant’s usual practice of using the terms ‘form of intuition,’ ‘pure intuition,’ and ‘formal intuition’ (also ‘form of appearances’ and ‘form of sensibility’) interchangeably. Only on rare occasions did he distinguish them, none of which are important here. For discussion, see KMM part I and KAIM chapter 5. 13. That the form external appearances take in humans is space rather than something else is, for Kant, a fact about the innate constitution of our minds. This, however, does not make appearances or the formally intuited space that contains them innate representations. In the same way dreams are not innate representations but instead have to be synthesized from data originally deriving from sensation, formally intuited space and the appearances it contains are likewise synthesized from sensations. The only difference from dreams and other forms of reproductive imagining is that no consciousness of particular sensations—possible only a posteriori—is required for their synthesis. Because the products of productive imagination are not derived from consciousness of sensation but instead original to imaginative consciousness itself, Kant saw fit to characterize them as original acquisitions: see On a Discovery, pp. 221–3.

226  Nature in Mind: Through Kant’s I’s 14. Even Kant specialists typically fail to separate out everything contributed by the understanding from their analysis of the purely sensible space and time of transcendental aesthetic. Yet, Kant could hardly have made his intentions clearer: “Transcendental aesthetic . . . isolate[s] sensibility by . . . separating out everything that the understanding thinks in it by means of its concepts so that nothing but . . . intuition remains” (CPR A22/B36). This text also calls for abstraction from everything to do with sensation (experience) so as to leave only intuition that is pure, which is likewise only rarely achieved. It is important to note, however, that what it does not call for is abstraction from imagination (as Hegel remarked in Faith and Knowledge). Moreover, there are numerous texts, some cited later in the chapter, that show the a priori, purely productive iteration of this faculty to be essential to Kant’s conception of pure sensibility. For detailed discussion, see KMM part I and KAIM chapters 3–5. 15. I have adapted the formula “uniquely differentiated from and completely determined in relation to” from CPR A581–2/B609–10, a text in the Transcendental Ideal chapter that Kant expressly related to the Transcendental Analytic, i.e. the portion of the Critique specifically concerned with the psychologistic explication of objectivity. 16. Although Kant also termed the data of internal sense ‘internal affections’ and ‘self-affections’, I  have opted for ‘internal sensations’ (an expression Kant used at Anthropology §7 and AA 23 28 E LX) to emphasize their generic identity with external sensations. Examples of internal sensation can be found at CPR A357–8. 17. AA 15 §224 (1783–4). Since Kant never specified the internal senses he meant to distinguish and neural anatomy in his time was far too primitive to distinguish organs of internal sensation, one can only conjecture what he meant when he referred to internal senses in the plural. Since my brief in this book includes updating his views and neuroanatomy today has reached the point where it might be said to distinguish different internal senses (e.g. the amygdala as important for emotion), it seems possible to fill in the gaps he was obliged to leave, if still only to a modest extent. 18. See chapter 10 note §8. 19. Anthropology, p. 153. 20. T 1.3.5 ¶3. 21. “They are the successive perceptions only that constitute the mind” (T 1.4.6 ¶4), so that apart from idea-enlivening associative imagination, “we could only admit of those perceptions [in that succession] which are immediately present to our consciousness, nor could those lively images with which the memory presents us be ever received as true pictures of past perceptions” (1.4.7 ¶3). See HTC chapters 2-C, 6-A, and KEUU chapter 16-F. 22. T 1.2.3 ¶10. Hume used ‘perception’ to designate internal and external sense impressions and ideas stored in memory or fashioned in imagination. 23. CPR B422. 24. CPR A37/B54. 25. Discovery, p. 203. I have amended the original, which concerns both space and time (i.e. formal intuition in general). Here is the whole without changes and excisions: “One can and must concede that space and time are mere thought entities and creatures of the imagination. But because they are the essential form of our sensibility and the receptivity of its intuitions whereby in general objects are given to us, and because the universal conditions of sensibility must at the same time necessarily be a priori conditions of the possibility of all objects of the senses as appearances and so agree with these, they are not fictitiously invented by the imagination but underlie all its

The Aesthetic I 227 compositions and creations.” Among numerous other texts to the same or similar effect (see KMM introduction and part I) is the following: “Space and time are of course not objects of intuition but merely its subjective forms. They do not exist apart from representations and are given only in the subject, i.e. their representation is an act of the subject itself and a product of the imagination for the sense of the subject. Yet, their representation is the cause of the object in appearance (phænomenon), and is not derived (repræsentatio derivativa) but original (originaria)” (AA 22 76). Regarding the original, non-derivative character of formal intuitions generally, see note §13 earlier in the chapter. 26. Some readers may already have realized that Kant’s a priori psychologism obliges its upholders to reject any claim that objective, physically real space and time are consciousness-independent. For even while agreeing that they are independent of empirical consciousness, their dependence on pure consciousness follows if scientific and mathematical representations of space and time can be shown to depend on subjective, consciousness-dependent formally intuited space and time and pure concepts of the understanding. Kant’s case for holding that they do will be examined in the next chapters. 27. By way of reminder, I am here using ‘subjective’ merely epistemologically, by contrast with the objectivity that emerges only after transcendental synthesis has been applied to the appearances of sensibility. Psychologically, the only subjectivity in Kantian theory is that predicated on the contentless pure consciousness ‘I’: see chapter 6. 28. CPR B67. 29. Anthropology, p. 135. 30. The same objection can be leveled against Berkeley’s associationist account of sense-divide transcending spatiality. However, to appreciate that and how his account breaks down in consequence, one needs to factor in how Kant’s psychologistic explication of objective time connects up with that of objective space (chapter 9) and their extension to physical reality (chapter 10). 31. Appendix on personal identity to A Treatise of Human Nature ¶12. Hume also noted that objectively necessary causal relations could provide a way out of his quandary but saw no way of countenancing them given that his unrenounceable principles imply that it is impossible to represent such relations purely psychologically, without recourse to linguistic convention: see chapter 5; also KEUU chapter 3. 32. AA 18 §5653. 33. AA 18 §5653. Kant made this point as early as 1770: “Time . . . more nearly approaches a universal and rational concept, for it embraces in its relations absolutely all things, namely, space itself and, in addition, the accidents which are not included in the relations of space, such as the thoughts of the mind” (ID 405). See also CPR A34/B50–1, A155/B194, and B427. 34. “The principle of the synthetic unity of apperception is the supreme principle of all employment of the understanding” (CPR B136, title §17). That this includes the unity of sensibility at the focus of transcendental aesthetic is evident in this crucial addendum (subjected to detailed analysis in KAIM chapter 5-B): “Space and time and all their parts are intuitions, hence individual representations with the manifold they contain in them (see the Transcendental Aesthetic). Not mere concepts through which one and the same consciousness is contained in many representations but many contained in one and in the consciousness of that one, and thus as composite. Consequently, unity of consciousness that is synthetic yet also original is met with in them. This individuality of space and time is important in application” (CPR B136n). This could not be more explicit: purely aesthetic space and

228  Nature in Mind: Through Kant’s I’s time constitute an original synthetic unity of all their manifold in one consciousness. See also KAIM chapter 9-B. 35. CPR B133. If any one text can be regarded as the most important Kant ever penned, this, in my opinion, is it. No interpretation that fails to fully account for every word and phrase (including the accompanying footnote) can be deemed tenable. 36. See CPR B135 and B138. Analyticity, though not explicitly asserted, is implied at CPR A107, A113, and A129. 37. CPR B133–4n. This is one reason why it is so important not to construe apperception as constituted by the categories (rather than their presupposed ground—see note §3 earlier in the chapter). 38. CPR B132.

9 The Objective I

Kant surpassed even himself in his elaboration of the thesis that the synthetic unity of apperception is the ground of the objectivity of space, time, nature, and their cognition in ordinary experience, mathematics, and science. There is a depth and virtuosity to his philosophizing he never surpassed and rarely equaled, sustained page after page, on the topic he deemed philosophically preeminent. It is Kant’s supreme achievement, which, given his stature in the history of philosophy, makes it a high point of the discipline. Although the relevant texts are dauntingly difficult, my prime concern in this book is with ideas. Accordingly, I will focus solely on the thesis at their heart. The present chapter will deal with its strictly a priori side: the role of pure concepts of the understanding in the objectification of the purely aesthetic formal space and time considered in the preceding chapter. In the next chapter, I will examine Kant’s account of the role of pure concepts of the understanding in the objectification of the products of empirical psychology, which, on his telling, is the story of nothing less than the possibility of physical objects, their laws, and nature itself as their systematic totality. What both share is the idea that unity of apperception—individual and universal self-consciousness—underlies all objectivity, so that empirical and mathematical objects alike derive their objectivity entirely from the most quintessentially subjective of all representations, the I.

Conceiving Space and Time Free of Extrinsic Limitations Some scientists speculate that an eventual quantum field theory that includes gravity will render space and time redundant. By “space and time” they mean the spacetime of relativity theory. The idea, as I understand it, is that if continuous spacetime, in which no space is the smallest and no time the briefest, can be resolved into a theory that recognizes nothing except discrete field quanta, then, poof!, space and time disappear entirely and nothing physically objective remains to which the names “space” and “time” can, in strict propriety, be attached. In that

230  Nature in Mind: Through Kant’s I’s case, space and time will have proved to be emergent properties in the way heat was shown to be an emergent property of molecular motion: real enough for most purposes but not something that needs to be mentioned in any final accounting of physical reality. Some say that continuous spacetime has already been resolved into discrete field quanta. Here is how physicist Vlatko Verdral explains it: [T]he fact that quantum mechanics applies on all scales forces us to confront the theory’s deepest mysteries. We cannot simply write them off as mere details that matter only on the very smallest scales. For instance, space and time are two of the most fundamental classical concepts, but according to quantum mechanics they are secondary. The entanglements are primary. They interconnect quantum systems without reference to space and time. If there were a dividing line between the quantum and classical worlds, we could use the space and time of the classical world to provide a framework for describing quantum processes. But without such a dividing line—and, indeed, without a truly classical world—we lose this framework. We must explain space and time as somehow emerging from fundamentally spaceless and timeless physics.1 But even if Verdral is correct, it is one thing to say these things are true of the world constituted through the four known fundamental forces (i.e. the quantum fields in which their fermions and bosons exist), but quite another to say they are true unqualifiedly and absolutely. We only need consider how meager our access to the physical universe really is. Our bodies and the world accessed through their sense organs are, for all physical intents and purposes, expressions of electromagnetic force: this is what holds our atoms and molecules together; the photons that strike our retinas are its bosons; electrochemical gradients are essential to the workings of all prokaryote and eukaryote cells, neurons not excepted; and the same electromagnetic character extends to all the devices scientists have built and are ever likely to build. The strong and weak forces are accessible to our senses only insofar as they give rise to electromagnetic effects on macro-scale electrically powered detectors composed of atoms and molecules held together by electromagnetism. Even gravity, which in Einsteinian truth is just the geometry of spacetime, is sensible to us (“felt”) only insofar as the electromagnetic forces acting in and on our bodies conform to that geometry.2 But what if there is more to physical reality than can be detected by electromagnetically constituted creatures and their technology? There might well be physical forces that have no electromagnetic effects, direct or indirect, at least at our scale and/or in our dimensions, and are therefore opaque to our senses and detectors, presently and perhaps forever. Any such realities, including kinds of conscious mind they might make

The Objective I 231 possible, would be just as physical as we are but lie entirely beyond human cognitive access. And what if the nature of those realities could not be explained in terms of discrete quantum units but instead entailed the physical reality of continuous space and time: do we not have to at least leave open a place in our representation of objectively real space and time for the possible physical reality of the latter too? Kant’s transcendental psychologism certainly requires it. Forces inaccessible to our electromagnetically constituted bodies are not, for that reason, any less likely to exist than electromagnetism itself, or any and all other forces detectable through their electromagnetic effects at the spatial and temporal scales at which we and our devices exist. Physical space and time should therefore be analyzed as Kant proposed: with complete surety that no creaturely physical limitations are surreptitiously incorporated into the result. Otherwise, the analysis cannot be fully and truly objective. Kant was equally insistent that creaturely mental limitations not be incorporated into transcendental analysis. Among the most insidious of these is the tendency to conflate space and time with the sensations through which they are experienced. For example, if space is a pure formal intuition comprising a manifold of non-sensational spatial appearances, then the fact that only three-dimensional appearances are required to exhibit visual and tactual sensations tells us nothing about space itself. If we had other senses, we might need only two- or even one-dimensional spatial appearances to exhibit external sensations. Or, alternatively, if we had senses that require spatial appearances of four or more, or even infinite, dimensions to exhibit their sensations, is there any reason to assume that our present formal intuitions would be inadequate to the task? To do so would, so far as Kant was concerned, be incorporating the characteristics of human sensation-producing external senses into the non-sensational spatial appearances that exhibit them—for which there can be no transcendental (i.e. a priori psychologistic) justification if, as he maintained, the latter are products not of the senses but of a productive imagination apprehending in accordance with a non-sensational form that “lie[s] ready for sensations a priori in the mind.”3 Indeed, as will become apparent shortly, formal space, understood purely transcendentally, is quite capable of accommodating any number of dimensions, including infinitely many. As for time, it is only in tandem with internal sensations, and so subjectively, in empirical apprehension, that we experience successions of temporal appearances as “taking up” time or being “refreshed” at such-and-such a “rate.” But productive imagination also enables us to intuit time purely formally, without regard to internal sensations, in pure apprehension. Here, temporal appearances quite literally take up no time, so that their succession can be represented in a single, unprotracted act of pure intuition. Its instantaneity is important because, on Kant’s

232  Nature in Mind: Through Kant’s I’s transcendental analysis, without this purely formal intuition of time we could not represent succession at all, pure or empirical, continuous or discrete, abstract or concrete. This includes everything from the most primitive, purely aesthetic succession in empirical consciousness such as the perception of one’s own train of thought to successions as abstract and arcane as algebraic functions successively ranging over a domain of arguments. Pure time, for Kant, is nothing less than the formal condition of all series, so that successively running through any multiplicity, however concrete or abstract, would be completely impossible without it. Yet it cannot be stressed too strongly that Kant’s claim is premised on time being a pure, purely formal intuition original to productive imagination, not a sensation or sensation-incorporating representation, and so intrinsically distinct, even if only in dark consciousness, from all that sensations and their exhibition contribute to our empirical apprehensions of successions. Pure sensibility is thus an essential step in Kant’s argument, though by no means the last, as we shall now see.

How Transcendental Space and Time Are Necessary for Mathematics We have seen why Kant’s conception of objective space and time cannot be contradicted by any physical theory and is completely independent of how they are experienced through sensations, human or any other. But what about mathematics? The assumption is all but universal among philosophers and scientists that Kant’s claim that pure mathematics presupposes formal space and time should be understood to apply only to space, to concern no branch of mathematics but geometry, and more particularly Euclidean geometry of the kind practiced today by high schoolers with rule and compass; while Kantian time is consigned to applied mathematics, as in classical era physics of motion. But aside from ignoring his iterated insistence that his theory explains the possibility of all mathematics, anyone who makes this assumption is failing to carry out Kant’s express instruction to keep the purely aesthetic space and time of sensibility distinct not only from sensation and the contributions of pure understanding via its pure concepts, but also from concepts and logically structured propositional thought generally (discursive understanding in general)—not excepting the concepts and propositions that define Euclidean geometry.4 What Kant actually held is that once the purely formal manifolds of aesthetic space and time—infinitely iterable juxtaposition and succession within and without juxtaposition and succession without end—are determined conformably to the pure concepts of the understanding they become objective space and time. He never claimed that determination by these concepts suffices also to transform formal space and time into any properly mathematical space or time, be it Euclidean or any other.

The Objective I 233 Instead, he characterized the result of this determination as objective transcendental space and time—a term he used to indicate that they precede and make possible all other representations of objective space and time, whether in pure mathematics, physics and other mathematizable sciences, or non-scientific quantitative representation of every kind from bookkeeping to numerology. In the case of geometry, Kant’s focus was determined by his endeavor to refute Humean skepticism after having himself extended it from causal relations to mathematics (Chapter  5). He held up his own account of space in terms of pure sensibility to counter the notion that the only alternative remaining once the platonism of Descartes, Leibniz, and others regarding space is discarded is that space and its geometrical determination are dependent on a posteriori consciousness of the particular sensations used in demonstrations (visual squares, lines, etc.). In contrast to a posteriori cognition, Kant’s a priori psychologistic explication of space provides a basis for supposing that geometry confers the kind of unalloyed necessity and universality on its definitions, axioms, postulates, and theorems that he deemed distinctive of mathematics overall and a priori cognition generally. Since the only propositions geometers had succeeded in mathematically validating in his day were Euclidean, Kant’s argument took the form of affirming the a priori necessity and universality of Euclidean propositions. But this anti-empiricist thesis must not be confounded with the dogmatic claim that geometry itself is necessarily and forever Euclidean. Its meaning is not geometrical but transcendental, and so devoid of any properly geometrical meaning or implications. The thesis merely asserts that whatever properly geometrical propositions are accepted by geometers—definitions, postulates, axioms, demonstrated theorems—owe their a priori (necessary, universal) validity to their grounding in transcendental space. “Whatever geometers accept” thus includes what Riemann constructed in the nineteenth century, what geometers and topologists today construct, and whatever future mathematicians may succeed in constructing. For, again, Kantian transcendental space, being informed solely by transcendental concepts, is mathematically completely indeterminate, and so capable of accommodating any and all systems of geometry or topology, or indeed any and all mathematics whatsoever.5 What would transcendental space need to be in order to accommodate all mathematics? Consider Kant’s a priori psychologistic thesis that the natural numbers depend on both space and time: We cannot represent any number except by successive enumeration in time and then taking this plurality together in the unity of a number. But this taking together cannot occur unless I place them next to one another in space; for they must be thought as given simultaneously, i.e. as comprehended into one representation; otherwise,

234  Nature in Mind: Through Kant’s I’s this many does not constitute a magnitude (number). Simultaneity is, however, not possible to cognize other than that I can apprehend (not merely think) the plurality as given forwards and backwards outside of my act of composition. Thus, an intuition must be given in which the manifold is represented outside one another and (simultaneously represented) next to one another, i.e. an intuition of space.6 Let’s start with time. The natural number series is a succession—not, to be sure, the protracted kind occurring at a particular rate that we are aware of empirically in internal intuition through sensations, or the kind we are aware of externally when our senses encounter motion. For, on Kant’s a priori psychologistic analysis, succession is a pure formal intuition that is representable through productive imagination completely independently of empirical consciousness in visual, tactual, or any other kind of external or internal sensation. Without this non-sensational formal intuition, we could not represent succession of any kind, whether the purely aesthetic sort accessible in apprehension such as drops successively dripping from a tap, or the most abstract, formal, and language-dependent successions such as the alphabet or prime number series (“time is in itself a series and the formal condition of all series”).7 Moreover, the representation of succession required for representations like number is one in which each step in the series is objectively (i.e. sensation-independently) differentiated and determined in relation to every other. For example, if five is to come right before six and be separated from two by first three and then four—or 10,000,005 to precede 10,000,006 and be separated from 10,000,002 by first 10,000,003 and then 10,000,004—the undifferentiated, indeterminate formal succession made possible by purely aesthetic time will not suffice. Their succession must instead be grounded in a single, one-dimensional series composed of infinitely many purely formal times, each uniquely differentiated from and completely determined in relation to every other. Thus, according to Kant, it is not merely purely aesthetic formal intuited time but fully objective transcendental time— the result of determining the manifold of aesthetic time conformably to pure concepts of the understanding via transcendental synthesis—that needs to be incorporated into all representations of number. But number has a spatial as well as a temporal component. Formal time is simply succession. No two parts of any succession can be simultaneous since, by definition, they are separated by a time (“simultaneity is not a mode of time itself, as in time no parts are simultaneous, all are successive”).8 Distinct spaces, by contrast, are always simultaneous, never successive, and so all stand together in formal permanence—the same purely formal permanence that pertains to A, B, C, and the rest of the alphabet or any other abstract serial listing.9 Accordingly, it is owing solely to their formal permanence as spaces that I can prevent preceding members of a series from being lost from my representation with the

The Objective I 235 transition to the next: I simply “place them next to one another in space” and am thereby enabled to apprehend them together at one time, which, given formal time alone, I could not. The natural numbers are a case in point. To represent a formal temporal plurality as a number, I have to be able to unite them in a single act of thought, and this I can do only by juxtaposing each with every preceding member of the series in formal space, which confers on them the formal permanence requisite to represent simultaneously what otherwise is only representable successively. Nor is it enough merely to set them outside one another in purely aesthetic, formally intuited space. Each number’s space must be uniquely differentiated and completely determined in relation to every other, so that the formal space used in the representation of the number seven is the one and only space that is adjacent to the formal spaces used in the representations of six and eight, separated from the space of five by that of six (and no other), and separated from that of ten by that of eight first and then that of nine, and so on ad infinitum. It is therefore not merely the undifferentiated, indeterminate formal space of pure sensibility that is required to represent the formally simultaneous natural number series, but the fully objectified transcendental space that Kant explicated as the determination of the manifold of aesthetic space by pure concepts of the understanding via transcendental synthesis. Kant’s inclusion of formal space in the analysis of natural number is almost certain to be misconstrued unless one is careful to abstract completely from both sensation and geometry. In the case of the former, it was never Kant’s intention to depart from the conclusion of Locke, Berkeley, and Hume that subjective internal experience of the modes of time—empirical succession, simultaneity, and duration—is completely independent of the representation of space. For Kant, however, “simultaneity” here refers only to internal sensations that are exhibited by one and the same temporal appearance via a single act of empirical apprehension and has no meaning apart from the particular sensations apprehended (Chapter 8). The same is true with subjective “duration,” as when a succession of non-sensational temporal appearances enables one to represent not only the pain of one’s headache but its temporal persistence. Since numerical representations in mathematics are pure, they are only subjectively, not objectively, intertwined with the sensational flux of empirical consciousness. Consequently, the simultaneity and duration met with in internal sense experience are useless when it comes to representing the fully objective (sensation-independent) formal simultaneity and permanence that mathematics involves, where simultaneity/persistence need to be representable completely a priori—something Kant held only the pure space intuition can make possible. This consideration helps explain the inclination of platonists to regard numbers as eternal, with their simultaneity as co-permanents falling directly out of it. For Kant, by contrast, it simply means that numbers

236  Nature in Mind: Through Kant’s I’s fall within the purview of transcendental rather than empirical psychologism. In particular, the only means that creatures constituted like ourselves have of representing the a priori numerical simultaneity and permanence required by mathematics is by employing, first, our formal intuition of space to apprehend them in juxtaposition, one outside the other, and, second, the pure concepts of the understanding to differentiate and determine their spaces as objectively unique individuals, each with a sensation-independent identity that distinguishes it from all the rest. Thus did Kant regard transcendental space as no less essential to the representation of the natural numbers than transcendental time: without the former, the formal simultaneity and permanence of numbers would be unrepresentable; without the latter, numbers could not be apprehended as a series, and space itself would be quantitatively unrepresentable (“space is expressed through a time, e.g. an hour’s walk”).10 The other risk mentioned earlier is misconstruing Kant’s claim that objective transcendental space is essential to representations of natural numbers by geometricizing it. The problem can perhaps best be appreciated in relation to the number line. Insofar as space and time enter into the representation of these numbers, and space and time admit of being represented not merely as discrete but as continuous magnitudes (i.e. never a smallest or briefest), the same must be true of the natural number series, which quite literally becomes a continuous, unbounded one-dimensional series, i.e. a line extending infinitely outward in both directions and infinitely inward within every delimited interval. This, of course, allows the natural number series to be expanded to include zero and negative numbers and divided into rational and irrational numbers. However, it is just as vital not to confuse this transcendental number line with its geometrical image as it is not to confuse it with its visual or tactual image. The Kantian number line is not a geometrical construction but a transcendental synthesis.11 It is not constructible with rulerand-compass, does not extend through a plane, cannot be divided into segments or intersected by other lines, and so forth.12 The number line is ungeometrizable for the simple reason that only transcendental concepts, not geometrical (Euclidean or non-Euclidean), are employed in synthesizing it. Indeed, the absence of mathematical concepts of every kind means that it is incapable not only of being brought into any kind of image but even of being represented by the most abstract mathematical symbolism as well. It is instead a purely transcendental product of the transcendental synthesis of the manifolds of aesthetic formal intuition by means of the pure concepts of the understanding. These are concepts whose sole and entire content consists in logical forms of judgment; they are extralogical only in the sense that they negate the logical reversibility implicit in these forms (Chapter  7). As such, they “can never be met with in any intuition,” “contained in appearance,” or “brought into any image whatsoever”—algebraic-symbolic not excepted.13 And since these are the

The Objective I 237 only concepts involved in the synthesis of objective transcendental space and time, the transcendental number line must never be conflated with any mathematical representation of a line, be it geometric, algebraic, or any other.14 Nor are numbers the end of it. For Kant, wherever objectified formal succession simultaneity and permanence are essential to non-­sensational representation, be it in consciousness or language (sensible signs and conventional symbols), transcendental time and space are presupposed. It is, of course, impracticable to attempt to prove that this holds for all mathematics on a case by case basis, including not only post-­Kantian mathematics and any and all mathematics developed after us but all mathematics without exception, whether humanly possible or not. So, some additional examples will have to suffice. Mathematical functions may be supposed to presuppose transcendental space and time insofar as they range over a fully differentiated, completely determinate manifold of formally permanent and simultaneous arguments that they (formally) successively map onto a manifold of equally permanent and simultaneous values. The arguments may or may not be numbers. Their scope may be restricted to some but not all numbers; or restricted to something other than numbers; or, as in algebra, to symbolic generalizations of numbers; or, as in mathematical logic, left completely unrestricted. But whatever they are, transcendental time and space may be supposed to be essential to their representation in the same way they are essential to geometrical and numerical representation: not just aesthetically, for formal permanence, simultaneity, and succession, but also objectively (1) to uniquely differentiate and completely determine each argument and value in relation to every other as a transcendental precondition for being able to pick out individually what the function is to map and what it maps to, and (2) to uniquely differentiate and completely determine every successive application of the function as a transcendental precondition for picking each out individually in distinction from all the applications that precede or succeed it. The same considerations apply to set theory. The elements of sets are one and all objective individuals uniquely differentiated from and completely determined in relation both to all other elements of the sets of which they are members and to the elements of sets of which they are not members. As such, all need to be represented as formally permanent and simultaneous, while the rules that generate the sets and the operations that relate them need to be represented as applied in formal succession to their inputs in the same way as numerical functions (indeed, set theoretical rules and operations may be regarded as simply a non-numerical species of mathematical function). Propositional functions likewise: a formally permanent, formally simultaneous manifold of serially (alphabetically, by subscript, etc.) distinguished propositional symbols, operators that successively map simultaneous co-permanent individual truth possibilities

238  Nature in Mind: Through Kant’s I’s onto simultaneous co-permanent truth-values, quantifiers that successively range over a domain of simultaneous individuals and map them onto simultaneous truth-values, etc. I  could continue in this vein, but the basic thrust of the Kantian claim that objective transcendental space and time are presupposed in all possible mathematical representation, mathematical logic included, should now be evident: they are essential formal constituents of algorithms of every kind, loosely characterized as a (simultaneous) list of (co-permanent) individual formation rules (successively) applied to (simultaneous) manifolds of uniquely differentiated, completely determined (co-permanent) individual items, supplemented by individual operations that (successively) map the results onto similar such (simultaneous) manifolds of (co-permanent) individuals. Among other things, it should now be clear why we can now be quite certain that Kant would have classified post-Fregean mathematical logic as mathematics rather than logic. Logic proper is confined to AUA-­ generality and has no place in it for individuality (Chapter 7). By contrast, mathematical logic, like mathematics generally, presupposes not only objective individuality but specifically the transcendental kind in which the individuals are uniquely differentiated and completely determined in terms of formal permanence/simultaneity and succession by means of pure concepts of the understanding like unity, plurality, and totality (which, according to Kant, are alone capable of yielding concepts of number and the denumerable). Mathematical logic would consequently be impossible if we had different formal intuitions of sensibility than pure space and time or if we had to derive our pure concepts of the understanding from different forms of judgments than categorical, universal, affirmative, et al. It is therefore phylogeny dependent in ways that logic proper is not. The latter, being confined to AUA-generality, is completely indifferent to pure sensibility and its particular formal constitution; and while logic would be altered in details if we had different forms of judgments, it would in essentials remain the same: a logical universe constituted by the universal self-consciousness ‘I think’ bifurcated according to the content of AUA-concepts, with the resulting logical representations of that universe rendered combinable by logical forms of judgments, be they ours or other equipollent forms, in a synthetic unity of consciousness. In contrast, it is utterly impossible for us, as presently constituted, even to conceive, much less imagine, what would exist in place of the mathematics and mathematical logic we know now if we had different transcendental intuitions and/or concepts.

Objectivity I argued in Chapter 7 that Kantian non-linguistic propositional representation is immune to Wittgenstein’s critique of private language because it makes no pretense of being a purely psychological, convention-­ independent counterpart to vernacular language. I  also contended that

The Objective I 239 this remains true even with the addition of pure concepts of the understanding to the human psycho-propositional repertoire. However, the only representational role that could be envisaged for non-linguistic propositional thought at that point was to provide new ways to sort and organize the deliverances of associative imagination. If that were the end of it, whatever benefits these additional logical and extra-logical conceptual categorizations may contribute to creatures’ fitness would pale by comparison with the organizational powers conferred by language and be at most an evolutionary footnote. But its aspect changes dramatically with the recognition that the true role of the pure concepts of the understanding in non-linguistic propositional representation has nothing do with taking us a (relatively small) step closer to language and everything to do with furnishing the psyche with representations of fully objectified transcendental space and time. Kant regarded transcendental space and time as not only the first genuinely objective representations our minds produce but an indispensable ingredients of all others, be it in ordinary experience, science, or mathematics. Their constitutive role in such representations thus signifies a huge elevation in importance for pure concepts of the understanding, one that sets them well on the way to earning the appellation of “categories.” But what exactly is their contribution to the objectivity of transcendental space and time and objective representations generally? A representation of space, time, or anything else cannot be objective if its existence begins and ends with its presence to consciousness in sensation. But even if this is not the case, as with Kantian non-sensational appearances, a space or time still cannot count as objective if it can be singled out from others only through the external or internal sensations exhibited in it, as, say, the wet floor, the red fruit against the green canopy, the distant scream from the rear of the auditorium, the approaching siren, the flickering embers, or the persistent feeling of foreboding that triggers the recollection of a disturbing childhood incident. So, at least at a first approximation, nothing can count as objective unless it is fully independent of sensation as regards both existence and individuality. In Kantian transcendental aesthetic psychology, independence of the existence of sensation is conferred by the appearances that exhibit external and internal sensations in intuition. For as non-sensational, there is nothing to prevent us from representing appearances as preexisting the particular sensations they exhibit, continuing after them, or existing even if those sensations are never present in consciousness at all. But if we are to single out spatial and temporal appearances without relying on sensations, they have to have an individuality all their own, an identity representable completely independently of the sensations they exhibit. And it is precisely such objective individuality that is lacking in the undifferentiated, indeterminate manifolds of purely aesthetic formal space and time. Pure concepts of the understanding make possible a transcendental synthesis able to make good precisely this want. By uniquely differentiating

240  Nature in Mind: Through Kant’s I’s and completely determining every space and time in relation to each of the infinitely many others possible in formal space and time, transcendental synthesis serves to individuate them completely independently of sensation, and so objectively. This objectivity is not, to be sure, absolute and unqualified, since, on Kant’s analysis, pure concepts of the understanding are, in the literal (i.e. psychologistic) sense, logically bound up with the subjectivity of the ‘I think,’ and, in addition, may be as specific to our phylogeny as the human complement of senses.15 If “objectivity” is construed to connote total consciousness-independence, then Kantian pure concepts of the understanding are not objective. But since, on Kant’s accounting, this definition would exclude the transcendental space and time presupposed by empirical and mathematical representation of space and time, nothing would then count as objective except the unintuitable, inconceivable thing in itself, which “is not cognized, and cannot be cognized, through these representations, and in experience no question is ever asked in regard to it.”16 Kant thus took the view that the only objectivity our cognition truly requires is the kind that renders representations independent of sensation as regards both existence and individuality, and that anything beyond that is otiose metaphysical hyperbole.

Toward a Solution to the Heterogeneity Problem: The Elements of Transcendental Synthesis We have seen that objective transcendental space and time have to be understood as able to accommodate a physics with not just four fundamental forces active in three dimensions but any number of forces active in any number of dimensions, including infinitely many of both. We saw too that they must be able to accommodate objective formal-mathematical succession, simultaneity and permanence of every kind, from Euclidean and non-Euclidean geometry to fields as diverse as number theory, analysis, algebra, and mathematical logic. We must therefore conceive them so as to exclude everything in our representations of space and time that derives not only from sensation but from physics and mathematics as well. Moreover, unlike the space and time of physics and mathematics, transcendental space and time are free of all dependence on language (which is not to say there could not be creatures capable of turning transcendental space and time to mathematical and scientific account purely psychologically, i.e. without depending on language). And since objective transcendental space and time are products solely of formal intuitions of sensibility and pure concepts of the understanding, their representation need never be supposed to stray over the evolutionary line separating gradualist a priori psychologism from massively saltationist platonistic metaphysics. Kant’s account of transcendental space and time in terms of the transcendental synthesis of the purely aesthetic homogeneous manifolds of

The Objective I 241 formal intuition by means of the pure concepts of the understanding is the core of his analysis of objectivity, including not only the objects of mathematics and science but the sense-divide transcending external and internal objects of ordinary cognitive experience as well. The principal explanatory obstacle he confronted was the radical heterogeneity between the two sorts of representation. On the one hand, pure concepts of the understanding have no content save that contributed by logical forms of judgments, and so are in no sense representations of sensible things, their doings, or their relations. On the other hand, the manifolds of formal intuitions have no contents save the non-sensational forms contributed by productive imagination, are thence purely sensible (non-intellectual) and aesthetic (non-logical), and so are entirely devoid of all the features that make representations propositional or, as with AUA-concepts, relatable in propositions. Given their radical heterogeneity, how can pure concepts possibly be supposed to determine the manifolds of formal space and time, as Kant’s psychologistic explications of objectivity require? Without a means of overcoming it, pure concepts of the understanding would lack all objective meaning and so could not be used to psychologistically explicate traditional metaphysical categories (fundamental concepts of objects) like cause and effect (Chapter 7). This would not only leave the latter wide open to Humean anti-platonistic skepticism (Chapter  5) but make Kant’s project of psychologistically explicating nature and natural law a nonstarter as well. He therefore needed to show that and how transcendental synthesis—and it alone—is capable of bridging the gulf separating pure concepts of the understanding from the manifolds of pure sensibility. Identifying the planks with which to build the bridge and fitting them together with sufficient strength to bear the full explanatory load plunged Kant into what even for him were unprecedented psychologistic depths and complexity.17 His account is grounded on what he deemed the supreme principle in this domain, pure apperception, which, as we have seen, is a self-consciousness with two, analytically related sides. One is the analytic unity of apperception (AUA) identified by him with the universal self-consciousness ‘I think,’ and so the source of the universality that alone can convert non-concepts into propositionally relatable concepts (i.e. representations universal in form, whatever their content). The other is the original synthetic unity of apperception (SUA) identified by Kant with a consciousness having an individuality commensurate with the universality of the ‘I think,’ which, in us, is the unity of sensibility constituted via the formal intuitions of space and time and their infinitely iterable homogeneous manifolds. The infinite totality of spaces and times within and without spaces and times of pure sensibility cannot be directly determined by the pure concepts of the understanding because of their radical heterogeneity. Yet, there is nothing to prevent a creature from incorporating AUA into its

242  Nature in Mind: Through Kant’s I’s representation of the synthesis responsible for producing these spaces and times, resulting in the representation of a further synthesis capable of generating a commensurate infinity of AUA-concepts having these spaces and times as their contents and ‘I think’–universality as their form. Each concept formed in this way will consequently be a perfect one-for-one proxy for the corresponding space or time, containing numerically the same content and differing from it solely by virtue of participating in the purely logical universality of the ‘I think’—a pure consciousness devoid of all content—that enables it to be incorporated into propositions. AUA-concept proxies of the spaces and times apprehended within formally intuited space and time are the first plank in Kant’s explanatory bridge. The rest are concerned with synthesizing these proxies with predicates in categorical propositions, combining pairs of these propositions to form more complex hypothetical propositions, and finally determining the entire propositional network conformably to the pure concepts of the understanding. It is this final plank that constitutes transcendental synthesis proper, with ‘transcendental’ used to indicate the role of pure concepts of the understanding in the objectification of the manifolds of formally intuited space and time. For while it is one thing for pure concepts to determine the AUA-concept proxies of the spaces and times in these infinite manifolds and another thing entirely to determine those intuitions themselves, the determination of the former cannot fail to carry over to the latter because the intuitions and their AUA-concept proxies are indistinguishable in content, differing solely by the addition of the contentless universal self-consciousness ‘I think’ to the AUA-concepts. Since the heterogeneity problem relates specifically to the difference in content between purely logically defined pure concepts and purely aesthetically defined formal intuitions, Kant could justifiably claim that transcendental synthesis successfully transcends their heterogeneity.18 But both in order to help clarify the kind of solution this problem requires and as a first pass at the formidable topic of transcendental synthesis itself, I shall briefly survey each plank of Kant’s explanatory bridge in turn, picking up where I left off: the application of the logical synthesis of categorical judgment to the AUA-concepts of spaces and times synthesized from the manifolds of formally intuited space and time. What would the predicates of these categorical propositions be? Spaces and times within formal space and time do one thing and one thing only: exhibit possible sensations. Each space is capable of exhibiting any external sensation whatsoever: red, green, blue, light blue, lighter blue; wet, hot, soft, or itchy; sugary, peppery, bitter, sour; barking, screaming, whispering; fragrant, pungent, stinky; and this is merely a partial list of just the human complement of sensations. Since space, for Kant, is a non-sensational representation, it could equally exhibit kinds of external sensation not found in humans, and do so in every degree, blend, and combination. Their non-sensational character also enables spatial

The Objective I 243 appearances to exhibit unsensed and even insensible realities as well, including those that require mathematics and science to represent and detect: the pressure where the earth’s core meets the mantle, the heat at the center of the sun, quantum fields, realities that exist only in still unknown extra dimensions, forces yet to be discovered or undiscoverable by human means, etc. So too times: each is capable of exhibiting any internal sensation whatsoever, be it a synthesis (apprehension, association, conceptual recognition), recollection, fantasy, emotion, desire, volition, or any other kind, whether found in humans or not, actual or possible. In addition, spaces and times can represent the absence of sensation, or indeed of reality of any kind (empty space or time). There are, in short, a potentially inexhaustible infinity of realities and voids, present or not in sensation, that each spatial appearance in formal space and each temporal appearance in formal time are capable of exhibiting. And simply by attaching the universality of the ‘I think’ to their representation, an equally inexhaustible infinity of AUA-concepts of exhibitive possibilities becomes available to represent as predicates of possible categorical propositions for purposes of transcendental synthesis. I want to stress that in none of these cases are infinities literally represented. What instead is represented are infinitely iterable aesthetic and logical form-based synthesizing operations whose scope extends to all representations, possible no less than actual, with unity provided by the formal individuality (SUA) and universality (AUA) of the pure selfconsciousness in which they take place. Since original SUA precedes AUA, its individuality is exclusively concerned with the aesthetic infinities of formal intuition. AUA, by contrast, is a logically universal consciousness able to accompany each and every possible representation and thereby convert it into an AUA-concept. It is thus a simple matter for a creature to incorporate AUA into both aesthetic syntheses: the one responsible for the infinite totality of the manifold of formal space and/or time— i.e. every spatial or temporal appearance productive imagination may synthesize—and the one responsible for the infinite totality of exhibitive possibilities pertaining to each such appearance—thereby yielding AUA-concepts corresponding to each. An extended SUA can then be represented by means of the logical form of categorical judgment that enables us to synthesize the AUA-concepts derived from the manifolds of formal space and time as subjects with the AUA-concepts derived from their exhibitive possibilities as predicates to yield an infinity of categorical propositions. And this SUA can be extended still further by using the logical form of hypothetical judgment to relate each of this infinity of categorical propositions to every other as ground to consequence in an infinity of hypothetical propositions. In this way, the infinitely iterable spaces and times of our formal intuition are first transformed into an infinite system of (non-linguistic) categorical propositions built entirely from AUA-concepts that differ from these spaces and times and their exhibitive

244  Nature in Mind: Through Kant’s I’s possibilities solely by virtue of the addition of the contentless formal logical universality of the consciousness represented in the ‘I think.’ And then, in similar fashion, a no less infinite system of hypothetical propositions is synthesized from these categorical propositions to form an allencompassing (non-linguistic) propositional network. The last plank in the bridge of synthetic unities Kant erected to span the chasm between pure concepts of the understanding and the otherwise incommensurable manifolds of formal intuitions is the incorporation of these concepts into the synthetic procedures responsible for generating the systematic network of propositions just described. Considered purely logically, the categorical and hypothetical propositions constituting the network are one and all reversible. Categorical form leaves us free to reverse the logical positions of the subject and predicate concepts so that the original predicate becomes the subject and the original subject its predicate. The pure concept of substance and accident derived from this logical form eliminates precisely this freedom (Chapter 7). By applying it to the conceptualized pure spaces that occupy the subject position in each of the categorical propositions in the system, they can be represented as final subjects of their predicates, i.e. subjects that can never in turn become predicates of anything else. This categorization is fitting because, in intuition, spaces are all (formally) simultaneous and permanent. Since successive times are neither, the AUA-concepts derived from them are necessarily categorized as accidents, i.e. final predicates. This means that in all categorical propositions in which AUA-concepts of spaces and times are related, the spaces must be represented as final subjects (substances) and the times as final predicates (accidents) (“succession itself cannot be represented otherwise than through the permanent . . . with which the successive is simultaneous, i.e. space”).19 But since this status does not prevent AUA-concepts of times from occupying the subject position in propositions not involving spatial final subjects, AUA-concepts derived from the exhibitive possibilities of pure times can still be represented as their predicates in transcendental synthesis. Yet, the pure concept of substance and accident can do nothing by itself to differentiate and determine the conceptualized spaces and times it converts into final subjects and final predicates. This happens through the pure concepts that are prima inter pares so far as transcendental synthesis is concerned: cause and effect and possibility-impossibility. Their sole and entire content consists in the determination of hypothetical propositions so as to fix the ground-consequent order of the constituent propositions, determining the one in the grounding position as being always and only the ground of the one in the consequent position and the one in the consequent position as being always and only the consequent of the one in the grounding position. Causal series are formed when the ground is further determined as the irreversible consequent of another proposition, this proposition the irreversible consequent of yet another,

The Objective I 245 and so on, and/or when the consequent is further determined as the irreversible ground of still another proposition, this in turn the irreversible ground of a further proposition, and so on. In such series, a proposition is the immediate ground of one proposition, the ground at one remove of another, the ground at two removes of still another, and so on, as well as the immediate consequent of a further proposition, the consequent at one remove of some other proposition, the consequent at two removes of still another, and so on. Each propositional link in the causal chain is thereby uniquely differentiated from and completely determined in relation to every other. If the propositions concerned in a causal series are those that have AUA-spaces and AUA-times as their subjects, the same differentiation and determination will result but now possessed of a spatial and temporal significance. For example, take any two times in purely aesthetic formal time. Since both are comprehended within the individuality of formally intuited time, they are ipso facto successive. But with nothing to connect them, their order of succession can equally well be represented reversewise; nothing determines their succession as immediate or as separated by intervening times; no third time is determined as intervening or not intervening between them, as nearer the one than the other, as containing or being contained by one or both, as bounded or bounding either, etc. This is why times apprehended in purely aesthetic formal intuition are completely undifferentiated and indeterminate, and it is precisely what the pure concepts of cause and effect and possibility-impossibility change. With one time connected to the other as (non-final) subjects of two categorical propositions irreversibly related as immediate ground to immediate consequent, their relative position in formal time is irreversibly determined—one permanently fixed as predecessor, the other as successor—as is the immediacy of their positions in the series of times— no intervening times, i.e. temporal contiguity—and the order (irreversible direction) of time—from the time of the grounding categorical proposition to the time of its consequent. However, if instead of being immediately connected by a ground-consequent relation, there was a proposition with a third time as subject of a proposition that was the immediate consequent to the first and immediate ground to the second, then the two times would be determined as succeeding at one remove with the third time intervening. Causal series thus permit every time in formal time to be differentiated from every other as either its immediate successor or predecessor, its successor or predecessor at one remove, at two removes, three, or however many (including infinitely many removes). And in this way the completely undifferentiated, indeterminate manifold of purely aesthetic formal time is transformed into fully objectified, irreversible, linear transcendental time. What about space? As comprehended within purely aesthetic formally intuited space, spaces are one and all in juxtaposition. Yet, with nothing

246  Nature in Mind: Through Kant’s I’s to connect them, their juxtaposition is not objectively (irreversibly) determined as either immediate or non-immediate, and no third space is objectively (irreversibly) determined as interposing between them, as nearer the one than the other, as containing or being contained by one or both, as bounded by or bounding them, or anything else. Now suppose the two formally (i.e. non-protractedly) successive times immediately related by a causal (i.e. irreversible ground-consequent) relation considered earlier are final predicates of two of these juxtaposed spaces. Since the immediacy of the relation precludes the intervention of a time in the transition from ground to consequent, it equally precludes the interposition of any space between the juxtaposed spatial final subjects, thus determining their juxtaposition as adjacency (spatial contiguity). However, if a causal relation involving a third time as subject intervened between the causal relations involving those times and that time was itself the final predicate of a third space, then there would be an intervening space between the two spaces, and a time would have to elapse before anything could pass from the first space to the second; and so on for a fourth, a fifth, or however many more causal relations involving times predicated of different spaces. In this way, causal series permit every space in formal space to be objectively differentiated from every other as either adjacent or not, and if not, then at one, two, or however many (including infinite) removes along any particular (formal) spatiotemporal route from one to the other. And while distance, motion, direction, and dimension have yet to be explicated, the foregoing should suffice to give some idea of how transcendental synthesis serves to transform the completely undifferentiated, indeterminate manifold of purely aesthetic formal space into the fully objectified manifold of transcendental space. Before proceeding, however, I want to focus more closely on the crux of Kant’s solution to the problem posed by the radical heterogeneity between the purely logical content of pure concepts of the understanding and the purely aesthetic content of formal intuitions of sensibility. Although pure concepts are actually only able to directly determine the system of propositions formed from AUA-conceptualizations of the spaces and times within formally intuited space and time, they nevertheless can be regarded as indirectly determining these very spaces and times themselves on the ground that there is no real (contentful) difference between them. For if the only thing that distinguishes these AUAconcepts from the space or time from which they derive is the addition of the contentless, purely formal consciousness ‘I think’, then, so far as their content is concerned, the former are not merely homogeneous with the latter but quite literally identical. So, how can the application of pure concepts of the understanding to this system of concepts fail to carry over to the corresponding spaces and times as well? After all, such an extension remains well within the precincts of a pure selfconsciousness (SUA/AUA) that encompasses all representations equally,

The Objective I 247 intuitions and the sensations they exhibit no less than concepts and propositions. It was on the basis of this numerical identity in content that Kant concluded that the subordination of concepts formed by attaching AUA to the manifold of formally intuited space and time to the pure concepts of the understanding not only can but must extend to the spaces and times themselves. Pure concepts of the understanding may be incapable of directly determining the manifold of formal intuitions. Yet, by directly determining the totality of concepts with that manifold as their sole and entire content, their determination must be regarded as applying to the manifold itself as well. Hence, notwithstanding their radical heterogeneity, pure concepts of the understanding are the basis of a transcendental synthesis whereby the undifferentiated, indeterminate manifold spaces and times within purely aesthetic formal space and time become the uniquely differentiated, fully determinate objective spaces and times of objective transcendental space and time. And with objectivity thus incorporated into their meaning, the way is opened for the psychologistic explication of the categories of traditional metaphysics in terms of pure concepts of the understanding (Chapter 10).

The Transcendental Synthesis of Space and Time I shall now resume the sketch of transcendental synthesis begun earlier. Granted that the sole and entire meaning of the pure concepts of cause and effect and possibility-impossibility in relation to hypothetical propositions consists in eliminating the logical freedom to reverse the order of ground and consequent, while at the same time both restricting the ground to being the ground of that consequent and no other, and restricting the consequent to being the consequent of that ground and no other (Chapter  7). Since a self-grounding hypothetical cannot be determined as irreversible, the concept can apply only to hypotheticals that relate different propositions. But if each ground has a unique but different consequent and each consequent a unique but different ground, then in any causal series, even one of infinite length, none can recur. Moreover, if the pure concept of possibility-impossibility prevents any ground from having multiple immediate consequents and any consequent from having multiple immediate grounds, causal series are in all cases unbranching, never converging irreversible linear sequences. Thus, a series of groundconsequent relations determined by the pure concepts of cause and effect and possibility-impossibility results in an equally formal ever-changing, unbranching, unconverging, irreversible linear sequence. If, as indicated earlier, ground-consequent relations determined by the pure concept of cause and effect are formed from a system of propositions that derive their subjects exclusively from pure times in formal time and their predicates exclusively from possibilities of temporal exhibition,

248  Nature in Mind: Through Kant’s I’s then the formal ever-changing, irreversible, linear series that results can be none other than objective temporal succession. For while the pure concept cannot directly determine intuitions, it can apply to hypothetical propositions synthesized from categorical propositions consisting of (non-final) subject AUA-concepts formed directly from the infinite totality of possible times, and predicate AUA-concepts formed from any of the infinite totality of each time’s exhibitional possibilities. Since the only difference between the conceptualized times in this lineage and the purely aesthetic intuitive times comprehended within formal time is the purely formal logical, non-contentful addition of AUA to the former, the ordering of these concepts of times into a single ever-changing, unbranching, unconverging, irreversible causal lineage infinite in both directions carries over to the intuitions from which they are derived. And the result is the objectification of aesthetic time as transcendental time. To illustrate this, take the causal relation “Tm is pm” ⇒ “Tn is pn,” with ‘T’ signifying an AUA-concept of a time ‘t’ in formal time, ‘p’ a concept of a possible temporal exhibition, and ‘⇒’ a hypothetical relation determined as irreversible by the pure concepts of cause and effect and possibility-impossibility. It constitutes an irreversible, non-repeating, linear sequence in which the time featured in the grounding proposition is determined as immediately preceding the time in the consequent proposition in the order of times. Now suppose the series continues infinitely in both directions as indicated: . . . ⇒ . . . ⇒ “Tm−2 is pm−2” ⇒ “Tm−1 is pm−1” ⇒ “Tm is pm” ⇒ “Tn is pn” ⇒ “Tn+1 is pn+1” ⇒ “Tn+2 is pn+2” . . . ⇒ . . . ⇒ . . .20 Since the AUA-concept T derived from formal time t is literally identical with t in content, this causal series cashes out as a genuinely temporal series, with tm determined as immediately preceding tn in the order of times, tm−1 determined as immediately preceding tm, tn as immediately preceding tn+1, tm−2 as immediately preceding tm−1, tn+1 as immediately preceding tn+2, and so on, resulting in a single, always unbranching, never converging, ever-changing, irreversible linear series comprising the infinite totality of possible times extending without end in both directions. And the transcendental time thus synthesized qualifies as objective because each time possible within formally intuited time not only exists sensationindependently but is individuated by its unique place in the series, entirely a priori, without reference to any sensation, internal or external. Since conceptualized transcendental times are final predicates, transcendental synthesis must map their formal succession onto final subjects determined conformably to the pure concept of substance, and so onto formally simultaneous co-permanent spaces. If, in addition to temporal predicates, we consider only those causally related exhibitive predicates that pertain to different spaces, a transition from the time of the grounding proposition to the time of its immediate consequent will at the same time constitute an immediate transition from the space of the first to the space of the second. Since an immediate transition from one space to

The Objective I 249 another constitutes adjacency (spatial contiguity), a causal series consisting of times all of which are predicated of different spatial final subjects serves to set the first space in the series adjacent to the second, the second adjacent to the third, and so on. Of course, the order of formally permanent and simultaneous spaces, unlike that of successive times, needs to be reversible. How can that be achieved given the irreversibility of the ground-consequent relations that compose causal series? The implication of this irreversibility is that no ground and no consequent may recur within the same series. Moreover, since in aesthetic formal space and time nothing is general, everything individual, AUA-conceptualized exhibitive predicates are as unique as the exhibitive possibilities of the space or time from which they are derived, and thence can occur only once in any causal series. So, how can formal causal succession across spaces ever proceed reversewise, and do so without reverting to its logical meaning, thereby sacrificing the extra-logical necessity requisite for Kant’s refutation of Humean skepticism? The answer is via irreversible ground-consequent relations connecting different exhibitive predicates of the same spatial final subjects. For example, the route from space A  to space M back to space A  can be represented as follows: “A  is pa” ⇒ “B is pb” ⇒ “C is pc” ⇒  .  .  . ⇒ “M is pm” ⇒ “M is pn” ⇒ “L is po” ⇒ “K is pp” ⇒ . . . ⇒ “B is pz” ⇒ “A  is paa.” Since each spatial final subject has an inexhaustible infinity of possible exhibitive predicates, this causal series can be indefinitely extended so that the route from A to M and back is retraced an unlimited number of times without ever needing to repeat the same predicate, thus yielding a purely formal yet fully objective—because sensation-independent—representation of perpetual oscillation between A  and M (other motions can be provided with purely formal objective representations in a similar fashion). The result is not, to be sure, linear in the geometrical sense. Causal series are not sequences of points, discrete segments, or any other properly geometrical unit from which an actual image of a line apprehensible in intuition might be constructed. No geometrical or other mathematical concepts are involved in their synthesis. Rather, they are purely transcendental propositional compositions synthesized from irreversible ground-consequent relations between substance-accident relations of AUA-concepts determined as substances (final subjects) and accidents (final predicates). Nevertheless, these transcendental syntheses are fully as spatial or temporal as any empirical or mathematical representation of space or time. It should thus be apparent why Kant deemed formal causal series of the kind just described to constitute the purest, most fundamental objective representations of one-dimensional space, length (A + B + C + . . . + M + L . . . + A + B . . .), and distance (e.g. from C to G = C + D + E + F + G). To get additional spatial dimensions, we simply need to represent causal series issuing from space A that, in addition to the route through

250  Nature in Mind: Through Kant’s I’s space B to space C, etc., proceed from A to space B´ to space C´, etc., from A to B´´ to C´´, etc., from A to B´´´ to C´´´, etc., and so on. Since, in purely aesthetic formal space, juxtapositions are possible within and without juxtapositions without end, we have to imagine an inexhaustible infinity of causal series issuing from A  through an inexhaustible infinity of adjacent spaces, and from there, one by one, to all other spaces without exception, no matter how distant in the series, even infinitely distant. This defines a second transcendental spatial dimension and provides for a richer notion of distance. For in addition to the causal series just described, the inexhaustible infinity of exhibitive predicates allows us to describe infinitely many more or less direct routes from any space to any other. On one path, the route from space A to space B involves no distance it at all, i.e. no intervening spaces to pass through. But another route passing from A to B´ might go through space N´´´´, space U´´, space Q´´´´´´, until it ends up at space B. Accordingly, the distance between one space and another in more than one dimension can be represented as the route from one to the other that passes through the fewest intervening spaces (i.e. the minimum of irreversible ground-consequent relations connecting spatial final subjects). Moreover, with an infinity of paths from A into adjoining spaces in addition to the route directly into B, infinitely many different directions out of A can be represented as well. The transcendental (a priori psychologistic) explication of higher dimensionality is somewhat more subtle. Given the inexhaustible infinities of concepts and propositions at which transcendental synthesis is directed, the pure concept of cause and effect confers on each space (final subject) in formal space an inexhaustible infinity of possible adjacent spaces. The most natural way to think about the routes into these spaces is in terms of three dimensions since that is all we can visualize, tactually image, or otherwise empirically imagine to ourselves. But while this may be a genuine limitation of human sensation-based powers of reproductive imagination, can it also be supposed to restrict the non-sensational space of pure productive imagination and its determination as objective transcendental space through transcendental synthesis? Spatial differentiations and determinations in the latter are purely propositional, a matter entirely of restrictions of logical freedom represented through pure concepts of the understanding. No objective space representable empirically or geometrically with compass and straightedge is or could be produced by a transcendental synthesis utilizing only concepts that “can never be met with in any intuition,” “be intuited through the senses,” “contained in appearance,” or ever “be brought into any image whatsoever.”21 Yet, for precisely this reason, there can never be any basis for supposing that transcendental representations of space are restricted to three dimensions the way spaces represented in empirical imaging-dependent ways are. How then does transcendental causal synthesis generate additional dimensions? Given the infinity of adjacent spaces it produces for every

The Objective I 251 space in formal space, there will be a commensurate infinity of directions radiating out from each into the spaces adjacent to it. However, in contrast to empirical imaging-dependent representations of spaces, this infinity can no more be exhausted by the infinity of directions radiating out from a space in three dimensions than it can be exhausted by the infinity of directions radiating out from it in two. Instead, the spaces ordered by this synthesis can accommodate an infinity of directions in an infinity of dimensions, and so are capable of supplying fully differentiated, completely determinate, formally permanent/simultaneous n-dimensional manifolds of the kind presupposed by even the most recondite geometrical, topological, and purely symbolic (e.g. algebraic) mathematical constructions. Still, before the representation of transcendental space can be completed, we need to take account of the merely possible character of the irreversible ground-consequent relations that compose the causal network: until the predicative possibility expressed in the grounding proposition is actualized, the consequent too remains merely possible, and no transition from one space (final subject) to another is yet effected. Actualization depends on the actuality of a spatial final subject with the hypothetically grounding exhibitive predicate, which here simply means changing the determination of the categorical proposition in the ground position from conformity with the pure concept of possibility-impossibility derived from the logical form of problematic judgment to conformity with the pure concept of existence-nonexistence derived from the logical form of assertoric judgment (Chapter 7). If represented as existent, then its consequent will ipso facto be determined as existent as well (modus ponens in inference form), and the transition—the route—from the space represented in the grounding proposition to the space represented in the consequent will be actualized. But if the grounding proposition remains merely possible, its consequent will remain unactualized, with the result that the route must be represented as non-existent. Accordingly, by representing AUA-concepts of spatial final subjects and their predicates in propositions determined as possible rather than existent, one can limit objective transcendental space to as few directions in as few dimensions as one’s representational purposes require. One could, for example, use the pure concept of existence-nonexistence to represent objective formal space as consisting of as few as two spaces simply by stipulating that none of the inexhaustibly infinite totality of predicative possibilities of spatial final subjects related in irreversible ground-consequent relations are existent except the ones that allow the one space to be determined as adjacent to the other and the other to be determined as adjacent to it. Or one might represent an actualized network of formal spaces consisting not of causal series but solely of single causal relations connecting each space to one other and no more, so that progress along all routes would cease as soon as it began, with the result

252  Nature in Mind: Through Kant’s I’s that the space thus synthesized would be universally discontinuous. Or one might represent the actualization of a formal space in which every possibility of stopping and continuing on, including infinitely, exists simply by representing the modality of the infinite totality of grounding propositions as all changed from merely possible to actually existent. Admittedly, a fully actualized transcendental space would make every space in the network adjacent to every other in some dimension (including those that in other dimensions are infinitely removed or secreted in the other’s deepest interior), and so may seem completely unrecognizable as space at all. But this is a feature, not a bug. It means that each formally existent space is uniquely differentiated from and completely determined in relation not only to all other formally existent spaces but also to every one of the inexhaustibly infinite totality of other possible spaces. This makes them individuals, each with its own distinctive identity, in the fullest possible objective sense short of complete consciousnessindependence: independence of sense, memory, physical standpoint, conventional signs and symbolic modes of representation, and every other kind of subjective limitation other than dependence on pure formal intuitions and pure concepts of the understanding. And this maximally objective individuality is retained by each and every space or time even when some of its possible relations to other spaces are not actualized—or indeed if none of them are. To be sure, pure concepts of the understanding are directly concerned solely with the AUA-conceptualizations of the spaces in formal space. But given that the only difference between these concepts and formal spaces themselves is the addition to the former of the purely formal logical universality of the ‘I think’ (a consciousness, not a content), the ordering of these concepts in a single, all-encompassing propositional network determined by transcendental synthesis inevitably carries over to their contents as well, i.e. the intuitions from which they are derived. And the result is the objectification of purely aesthetic formally intuited space as transcendental space.

An Analogy From Mathematical Physics The foregoing is no more than a partial sketch, in broad strokes, of Kantian transcendental synthesis.22 Even so, the transcendental space it presents is sufficiently complex that its lineaments may prove elusive (transcendental time, being one-dimensional, poses fewer problems). To help rectify this, an analogy may be helpful. It is drawn from a passage in a book for the general reader concerned with “quantum graphity”—a way of understanding what Einstein called “spooky action at a distance” inspired by graph theory. I cite it at length because it describes a view of space as a network that parallels Kant’s transcendental synthetic conception in so many ways as to be almost uncanny:

The Objective I 253 The concept of space as a network goes back to the 1960s and the brainstorms of such innovative (and iconoclastic) theorists as John Wheeler, David Bohm, Roger Penrose, and David Finkelstein. Wheeler, for one, imagined taking a bucket of “dust” or “rings”— primitive grains of matter that do not exist within space, but simply exist—and stringing them together to form space.  .  .  . Today one of its strongest champions is Fotini Markopoulou, who pictures the stringing-together process as a graph akin to . . . Facebook diagrams. She and her colleagues call their approach “quantum graphity.” . . . The links between the elementary grains are as simple as can be. Two grains are either connected to each other or not, like Facebook users who can either be friends or not—just an on-or-off relationship. The resulting network looks like a string-art craft project in which you hammer nails (representing the grains) into a sheet of wood and stretch threads (the links) among some of them. . . . To breathe life into the network—to give it the capacity to transform and evolve— Markopoulou and her colleagues suppose that the links switch on or off depending on the amount of available energy . . . The energy maxes out in a fully interconnected network, where every grain is linked to every other grain. In such a network the principle of locality doesn’t hold; you can go from any grain to any other grain in one hop, without passing through any intermediate points. The network lacks the hierarchy of relations—near versus far, small versus big—which is characteristic of space. You can’t subdivide it into separate chunks; it’s an indivisible whole. “This thing has no notion of locality . . .,” Markopoulou has explained. “If you just put out your hand, you reach everybody in the whole universe.” To see why the high-energy network is not spatial, try assigning locations to the grains. Every grain has to be equidistant (a single hop) from every other. For the first three grains, that’s no problem: arrange them in an equilateral triangle. Four can be stacked in a pyramid. But where does a fifth go? There’s nowhere equidistant to the first four points, at least not within ordinary, three-dimensional space. You need a four-dimensional pyramid. In fact, each additional grain requires a whole new dimension of space. Before long, you enter an ultra-higher dimensional realm beyond our capacity to visualize. And most of that vast venue is wasted: the network is only one hop wide in any direction and does a good impression of a balled-up spider’s web. So although you might still talk of the network as existing within space, it’s not the kind of space we want: three dimensions that extend as far as we can see in every direction and provide an economical description of the relations among objects Lower-energy patterns are a different story. They’re just what we want. Each grain connects to just a few others, forming a regular grid like a honeycomb or woven fabric. The notion of distance regains

254  Nature in Mind: Through Kant’s I’s meaning: some grains are close together, the rest far apart. The network is nice and roomy. The principle of locality holds: for an influence to go from one place to another, it can’t hop straight there, but must work its way through the network. The passage of the signal takes time, which would explain why the speed of objects through space is limited (by the speed of light).23 The difference between an inexhaustible infinity of primitive grains of matter that simply exist and a space of any magnitude and number of dimensions constituted by links between these grains parallels the Kantian distinction between the inexhaustible infinity of undifferentiated, indeterminate spaces within purely aesthetic formally intuited space and the maximally interconnected network of objectified transcendental space capable of being actualized to any magnitude and number of dimensions. Similarly, the simple on-off character of the amount of available energy at any link closely parallels the differential on-off actualization of grounding propositions effected wherever the modal concept existent-­nonexistent is substituted for possible-impossible. In particular, just as quantum graphity’s available energy may suffice only for a finitely divisible (to the Planck length but no further), finitely extendable three-dimensional space (the volume of the Big Bang universe), so too the predications of spatial final subjects represented through transcendental synthesis as existent may suffice to actualize only a finitely divisible, finitely extendable threedimensional transcendental space. Finally, the balled-up spider’s web that results when quantum graphity’s available energy maxes out parallels the case where the concept existent-­nonexistent is used to actualize all possibilities of connection within the network of transcendental space so that each of the infinite totality of spaces is adjacent to every other in some dimension. Thus, both quantum graphity and Kantian transcendental synthesis define the objective individuality of spaces exclusively in terms of their positions in a network constituted by all the links possible within it, and define the existence of these spaces in terms of which of their possible links happen to be actualized.24 Of course, the whole point of Markopoulou’s use of quantum graphity is to deny the spatiality of both the grains and the balled-up spider’s web. But this can be reconciled with Kant’s conception of space by taking into account quantum graphity’s implicit assumption that representations of space are all founded on experience. Take the example of the four-dimensional pyramid. This is supposed to not be spatial because we have no experience of such a figure, and experience provides us no materials with which to imagine it (visualize it, tactually imagine moving in the fourth dimension away from one part to another part of the figure, etc.). But Markopoulou does not take into account a view like Kant’s that considers purely aesthetic spatial appearances (the counterpart to her “primitive grains”) to be not sensations but non-sensational

The Objective I 255 intuitions generated in productive imagination, the spatial character of which is wholly independent of sensation-based empirical imagination and its limitations. Similarly, she neglects the possibility of conferring objective individuality on spatial appearances entirely a priori by means of pure concepts of the understanding that derive their meaning neither from mathematics nor from experience but from Aristotelian logic as explicated via Kant’s a priori psychologism. To be sure, even if one grants Kant’s transcendental conception of space, at least for argument’s sake, one may still deny that a fully actualized transcendentally synthesized space (“balled-up spider’s web”) can legitimately be regarded as “space.” Yet, just because every space can be reached from any space in only one hop does not mean it must be reached that way: other routes from one space to another involving many hops, even infinitely many, are likewise delineated in it. Suppose, for example, that in addition to being formally fully actualized, transcendental space were fully actualized materially. Does it follow from this that electromagnetically constituted creatures like us would be able to reach every part of it instantaneously? Not if quantities or other aspects of the four familiar forces that have yet to be encountered and/or additional, humanly undetectable forces differentially actualized links between the spatial final subjects composing it. The physically real three-dimensional space familiar to us might then be simply the parochial by-product of our electromagnetic natures, and our consequent acquaintance with only four fundamental forces operating in only three dimensions in quantities and other ways that happen to be observable by us. Other aspects of these forces and/or a fifth, sixth, or more additional forces might, unbeknownst to us, actualize other links, so that physical space in truth consisted of ten, a hundred, or any number of dimensions. In that case, there would be every reason to regard even an infinite dimensional, albeit mostly undetectable physical space as a more purely objective realization of space than the three-dimensional realization familiar to humans. Similarly, the light speed limit would then have to be regarded not as an objectively absolute bound of motion through physical space but as a merely human limitation, no different in principle from the human incapacity to produce sensation in response to ultraviolet light. Indeed, given that quantum entanglement is a fact of nature, it should be deemed a plus for Kant’s psychologism that the objectified space of transcendental synthesis is quite capable of accommodating “spooky action at a distance.”

Objective Unity of Apperception That transcendental space and time can furnish the kind of n-dimensional, formally simultaneous, co-permanent manifolds orderable in serial succession required for higher mathematics should not be construed to mean that Kant supposed that our minds evolved the capacity to produce these

256  Nature in Mind: Through Kant’s I’s representations in order that our species might someday spawn a class of professional mathematicians. Instead, he would no doubt have us view it the same way we do capacities like the senses and ordinary imagination: just as we have the ability to savor the special flavor of a fruit unique to a planet in a distant galaxy but never actually will, productive imagination and pure understanding equip us to represent the objective individuality of spaces up to and including an infinity of dimensions even if we never have occasion to represent any but three-dimensional spaces. All that matters is that these two faculties together provide the means to represent spaces that not only exist independently of the sensations they exhibit but have their own independent individuality and identity as well. For how could anything be represented as an objective individual if the space and time in which it is situated could not be? Even the most arcane mathematics and mathematical physics have to take their start from a manifold of n-dimensional, formally simultaneous, co-permanent, objectively individuated spaces determinable (according to rules, operators, functions, etc.) by means of an equally formal succession of objectively individuated times. Just as important for Kant’s anti-platonist purposes is that representations of transcendental space and time be recognized as well within the reach of even quite primitive species of mind. Their only real psychological presupposition is the a priori representation ‘I think.’25 As an analytic unity of apperception, the representation ‘I think’ entails the prior representation of a synthetic unity of apperception as well. Since the latter is prior to all AUA-conception, it necessarily precedes all conceptual thought, and so must consist exclusively of sensible appearances of the kind made possible by the purely aesthetic formal intuitions of space and time that lie ready for sensations a priori in the mind and underlie all representation of sense-divide transcending objects. Nor is that all. The AUA-concepts that creatures endowed with the representation ‘I think’ are capable of producing require logical forms of judgment to combine them in propositions if they are to be put to representational use. Since these forms are all that is needed for the derivation of pure concepts of substance-accident, cause-effect, possibility-impossibility, etc., and since the application of these concepts to the manifolds of formal space and time is sufficient to synthesize representations of fully objective transcendental space and time, even creatures lacking language and everything else that depends on human sociality and lifeways (science and mathematics included) would still have the psychological wherewithal to produce them. Perhaps the most striking feature of Kant’s account of the advent of objectivity through representations of transcendental space and time is its derivation entirely from non-objective inputs. Far from needing to eliminate everything psychological from a representation in order to render it objective, everything from the most ordinary to the most highly

The Objective I 257 mathematized scientific objectivity is, for him, wholly psychological. For if it is granted that the objectivity of a representation is essentially a matter of independence of sensation as regards both existence and individuality, it requires the exclusion of only so much psychology as is beholden to a consciousness of particular sensations, i.e. a posteriori psychology. Of course, in the case of any previous philosopher, the exclusion of empirical psychology would be tantamount to the exclusion of psychology tout court. But this is not the case with Kant. Everything in his psychologistic account of objectivity is directly bound up with the universal self-­ consciousness ‘I think,’ from the pure intuitions of productive imagination requisite for the synthetic unity of apperception it presupposes to the a priori logical structure contributed by pure understanding requisite for non-linguistic propositional thought, transcendental synthesis included. So, paradoxical though it may seem, objectivity is for Kant wholly an expression of an a priori psychologism centered on the ‘I think’—the most quintessentially subjective representation of all. Does this mean there is such a thing as objective unity of apperception? To answer this, one should first recall that pure self-consciousness has two analytically related sides, the original synthetic unity of apperception and the analytic unity of apperception. In beings constituted like ourselves, original synthetic unity of apperception is an amalgam of purely aesthetic formal space and time, while the combination of analytic unity of apperception with logical forms of judgments makes possible pure concepts of the understanding such as substance-accident, cause-effect, possible-impossible, and existence-nonexistence, with objective transcendental space and time the product of the determination of the former conformably to the latter via transcendental synthesis. Since transcendental space and time are consequently nothing but purely aesthetic space and time objectified, the fact that Kant deemed the amalgam of the latter constitutive of the synthetic unity of apperception (Chapter 8) means that the amalgam of the former can be nothing other than the objectification of that same synthetic unity, and so an objective unity of apperception. Nor is “objective unity of apperception” simply another name for transcendental space and time. Kant was quite careful not to preclude the possibility that transcendental synthesis of sensation-independent objectivity can be achieved by forms other than ours, whether different formal intuitions of sensibility, different forms of judgment/pure concepts of the understanding, or both.26 Thus, from a transcendental psychologistic perspective, the objective I is not only more general but also more fundamental than transcendental space and time. Among the consequences of this distinction, none is more important than the support it provides for the contrast between understanding and knowledge presented in Chapters 2 and 5. If all the components essential to making a representation objective are themselves subjective, then the understanding that informs knowledge of objects may vary from subject

258  Nature in Mind: Through Kant’s I’s to subject. This does not prevent the knowledge it produces from relating to a single consciousness-independent objective reality, but it does mean, pace platonism, that the understanding of that reality is anything but phylogeny- and consciousness-independently objective. Its objectivity is limited strictly to sensation independence, which is something that may well be achievable by forms of sensibility and/or understanding different from ours. If so, then there may well be creatures that rival or surpass our technological achievements to which not only our ordinary experience of objects but our mathematics and science too would be quite literally unintelligible. I will return to this topic in the next chapter. But before concluding the present one, I want to stress that my presentation of Kant’s account of transcendental space and time has touched on the roles of only four of the twelve pure concepts of the understanding he deemed requisite for transcendental synthesis. A  more complete exposition would also supplement the case presented earlier that transcendental space and time are necessary to the number line with the case for their being sufficient as well (viz. via transcendental synthesis of the pure quantitative concepts of the understanding—unity, plurality, and totality). It would further show how transcendental space and time furnish the framework from which empirical and mathematical representation of space and time take their start (via transcendental synthesis of the pure qualitative concepts— reality negation and limit). And it would show how no space or time is extra-logically possible outside transcendental space and time (via transcendental synthesis of the pure concepts of community and necessitycontingency). This, however, is not the occasion for a concept-by-concept elucidation of Kantian transcendental space and time.27 For our purposes it suffices to understand it well enough to appreciate its uniqueness in the annals of the philosophy of space and time. To wit . . . Kant’s is the first and only transcendental (i.e. a priori psychologistic) analysis of the objectivity of space and time. All others, before or since, have been physical, mathematical, and/or metaphysical, and either implicitly or explicitly dependent on sensation, linguistic convention, and/ or inputs from platonic heaven. Kant’s, by contrast, psychologistically explicates their objectivity as a wholly natural product of the intelligent mind, understood exclusively in terms of the universal self-consciousness ‘I think,’ its purely aesthetic representational conditions, and the logically structured, yet language-independent propositional representations it makes possible (propositional being the critical mark of the kind of intelligence he had in mind). This transcendental character is all that is required for spatial and temporal appearances to transcend the limits of sensation and empirical imagination so completely as to accommodate the infinite dimensions of a maximally energetic, light-speed-limit superseding “quantum graphity” universe, or to provide even the most abstract, complex systems of mathematics and mathematical logic with

The Objective I 259 the kind of formally simultaneous co-permanent objectified manifolds capable of being ordered in objective formal succession from which they take their start. In my opinion, that is more than enough to certify Kant’s as the most ingenious and original philosophical analysis of objective space and time of its era, and quite possibly ours as well. One final remark. Neither this chapter nor the next could have been written without taking up Kant’s invitation28 to extend his explanation of the pure concept of substance and accident to the remaining pure concepts of the understanding and then applying the result to the analysis of the transcendental synthesis those concepts make possible. Given a choice between staying within the confines of what stands in the text at the price of grossly distorting the philosophy it is intended to convey and utilizing all available clues to fill in the gaps in the manner best suited to realizing Kant’s express philosophical intentions, it seemed to me a no-brainer: scholarly inhibitions, however well motivated, can never justify the payment of so debilitating a philosophical price. Accordingly, I  challenge anyone who disputes any aspect of my reading, especially my exoneration of Kant from the charges of Euclidean and Newtonian dogmatism, to provide an analysis of the kind I  have sketched here and developed more fully elsewhere, showing that when transcendental synthesis is fully worked out, there emerges a transcendental space and time that do not merely suffice for the possibility of mathematics and physics—be it those of Kant’s, ours, or any subsequent era—as I contend, but instead actually imply the truth of Euclidean geometry/falsehood of non-Euclidean geometry or the truth of Newtonian physics/falsehood of relativity theory and quantum theory. If you can’t do that, then you have to concede that Kant’s transcendental philosophy does not have these consequences. And afterward, with your previous certainty shaken, I  invite you to reread what Kant actually says and I  suspect you’ll end up agreeing with me that there is in fact nothing remotely close to a commitment to dogmatic Euclideanism or dogmatic Newtonianism to be found there.

Notes 1. “Living in a quantum world,” in Scientific American, June 2011: 24. 2. If some physicists include the Higgs or other fields among fundamental forces of nature, it is enough for my purposes that they all fall within the scope of the Standard Model. I shall therefore ignore the question of how many forces count as fundamental and simply take four for granted. 3. CPR A20/34. See chapter 8 note §12 regarding my use of Kantian aesthetic terminology. 4. See chapter 8 note §14. 5. A case in point is Kant’s claim at CPR B41 that the proposition that space has three dimensions is a priori, i.e. necessarily and universally true. While this may seem to be a clear-cut instance of Kant’s dogmatic Euclideanism, scholars who assume this almost invariably overlook the fact that he cited this proposition as an example of geometrical, not transcendental a priori truth.

260  Nature in Mind: Through Kant’s I’s It is thus perfectly consistent with my reading of Kant’s transcendental philosophy as leaving it to geometry to settle properly geometrical questions like this. Since the only geometry in Kant’s time was Euclidean, only threedimensional space met the era’s standards of geometrical construction. But, as I argue later in this chapter, the objective space yielded by transcendental synthesis is perfectly able to accommodate a space with higher dimensionality, including infinitely many. So, on my reading, the fact that 19th geometers discovered how to construct higher dimensional spaces (albeit symbolically rather than ostensively) poses no problem for Kant’s transcendental psychologism. Unfortunately, the failure of nearly all Kant scholars to properly distinguish transcendental space from geometrical has spilled over into the sciences. Quantum scientist David Deutsch is just one of innumerable examples I could cite: “Kant . . . concluded that Euclid’s theory of geometry was self-evidently true of nature. Hence he believed that it was impossible rationally to doubt that the angles of a real triangle add up to 180 degrees. And in this way he elevated that formerly harmless misconception [that there is no difference between mathematical and physical space] into a central flaw in his philosophy, namely the doctrine that certain truths about the physical world could be ‘known a priori’—that is to say, without doing science. And of course, to make matters worse, by ‘known’ he unfortunately meant ‘justified’ ” (The Beginning of Infinity: Explanations That Transform the World, New York: Viking, 2011, pp. 182ff). It is true that Kant held that transcendental space underlies mathematical, scientific, and ordinary representations of objective space, but it is false that transcendental space itself is any more Euclidean than it is non-Euclidean. It is mathematically completely undetermined, and so intrinsically incapable of imposing any properly mathematical framework either on nature itself or on our scientific representation of it. 6. AA 18 §6314 [1790–1]. Also: “Number requires space and time for its intuitive representation” (AA 17 §4629 [1771–3]). 7. CPR A411/B438. 8. CPR A183/B226. Simultaneity “must therefore have a different ground” (AA 18 §6313), i.e. space. See also AA 17 §4509 and AA 18 §6346. 9. Since “the representation of something permanent . . . is not the same as a permanent representation” (CPR Bxli n.), the representation of formal permanence, as apprehended in internal intuition, is as instantaneous as any and all other representations. 10. AA 18 §6359. A walk is a series of space-traversing steps, and so grounded in time, the formal condition of all series. Where the spatial unit (here one step) is conceived as a measure, temporal successions of units measure off spaces, thereby numerically quantifying them. 11. Kant’s transcendental schema of number (CPR A142–3/B182) is an application of transcendental synthesis, but not that synthesis itself. Since the analysis of the latter is quite complicated, I refer readers interested in the details to KAIM chapter 15-E. The difference between transcendental synthesis and schematism will be discussed in the next chapter. 12. What about the imaginary y-axis mathematicians use for complex numbers, the imaginary z-axis for quaternions, and further axes for octonions, sedenions, etc.? Since construction of these extra dimensions requires properly mathematical operations, starting with taking the square root of −1, complex numbers et  al. should be regarded as mathematical structures built on transcendental foundations rather than as anything possible purely transcendentally. 13. CPR A137/B176 and A142/B181.

The Objective I 261 14. My objective in the foregoing was to explain why Kant regarded transcendental space and time as necessary for representations of number series. Those interested in why he also regarded them as sufficient should consult KAIM chapter 15-E. 15. See CPR B145–6 and A230/B283. 16. CPR A30/B45. The thing in itself will be discussed in the next chapter. 17. Kant termed this portion of his investigation the “subjective transcendental deduction of the categories.” See KAIM introduction to part III and chapter 12 for discussion of the relevant texts. 18. Failure to transcend it results in what Kant termed a “transcendental amphiboly” fallacy: see CPR A260/B316 ff. 19. AA 18 § 6313 [1790–1]. 20. For ease of inspection, I have shortened the causal relations expressed here from (“Tm is pm” ⇒ “Tn is pn”), (“Tn is pn” ⇒ “Tn+1 is pn+1”), etc., to “Tm is pm” ⇒ “Tn is pn” ⇒ “Tn+1 is pn+1,” etc. Subscripts are for convenience only. For if without the pure concepts of the understanding there can be no differentiation from or determination in relation to other spaces or times, and so neither objective formal simultaneity permanence nor succession, then there cannot yet be numbers, alphabetical order, or any other kind of series. For Kant, number in particular is a transcendental concept that is as much a product of the transcendental synthesis of formal space and time as transcendental space and time as such. See analysis in KAIM chapter 15-E. 21. CPR A137/B176 and A142/B181. 22. A more detailed exposition of transcendental synthesis can be found in KAIM chapter 15. It does, however, minimally presuppose familiarity with the analyses of the logical forms of judgment in chapter  10 and the pure concepts of the understanding in chapter 11. 23. Spooky Action at a Distance, pp. 182–5, by George Musser, New York: Scientific American / Farrar, Straus and Giroux, 2015. 24. One difference is that while Wheeler’s primitive grains of matter are not further determined, Kant’s purely aesthetic formal spaces have to be determined as spatial final subjects by the pure concept of substance-accident before they can be linked in the network. But this difference does not affect the analogy. 25. Although Kant used pure apperception as a criterion for distinguishing human from non-human terrestrial minds, I  suspect that an acquaintance with contemporary sciences like ethology and paleoanthropology might have opened him up to the possibility that apperception may have figured in earlier Homo species, possibly also Australopithecus, hominoids, and even perhaps cetaceans, corvids, and others. 26. See CPR B145–6 and A230/B283. Creatures with space and time but different forms of judgment/pure concepts of the understanding would still represent transcendental space and time, but the individuality and identity of the spaces and times within them would be conceived in a manner incommensurable with ours—e.g. quantity would not take the form of number. 27. See note §22. 28. At CPR B128–9.

10 The I of Nature

Transcendental synthesis is the key to the solution of the heterogeneity problem of Chapter 7. However, its complete solution required Kant to show how transcendental objectivity can combine with physical reality in such a way that pure concepts of the understanding can finally merit their designation as categories—and, in the process, psychologistically explicate nature itself. The explication falls short of idealism insofar as the existence of physical things remains completely independent of consciousness, even if, by Kant’s a priori psychologistic reckoning, everything about them—their spatiality, temporality, substantiality, causality, mathematical quantifiability, universal subordination to natural laws, and laws of nature themselves—does not. In particular, in the absence of consciousness, physical existence remains in the form of what Kant termed things in themselves. After addressing a number of topics relating to mathematical physics, the chapter  concludes by laying out, step by step, Kant’s refutation of the Humean skepticism outlined in Chapter 5.

Objectivity and Physical Reality Consciousness has access to physical reality via sensations of the external senses. It may be true that both internal sensations and every other conscious representation have neural correlates, and so, to that extent, give “access” to physical reality via the brain. But the brain is no more part of internal conscious experience than the individual neurons, glia, and other cells that compose it, much less the molecules, atoms, and everything else down to the bosons and fermions that compose them. In fact, we have no conscious access to our brain at all unless the skull is removed and it can be seen (in a mirror), touched, or otherwise accessed via external sensations. Sensational access to physical reality starts with access to our own body via tactual sensation in and beneath the skin, and proceeds to access bodies external to our own via visual, auditory, and other sensations. Everything else in consciousness, internal sensations included, gives access to physical reality only at second hand, as when external sensations are recalled, imagined, and given universal form in concepts

The I of Nature 263 employable in propositional representations, or purely reactively in emotions (seeing a threatening animal), desires (smelling French cooking), and volitions (rising to answer the doorbell). Neural correlates notwithstanding, external sensations are the only access to physical reality any of us, scientists included, ever have, so that in their absence we would be locked into the kind of radically solipsistic existence Descartes supposed to be the case before and after the mind’s union with the body. Like Locke before him, Kant emphasized the indispensability of external sensation as our immediate and, ultimately, only access to physical reality. Its immediacy gives external sensation the same relation to physical reality the representation ‘I think’ has to the conscious subject, and so confers the same immunity to doubt regarding the existence of the external world that the ‘I think’ confers on the existence of oneself. Its being our only access to physical reality means that no representation of the latter can be accorded the least cognitive worth unless it relates to external sensation, either directly when the represented object itself is present in external sensation, or indirectly when something present in external sensation is causally or otherwise existentially dependent on something not present in it. Any ostensible representation of physical reality that is incompatible with being given directly or indirectly through external sensations, human or any other, must be adjudged inaccessible in principle and so cognitively null. For every external sense as such, whether present in humans or not, whether actualized in a phylogeny or requiring futuristic genetic (or other) engineering, shares the same immediacy to physical reality and indispensability to accessing it that is true of human touch, sight, et  al. Indeed, Kant went so far as to single out external sensation as that alone which gives conscious representational access to the subjectively unconditioned transcendental reality of objects existing mind-independently in themselves.1 Yet the unique status Kant accorded to external sensation vis à vis physical reality exposes a gap in his psychologism of such significance that its tenability hinges on closing it. Transcendental space and time are, for him, as emblematic of and essential to objectivity (Chapter 9) as external sensation is with respect to physical reality. But since transcendental space and time are so pure and formal as to exclude all contributions of sensation from their representation, whereas external sensations include nothing whatsoever of space and time in theirs (Chapter 8), it is difficult to see how transcendental space and time could have the least physical reality or the real in sensation the least objectivity. Kant was well aware that unless he could explain how the objectivity of the one can combine with the relation to physical reality of the other, transcendental space and time would have to be dismissed as mere “phantoms of the brain,”2 while sensations “would be without any object and so be nothing but a blind play of representations, i.e. less than a dream.”3 Thus, the be-all and end-all of the cognitive phase of his psychologism became

264  Nature in Mind: Through Kant’s I’s showing that and how the mind is able to physically realize its representations of the former and objectify its representations of the latter, and, moreover, do so in such a way as to yield a representation of nature as an all-encompassing system of fully objective realities everywhere subject to natural law, i.e. to psychologistically explicate nature itself. Given its importance, I shall briefly postpone discussion of Kant’s solution in order to examine the problem more closely, starting from the side of external sensation and concluding with that of transcendental space and time. In the former regard, it is natural to ask what access to physical reality through sensation would amount to in the absence of transcendental space and time. One response is that it would make little or no difference. For how could their absence impede our access to ordinary physical realities like stones, trees, sky, ground, chairs, cats, computers, and, of course, our own bodies? This, however, is not to reckon with the fact that these, like external objects generally, essentially involve a single multi-sensory external sense capable of representing the same object through sight, touch, and the other senses—i.e. they are all sense-divide transcending objects, potentially if not actually (e.g. people may one day see their shadows on the Martian surface and feel its soil beneath their boots). Sensation by itself cannot be the source of such representations since sensations across sensory modalities are radically incommensurable both qualitatively and relationally, and their correlations seem completely arbitrary (Chapter 3). Nor does association of sensations alone suffice to make good this want. Insofar as association is conscious and owes its representational primacy to custom (Chapter 4), it involves consciousness of the frequency and constancy of the occurrence and co-occurrence of sensations, and so, by Kant’s reckoning, is dependent on a representation of time that neither sensation nor association can yield: an ever-changing (non-repeating), irreversible (ordered), linear (unbranching) succession in which each moment is objectively (i.e. sensation-independently) individuated by being uniquely differentiated and completely determined vis à vis every other (Chapters 8 and 9). This is not to deny that creatures capable of sensation and cruder forms of association do not have other, most likely non-conscious mechanisms to organize sensations across sensory modalities whereby to regulate their behavior. But, in Kant’s view, genuine cognitive representation of sense-divide transcending external objects, much less representation of an all-encompassing realm of nature and natural law, is impossible for such minds.4 Transcendental space and time go some way toward making good this want. Since everything in them exists and has its own individual identity fully independently of the sensations they exhibit, their contents are not dispersed among diverse, incommensurable sensory fields as heterogeneous sensations are, but instead stand immediately together, a single infinitely iterable homogeneous manifold, all contained within the fully objectified unity of an all-encompassing, individual sensible

The I of Nature 265 consciousness. Nevertheless, as nonsensational, these sense-divide transcending spaces and times have nothing to connect them to physical realities that are accessible uniquely through sensations. For formally intuited space and time, and so too every spatial and temporal representation possible in and through them, are products not simply of imagination but of productive imagination quite specifically, and so lack even the derivative, second-hand relation to physical reality characteristic of even the wildest sensation-imbued products of reproductive imagination (riding a unicorn at the derby, leaping to Enceladus in a single bound, etc.). They are beholden to sensation for 0% of their content and 100% to forms that “lie ready for the sensations a priori in the mind.”5 The same is true of the objective differentiation and determination conferred on the manifolds of formal space and time via transcendental synthesis in accordance with pure concepts of the understanding: these concepts are products of universal self-consciousness that derive 100% of their content from logical forms that lie ready a priori in the mind and 0% from sensation. The implication therefore seems inescapable: the objectivity of transcendental space and time and everything in them is 100% devoid of relation to physical reality. If this were all there were to Kant’s account of objectivity, then transcendental space and time would indeed be mere phantoms of our brains, and the pure concepts of the understanding to which they owe their objectivity could never merit the status of categories—fundamental concepts of objects that precede and make possible all other concepts, ordinary and scientific alike, of physically real objects. Surprisingly, the reigning assumption among philosophers is not only that Kant’s account does not proceed further, but that doing so was never his intention. This is maintained notwithstanding his explicit statements that pure understanding is nature’s creator and legislator,6 and his repeated warnings that this “Copernican” thesis, which seemingly reverses the order of nature by subordinating physical reality to (a priori) psychology, would be certain to appear “surprising,” “bold,” “strange,” “exaggerated,” “a riddle,” even “absurd.”7 The typical response is to dismiss Kant’s rhetoric as at best hyperbole and at worst self-delusion, so that the more well-meaning tend to feel obligated by considerations of philosophical charity to set something far more modest in its place, such as the thesis that space, time, the categories, and even the I itself are components of an innate conceptual scheme, or framework, with which our understanding operates in representing the realities experienced through sensation. The implication is that transcendental representations are a kind of lens through which physical reality is viewed rather than a law unto that reality itself. For the necessity of using this lens is a necessity that can be supposed to apply only to our representations of physical realities, not to those realities themselves. Since the latter are under no more obligation to conform to necessities of our representations of them than they are to any other, quite

266  Nature in Mind: Through Kant’s I’s possibly incompatible necessities of thought that govern other (actual or possible) creatures’ representations of them, this purely formal scheme of “objectivity” is either of merely subjective, phylogenetically relative validity or only serendipitously in agreement with physical reality. It therefore is unclear what such acts of “charity” actually preserve of Kant’s revolutionary ambition to establish that physical objects and even nature itself conform to conscious representations and not the reverse. And, in any event, none of this is to reckon with the fact that, in addition to its purely formal side, Kantian transcendental synthesis has a material side as well.

The Matter of Appearance We saw in Chapter 8 that sensations can figure in a priori psychologism only to the extent that none is consciously distinguished from any other, be it by quality, relation, existence, or anything else. For even dark consciousness of what is distinctive to particular sensations and differentiates some from others, possible or actual, is, for Kant, the very definition of empirical and so can have no place in a priori psychologism. The two genuinely a priori features of sensations this leaves are their manifoldness and the existence each immediately indicates that mere thoughts, whether in memory, fantasy, or propositional representation, never do.8 The manifoldness of sensations is exhibited a priori by the infinitely iterable manifolds of non-sensational appearances preceded, made possible, and united by formal intuitions of space and time that lie ready for sensations a priori in the mind. This we have already considered. But what about the existential side of sensations, and particularly the physical realities immediately indicated by external sensations? The answer is to be found in an aspect of the synthesis of apprehension in intuition at the heart of Kant’s account of pre-associative sense perception that we have yet to consider. According to this, the appearances yielded by productive imagination’s synthesis of apprehension exhibit not only the manifoldness of sensations but the physical realities indicated by them as well. Adapting Aristotle’s form/matter distinction, Kant designated the aspect of appearances concerned with the former their form and that concerned with the latter their matter.9 If an appearance exhibits an actual sensation, it is apprehended in productive imagination as having a matter as well as a form. This means that the existence indicated by the sensation carries over to the matter of the appearance as well, so that, in exhibiting a sensation, the matter of appearance must likewise be represented as corresponding to “the transcendental matter of all objects as things in themselves (facticity, reality).”10 Moreover, since the matter of appearances is no less non-sensational than their form, we are free to imagine any appearance with a matter not to have one and any appearance without a matter to have one. It thus provides us with a purely aesthetic, productive-­imaginative means of representing possible vs. actual existence/

The I of Nature 267 nonexistence that is completely independent of the purely intellectual means exemplified by the (logical-form derived) pure concepts of possibilityimpossibility and existence-nonexistence discussed in Chapter 7. In addition to exhibiting the realities indicated by sensations, the matter of appearances also serves to exhibit variations in sensational intensity. A sensation may vary in degree from zero (perfect insensibility) through every degree right up to the maximum intensity of which the sense is capable, all without affecting its quality, situation with regard to other sensations, or anything else. Similarly, in exhibiting the realities indicated by sensations, the matter of appearances can be represented as greater or lesser without in any way affecting their formal character (e.g. their spatial configuration or duration). Of course, since the matter of appearance is non-sensational, its intensity need not be supposed to go to zero when that of the sensation it exhibits does, nor need its intensity be supposed to be limited by the maximum intensity of which a particular creature’s sensory systems are capable. Thus, zero intensity of the matter of appearance exhibits the absence not just of sensation but of any reality whatsoever, while, at the other end of the scale, it is fully capable of exhibiting the maximum any reality can attain irrespective of the limits of our sensitivity to it (e.g. while no thermoceptive sensation of the heat within the hottest region of the sun’s corona may be possible, there is nothing to prevent its being apprehended in intuition via the matter of its appearance). The matter of appearances enables us to represent sense-divide transcending physical realities in the same way their form makes it possible to represent the manifolds of sense-divide transcending spatiotemporal objects. For example, the reality indicated by the warmth I  feel from wood burning in the fireplace can be combined or identified with the heterogeneous realities indicated by my visual and auditory sensations of its burning because the non-sensational matter of the appearances exhibiting these sensations is as homogeneous as their form, and so admits of being represented in accordance with empirical inputs either as coalescing to constitute a more intense conjoint reality or simply as one and the same reality accessed through different senses. Without the matter of appearances, representation of sense-divide transcending physical realities would be as impossible as the representation of sense-divide transcending spaces and times would be in the absence of the requisite forms. Both sides of appearances are for Kant equally essential. The only question is how they can be combined to yield objective representations of sense-divide transcending objective spatiotemporal physical realities.

Transcendental Schematism: Making Objectivity Physically Real and Physical Reality Objective The matter of appearances provides the access to physical reality that their forms do not. Yet, as it is in itself formless (aspatial, atemporal),

268  Nature in Mind: Through Kant’s I’s it is of no more use to represent that reality objectively than the subjective sensations it exhibits (objectivity being the product of transcendental synthesis of the formal side of appearances). There is, however, this crucial difference from sensations: because their matter is as indelibly a part of the appearances generated in productive imagination as their form, it partakes of the a priori unity appearances have as infinitely iterable homogeneous manifolds within formally intuited space and time (unity of sensibility). While these purely formally constituted unities do not provide access to physical reality, they nevertheless create the possibility for such access insofar as the unity of sensibility they yield constitutes an original synthetic unity of apperception (SUA). For with this unity in place, analytic unity of apperception (AUA) becomes possible as well. As logically universal self-consciousness, AUA can directly convert any non-conceptual representation that incorporates it into an AUAconcept. Insofar as SUA encompasses the matter as well as the form of non-­sensational appearances, AUA can therefore be represented a priori as attaching to the synthesis of appearances in respect to not only their form but their matter as well. When this is done, there results an infinity of AUA-concepts of the matter of the appearances that differ only in logical form (universality), not content, from the matter of sensible appearances themselves. Moreover, since the matter of appearances is capable of exhibiting all possible realities accessible through sensations directly or indirectly, AUA-concepts of this matter are representable as corresponding to every reality that can ever be given to us through sensation. We have then only to incorporate these concepts into a propositional network by coupling them in propositions with predicates that represent each of the infinite exhibitive possibilities inherent in the matter of any appearance for the prospect to emerge of subjecting the infinite totality of physical realities to the same transcendental synthesis as the manifolds of purely formal space and time. Since this would have the effect of extending objective transcendental space and time from the form to the matter of appearances, and from there to the physical realities indicated by the sensations they exhibit, such a synthesis would at once confer spatiotemporal objectivity on physical reality and physical reality on spatiotemporal objectivity. Transcendental synthesis of the matter of appearances cannot, however, proceed exactly like that of their form. For if what is to emerge from the synthesis is to be the physical realization of transcendental space and time, the propositional network formed from the AUA-concepts of sensation-accessible physical realities cannot be determined directly by the pure concepts of the understanding. The latter have no meaning, in and of themselves, except as restrictions on the logical freedom of concepts and propositions. Being devoid of spatial and temporal content, they are consequently incapable of conferring such content on anything that lacks it; and since space and time are restricted to the formal side

The I of Nature 269 of appearances, this includes their matter. It therefore follows that pure concepts are incapable of conferring spatiotemporal meaning on the non-spatial, non-temporal manifold of AUA-concepts of the matter of appearances. Fortunately, the situation is easily remedied by substituting in place of the pure concepts of the understanding that enter into transcendental synthesis of the forms of appearances the spatially and temporally meaningful concepts that emerge from it. As we saw in Chapter 9, when the former are combined with the purely aesthetic spatial and temporal forms of appearances in transcendental synthesis, they are transformed into concepts of spatial final subjects connected to other such subjects in temporally determinate causal relations, each uniquely differentiated from and completely determined in relation to the infinite totality of other spatial final subjects, both spatially and temporally. By applying these spatially and temporally enriched versions of pure concepts of the understanding to a propositional network formed from the infinite totality of AUA-concepts derived from the matter of appearances, transcendental synthesis is thus able to yield a new propositional network that combines any and all physical realities exhibited by that matter with objective transcendental space and time. Kant termed the transcendental synthesis that produces physically real objective transcendental space and time transcendental schematism, and the spatially and temporally enriched derivatives of pure concepts of the understanding utilized in it transcendental schemata.11 The synthesis is twofold in its effect. On the one hand, it confers on transcendental space and time the physical reality that makes them more than mere phantoms of the brain, and, on the other, it transforms formless physical realities into physically real occupants of otherwise empty transcendental space and time. To be sure, the expression “physically real objective transcendental space and time” signifies nothing more or other than the propositional network generated by transcendental schematism from AUA-concepts of the matter of appearances. The descriptor “physically real” is nonetheless warranted because the conceptualized matter in these propositions differs solely in logical form (the universality of consciousness) but not in content from the matter of appearances themselves, i.e. physical realities of a particular intensity. Accordingly, there is nothing to prevent the “filling,” or “occupation,” of transcendental space and time constituted by schematized AUA-concepts of the matter of appearances from carrying over fully and completely to the purely aesthetic matter of appearances themselves in the same way everything true of the AUA-concepts of the form of appearances carries over to purely aesthetic spaces and times in the first stage of transcendental synthesis (Chapter 9). There is, however, this difference from transcendental synthesis of the formal side of appearances: whereas unschematized transcendental space and time extend no further than the purely imaginary spaces and times

270  Nature in Mind: Through Kant’s I’s that exhibit the manifoldness of sensations, schematized transcendental space and time extend all the way to physical realities themselves. For whereas, on Kant’s analysis, there is nothing in the least spatial or temporal about sensations, the indication of existence/reality is an intrinsic feature of all (and only)12 sensations as such, and so too of the matter of the non-sensational appearances that exhibit these sensations. Thus, the matter of appearances confers on schematized transcendental space and time the very same physical reality that is found in trees, clods of earth, the sun, and all other sense-divide transcending physical realities (including one’s own body) representable as filling them. I remarked earlier that Kant took sensations to indicate physical realities indirectly as well as directly, and it is now possible to see why. Insofar as transcendental schemata incorporate the matter of all appearances into objective transcendental space and time, the sensation-­ indicated realities they exhibit are embedded in a physically real, temporally determinant causal network of spatial final subjects isomorphic to that described in the previous chapter with respect to their spatial and temporal forms. This means that every external sensation present in consciousness, the matter of appearance that exhibits it, and so too the physical reality its presence indicates will be represented a priori as determined in accordance with irreversible final subject–final predicate and groundconsequent relations. Sensations present in consciousness will therefore no longer just indeterminately indicate the existence of some reality but quite specifically indicate the actualization of a causally effected accident in an underlying physically real substantial spatial final subject substrate. This means that every sensational consciousness not only directly indicates the existence of that accident but also indirectly indicates the reality of two further things that may not be present in that consciousness at all: the underlying spatial substance in which the accident is represented as inhering and the cause represented as having produced it. Nor is that all. If the cause is the accident of a different spatial substrate than its effect, then the sensation indirectly indicates the existence of a second spatial substance, none of whose accidents need be present to consciousness in sensation at all. If that cause is in turn part of a causal series, then the sensation will further indirectly indicate the reality of everything prior to its immediate cause in the series. If each of those causes is in turn the accident of a spatial substrate it shares with no other in the series, then a multitude of physically real spatial substances, perhaps infinitely many, will be indirectly given with the sensation as well. And that still is not all. If the spatial substrate in which its causative accident inheres has additional accidents that causally connect it to the accidents of still other physically real spatial substrates, each different from the others, which are connected in the same way to still other such substrates, then, thanks to transcendental schematism, that one sensational consciousness may indicate the existence of an entire network of

The I of Nature 271 interacting spatial substances, each with a multitude of accidents causally connected, directly or indirectly, to other members of the network, none of which need be present to consciousness in sensation at all. The upshot is that, on Kant’s analysis, with each sensation—any ­sensation—the entirety of nature may be given as well. At a maximum, transcendental schematism allows a sensation to indirectly indicate the existence of an infinite network of spatial substances, each with an infinity of accidents that causally connect it to accidents of all the others. Such a network would give physical reality to the fully actualized transcendental space and time produced by transcendental synthesis that I compared to maximally energetic “quantum graphity” in the preceding chapter. Any less than maximal network—i.e. one where not all possible substances and/or not all possible accidents are actualized via the schematized pure concept of existence/nonexistence—would only partially physically realize objective transcendental space and time, but they would still to that extent not be mere phantoms of the brain. Thus, in giving objective spatial and temporal form to the manifold of sensation-indicated physical realities, transcendental schematism simultaneously gives physical reality to at least some of objective transcendental space and time. What about the future? The distinction between present, past, and future is as meaningless with respect to schematized transcendental space and time as it is for the “block universe” of relativistic spacetime. Present, past, and future depend on empirical, not pure, apprehension, for they are not independent of particular internal sensations either in existence or in individuality, and so can only be subjectively, never objectively, distinguished. Transcendental schematism, by contrast, is pure and so excludes all a posteriori consciousness of sensations whereby any may be distinguished from any other, but instead focuses exclusively on a feature common to all sensations as such: the indication of an objectively real physical existent. It thereby sets every reality indicated by sensation (via the matter of the appearance exhibiting the sensation) in a physically realized transcendental space and time, with the fully objective individuality and identity that implies. This makes the subjective individuality sensations confer on spaces and times, including whether they are present, past, or future, a matter of complete indifference where the objective physical reality of these spaces and times is concerned. Should one’s present sensations turn out to be one’s last, the physical realities indicated by them may well have objective physical effects in the objectively succeeding (“future”) time and so too indirect physical effects in times objectively subsequent to it, perhaps without end. But even if they are all termini of their causal series, any transient realities simultaneous with them and inhering in the same or causally connected permanent spatial substances will be determined by transcendental synthesis to have “future” effects in objectively subsequent times, potentially ad infinitum. Thus, the physically real objective transcendental space and time yielded

272  Nature in Mind: Through Kant’s I’s by transcendental schematism are fully consonant with the spacetime “block universe” of relativistic physics.

Nature Psychologized Kant’s readiness to affirm the possible existence of an inexhaustible infinity of physically real spatial substances, each with an equally inexhaustible infinity of determinations temporally connected in causal series of every possible length, must not be construed as a claim of physics, much less metaphysics. The reality whereby substances fill/occupy the space and time of transcendental schematism is never anything other than the non-sensational matter of appearances that exist only in and through productive imagination. The transcendental space and time they fill are nothing other than a propositional network formed from this matter via analytic unity of apperception, logical forms of judgment, and pure concepts of the understanding enriched with spatial and temporal content via transcendental synthesis of the forms of appearances. Being composed entirely of a priori psychological elements, the schematized representation of objective transcendental space and time can therefore never lay claim to anything more or other than the significance conferred on it by a priori psychologism. Nor can Kant’s schematized propositional network be viewed as in any sense a language, not even a private one. Languages reference physically real spatiotemporal objects, and to that extent are about the natural world. But languages are not physical theories, metaphysical theories, or theories of any other kind. They do not require, expressly or otherwise, that the objective reality they reference be something other than a propositional network. They require only that it be objective, spatial, temporal, and, above all, physically real. If a propositional network can be all of those things—if indeed only the propositional network of transcendental schematism is even capable of being all of them—then the world demarcated by schematized transcendental space and time is the natural world referenced by all other propositional networks, human languages included. And this, for Kant, is the pay-off that a priori psychologism alone is capable of delivering: the exposition of transcendental schematism as a propositional network constitutive not just of our thought about physically real objective space and time but of the very physically real objective space and time our thought is about. It is indeed nothing less than the psychologistic explication of nature itself. Kant was no doubt right to fear that a philosophical thesis revolutionary enough to warrant the mantel of Copernicus would seem “surprising,” “bold,” “a riddle,” “strange,” “exaggerated,” even “absurd.” But he could not gainsay the fact that the a priori psychology of the ‘I think’ leads inexorably to the psychologistic conclusion that pure understanding is “not merely a capacity to form rules through the comparison of

The I of Nature 273 appearances” but, through its pure concepts, “the legislator for nature” and “itself the creator of the experience in which its objects are found,” so that “apart from the understanding there would be no nature at all.” The notion that “the understanding is itself the source of the laws of nature, and thence of the formal unity of nature,” is indeed bold, for it implies that “the laws of nature . . . are just as little to be encountered outside the understanding as space and time are [to be encountered outside sensibility].”13 Nevertheless, it is the inevitable consequence of the thesis that schematized pure concepts of the understanding provide the basis for propositionally formed laws so fundamental as to be constitutive of nature itself and everything in it. We will see how this cashes out when we get to the principles of pure understanding. Here it suffices to note that Kant did not claim that the laws of nature we encounter via experience of physical reality are entirely the work of pure understanding. He claimed only that transcendental schematism guarantees that all possible laws of nature, regardless of the content represented in them, will take propositional form and, more particularly, incorporate into their content the propositional network constitutive of physically realized transcendental space and time.14

The Psychological Reality of Transcendental Schematism One may object to Kant’s psychologistic explication of nature on the psychological ground that it posits conscious activity too complex and sophisticated to be plausibly supposed to have evolved and operated in our hunter-gatherer ancestors, much less their evolutionary forebears. Nor is it simply that we do not find any such propositional network in ourselves; we cannot find it no matter how hard we search. In response, Kant would no doubt concede its absence from access consciousness but then remind us that the capacity to shine the light of attentive discernment is in general very limited, and that attentive discernment must never be confounded with consciousness as such, which is like “a great map on which only a few places are illuminated.”15 Scientists today, like ordinary people always (philosophers included), commonly infer the existence of non-conscious mental processes that affect subsequent consciousness. Computational mind theorists, for example, typically posit “hard-wired” neural algorithms for mental activities such as sight in three dimensions, creating spatial maps to navigate one’s environment, and mediating between the senses and motor activity, as well as such higher mental functions as proposition-dependent reckoning and planning. But to the extent that access consciousness is unneeded for these, why should it be supposed to be requisite for transcendental schematism? The latter would almost certainly be far less taxing for a brain than keeping the body atop a careening mountain bike or maneuvering fingers in just the right ways to perform a well-rehearsed Liszt Transcendental Etude—both of

274  Nature in Mind: Through Kant’s I’s which are executed most efficiently and economically by dispensing with access consciousness altogether. Why should transcendental schematism be supposed any different? As it happens, attending to schematism would be self-defeating since the whole point of the synthesis is to supply us with the physically real a priori spatiotemporal propositional network wherein alone genuine sense-divide transcending spatiotemporal physical realities can be attentively accessed through sensations a posteriori. Thus, just as in piano performance paying attention to the means of execution can only distract from the end, schematism operates best by drawing as little attention to itself as possible, and ideally none at all. A more pertinent objection perhaps is to doubt whether schematism requires any consciousness at all, even the dark kind. We have seen that philosophers previous to Kant inferred a wide variety of complex mental activity that takes place outside the spotlight of attention yet is nevertheless conscious. To distinguish it from non-conscious mental activity, they pointed to some feature of the activity that cannot be coherently supposed to exist in the absence of consciousness. For example, the feelings of facility and vivacity that Hume deemed essential to all properly associative relation and belief can no more exist non-consciously than feelings of pleasure and pain can, even though they completely escape our attentive gaze (Chapter 4). Kant, however, had an even stronger case to make. As concerned with a propositional network, transcendental schematism can take place only given logical forms for combining distinct concepts, i.e. representations universal in form. Since the universality in question has no content but is instead merely the representation of the ‘I think’—a consciousness (not a content) that can accompany, and so is logically common to all possible representations—schematism in the absence of consciousness is as impossible as pain is (indeed, the notion of a non-conscious cogito is, if anything, even more radically incoherent). Since both the synthetic and analytic unity of apperception are essential to all of its components, the possibility of non-conscious schematism can therefore be definitively precluded notwithstanding the complete inaccessibility of its operations to the spotlight of attention. There thus is no escaping the psychologistic implication of the consciousness-dependence of nature if one accepts (1) that nature is representable only via transcendental schematism and (2) that consciousness is constitutive of transcendental schematism. Nor is there anything in Kant’s psychologistic explication of nature that cannot plausibly be attributed to humans or any other creature in a manner fully compatible with their gradual evolution from minds with some but not all of the capacities requisite for this operation. In the first place, transcendental schematism is a natural extension of transcendental synthesis, which, in turn, can plausibly be reckoned an adaptationally advantageous realization of a potential implicit in the possession of formal intuitions of sensibility and pure concepts of the understanding. The

The I of Nature 275 latter can be supposed to be acquired in the natural course of employing forms of judgment to synthesize propositions, while the former are part and parcel of the synthesis that makes it possible to apprehend sense-divide transcending appearances in intuition. Both functions seem sufficiently primitive that I see nothing to prevent their being plausibly attributed not only to earlier Homo species but to other clades as well. Thus, Kant’s purely psychological, yet fully objective conception of nature and natural law seems entirely compatible with evolutionary theory.

Transcendental Judgment and the Advent of Empirical Cognition Transcendental schematism demonstrates that the objective transcendental space and time generated by transcendental synthesis from formal intuitions of space and time are no mere phantoms of the brain. But they would still count for little if their representation had no employment in the arena that matters most: experience of physical realities, both the everyday kind and mathematically informed scientific theorizing, observation, and experiment. The former is the territory staked out by empirical psychologists like Hume and consists of actions like recollection, discernment, comparison, association, custom, the formation of general representations, inference, belief formation, and propositional representation. Setting aside for the present those acts that depend on convention (especially linguistic), Kant’s general term for these acts was empirical synthesis, with association, conceptual recognition, and the logical combination of recognitive concepts in (non-linguistic) propositions singled out as most important for empirical cognition of fully objective sensedivide transcending spatiotemporal realities. Kant analyzed association as “preferential reproduction,”16 by which he meant the psychological operation whereby upon the appearance of one representation in empirical intuition, some specific other(s) is (are) reproduced in thought preferentially to possible alternatives, or at least is (are) more likely to be. It could be the result of features of the appearances themselves—directly observable qualities (resemblances) and relations (spatial and/or temporal contiguity)—or of habits of mind produced by patterns in their occurrence (constant co-occurrence or sequence). More importantly for Kant’s purposes, association is a largely automatic, reflexive process. We must be at least darkly conscious of both the features that trigger preferential reproduction and the feeling of facile transition that relates the representations to and from which the transition in thought is made. But that is all. There is no consciousness, not even a dark one, of the synthesis in its entirety, as the unitary product of various operations performed a particular way and in a particular order, i.e. according to a rule. Associators instead simply blindly obey mental instincts such as relating like to like, and blindly formed habits

276  Nature in Mind: Through Kant’s I’s that induce them to continue in thought experiences that partially conform to patterns encountered with sufficient frequency and constancy in the past, such as inferring in thought the existence of a grizzly from the sight of fresh grizzly tracks. This is where Kant’s synthesis of recognition in a concept comes in. By attaching analytic unity of apperception to an associative synthesis and transforming it into an AUA-concept, we are made conscious of even our most complex, protracted preferential reproductions as unities. The consciousness may be very “weak,” lacking “the clarity to stand out,” with the incumbent risk of misidentifying recognitive AUA-concepts with a consciousness limited to “the end result but not the action of generating the representation.”17 Nevertheless, it is enough to make us conscious of the rule our associative synthesis obeys: its components as well as the order and manner of how they belong together. Recognition of the rule of synthesis constitutes the difference between a mere aggregate of actions blindly performed and consciousness of the unity of the synthesis, so that the whole remains before our eyes from first step to last, each recognized, however darkly, as normatively necessitated by the previous and necessitating its successor. Moreover, insofar as conceptual consciousness of the rule renders facility affect redundant, we can recognize instances of the rule even where facility is weak, canceled out by something else, or otherwise absent. This enables us to make comparisons with similarly conceptualized syntheses freed from the affective leading strings of association and custom, thereby making possible representations of new, more abstract relations as well as consciousness of more general rules of which the original rules are thereafter recognized as instances—all without benefit of language. Thus, empirical synthesis of recognition results in a whole new, more intellectualized level of conscious representation featuring synthetic necessities under rules. However, recognition is not yet cognition of an object, nor are rules laws. A purely empirical synthesis of recognition merely gives the normative form of a rule to the deliverances of associative imagination, and association is far too beholden to the contents of sensation and feeling to be capable of objectively singling out spaces, times, or the physical realities that fill them (objectivity for Kant, it will be recalled, minimally requires independence of sensation as regards both existence and individuality). To rectify this, the transcendental synthesis responsible for schematized transcendental space and time needs to be incorporated into all empirical synthesis a priori, so that every product of the latter will be represented as embedded in the transcendentally constituted realm of nature. This is where the conceptual character of synthesis of recognition proves crucial. For while associative synthesis is purely sensible and so completely incommensurate with a propositional network like transcendental space and time, the concepts that result from affixing AUA to

The I of Nature 277 associative syntheses in synthesis of recognition are anything but. They can be incorporated into the network simply by adding the transcendental content furnished by schematized pure concepts of the understanding to their own. Once embedded, recognitive concepts, together with the associative syntheses represented in them, are ipso facto converted into representations of physically real, spatiotemporally individuated objects subject to laws of nature, thereby elevating empirical synthesis— subjective experience—to the epistemological status of genuine cognition— objective experience. Kant’s term for incorporating transcendental synthesis into empirical was transcendental judgment. “Judgment” here signifies the subsumption of the particular case under the general rule. When judgment is empirical, the cases are whichever realities instantiate the concept, as an actual bird instantiates the concept of a bird or an actual quantum computer instantiates the concept of a quantum computer. The instance validates the concept as a genuine concept of an object, as is evident when instantiation is lacking or impossible, as with the concepts of Big Foot, the Flying Spaghetti Monster, telepathy, prophecy, the trisection of an angle by rule and compass, and the whole number between seven and eight. In the case of purely recognitive concepts that do no more than give the normative form of a rule to otherwise blindly formed and executed associative syntheses, the instance is the physical reality experienced through sensation responsible for the association being formed in the first place. AUA-concepts like these, formed directly from physical reality without abstraction, generalization, or reliance on linguistic convention, are effectively their own instances, since the only difference between the recognitive rule and the associative case is the contentless, purely formal logical unity of universal self-consciousness supplied by the analytic unity of apperception—a difference that leaves no scope for either false instantiation or failure of instantiation. When recognitive AUA-concepts are embedded in the network of transcendental space and time, their instantiating realities are ipso facto subsumed under the transcendental schemata and so become instantiations of the latter as well. In other words, association provides particular cases from physical reality that confer on transcendental concepts the same empirical validation that only instantiated empirical concepts are otherwise capable of attaining. The result is judgment that is transcendental yet with genuine empirical-cognitive worth. Moreover, subsuming the physical realities underlying associations under the schemata in transcendental judgment not only confers empirical-cognitive worth on transcendental concepts that, as pure, otherwise lack it, but conversely confers spatiotemporal physical objectivity on realities that, being recognized strictly empirically, would be purely subjective were it not for their subsumption under transcendental concepts in transcendental judgments. So far as Kant was concerned, this reciprocal validation/objectification is all

278  Nature in Mind: Through Kant’s I’s that is required to prove transcendental judgment to be both necessary and sufficient for experience of genuinely objective spatiotemporal physical realities. In its absence, transcendental concepts are as empirically suspect as concepts of centaurs and leprechauns, while AUA-concepts originating in recognitive synthesis can never enable us even so much as to think, much less cognize, sense-divide transcending realities that are independent of sensation not only in existence but in individuality as well. Transcendental judgment thus emerges as the locus where a priori and empirical psychology converge to constitute a single unified, genuinely cognitive consciousness. And it is only in transcendental judgments that the pure concepts of the understanding utilized in transcendental schematism first truly earn the right to be deemed categories: concepts of objects so fundamental that they are presupposed by all others while themselves presupposing none. By way of illustration, imagine a person living alone in the wilderness. To completely bracket out the effects of linguistic convention, let us further suppose him to have been raised in one of the very last groups of humans who were complete strangers to anything resembling modern grammatical language and whose socialization was limited to being taught paleolithic survival skills. One day, judging from the position of the sun, he decides it is time to make his way home. Having traveled the terrain all his life, his associative imagination immediately reproduces the entirety of the route he must traverse to get there from where he is, including the various alternatives, their attendant obstacles and dangers, preferential rankings from, e.g., quickest to slowest, safest to riskiest, easiest to most difficult, etc., with volition jointly determined by the facility vivacity and desire/aversion affects involved in these actions. It does all this with such great facility and rapidity as to exceed his capacity to discern and attend to each step in the association individually, so that arriving at his decision seems to him like a single instantaneous ­representation—if it even breaks the surface of dark consciousness enough to be noticed at all. Being equipped with the same a priori psychology we have, the man is further able to transform his associative syntheses into AUA-concepts that provide him with recognitive consciousness of the syntheses as determined by rules that prescribe their components in their proper sequence and relations without any longer depending on the facility feeling constitutive of preferential reproductions. The way home is no longer chosen blindly, by affect alone, but intentionally, according to a regulative norm—even if the conceptual consciousness of doing so still remains too weak for him to discern and attend to all that is transpiring in him (i.e. all the appearances darkly apprehended in internal intuition). This, however, is as far as mere empirical syntheses of recognition in AUA-concepts can take him. To cognize the journey he is representing objectively, he must first embed it in schematized transcendental space and time, which alone can confer sensation-independent objective

The I of Nature 279 existence and individuality on the clearing, the stream winding through it, the trail and trees beyond, the hills, sky, and all the other spatiotemporal appearances he apprehends through sensation, recollects in memory, and reproduces in associative imagination. To do this, he need only use transcendental judgment to subsume his recognitive concept of the journey under transcendental schemata. In that way, what he otherwise represents purely aesthetically as spatial and temporal appearances, individuated merely subjectively via the sensations through which they are apprehended—the green meadow, the cold stream, the sweet-scented flowers, etc.—acquires the extra, purely propositional dimension of sensation-independent, properly categorial objectivity. One effect of this extra dimension is to make him conscious, if only darkly, of the extent of his ignorance even regarding the things most familiar to him. For once appearances are cognized as embedded in a vast, largely unperceived, partly inaccessible network of causally interacting, permanently existing spatial substances with their own sensationindependent quantitative and qualitative determinateness, he implicitly understands that the sphere of reality demarcated by his recognitive concepts may be no more than the tip of a vast, potentially infinite iceberg. This applies not only globally but to each object insofar as it now is recognized as having its own, sensation-independent individual nature in which a potentially infinite multitude of accidental properties and causal connections may lie concealed, just like the waxen nature underlying the variable appearance of Descartes’s beeswax. Consciousness of such immense ignorance cannot but help to stimulate our ancestor’s curiosity and turn him into a seeker after information to which he would otherwise be oblivious. To this end, his first impulse might be to use his senses to discover more associations so as to fashion new and better recognitive concepts, which he can then compare, abstract from differences, and thus use to form new, more generally applicable ones. But even though independent of any one associative synthesis, the resulting AUA-concepts cannot help being just as sensation-dependent and subjective as the syntheses from which they take their start, and so do nothing to narrow the gulf of his ignorance regarding genuinely objective properties and connections. Instead, he must use his encounters with objects in experience to refine his understanding quite specifically of their objective, transcendental schemata-derived character. It is then no longer enough to recognize, e.g., fires to be regular antecedents of smoke. He must additionally cognize fires as causes of smoke, so that their relation is no longer represented as merely a subjectively grounded normative rule but rather the objective law that whenever there is fire there is necessarily smoke. Since this law is not a priori, our ancestor also implicitly recognizes that it admits of being falsified by even a single contrary instance. Accordingly, if he encounters fire but no smoke or smoke without any fire preceding

280  Nature in Mind: Through Kant’s I’s it, his puzzlement may well induce him to undertake a minute examination of the circumstances and/or to perform trial and error experiments to determine if there is something he missed. That, for Kant, is the key: to elevate subjective rules to objective laws, our ancestor must break free of nature’s leading strings and let his reasoning be guided not only by the appearances whereby he exhibits his sensations but by the transcendental concepts contributed to synthesis of recognition by his own psychology a priori (“what reason cannot know from itself but has to learn from nature must be sought in what it has itself put into nature”).18 And what is true of our pre-linguistic ancestor remains just as true today when we analyze the blue of the same sky he beheld in terms of the causal relations between photons emanating from the sun and the molecules of earth’s atmosphere, the numbing cold of the stream he waded across in terms of the average rate of collision of the H2O molecules composing it, and so on.

The Mortal Soul and the Unity of Time Association was presupposed in the foregoing consideration of transcendental judgment. But at its most fundamental level, transcendental judgment precedes and makes possible association itself.19 This is because preferential reproduction, in all but its simplest, most immediate varieties, depends on custom (habitual transitions of thought), which in turn requires at least dark consciousness of the frequency and constancy of the occurrence or co-occurrence of sense perceptions. Such consciousness is possible only through their representation as events occurring in an everchanging (non-repeating), irreversible (ordered), linear (unbranching) time in which each moment is objectively individuated—uniquely differentiated and completely determined—in relation to every other. Only then, with the existence of each new and recollected perception represented as an event occurring objectively before, during, or after this, that, and every other—indeed so much before, so long during, and so much after—can their frequency and constancy be consciously experienced and a cognitively utilizable customary association formed. And, as analyzed in Kant’s a priori psychologism, this representation of time is possible only in and through the propositional network of objective transcendental space and time produced via transcendental schematism. Kant further recognized that if even customary association is impossible without the objectivity contributed by transcendental schematism, empirical self-consciousness must be no less so. Like Hume (Chapter 8), he held that “all that is in internal sense is in constant flux”; and since “in bare succession existence is always only vanishing and recommencing and never has the least magnitude . . . everything in what we call ‘the psyche’ is in continuous flux and nothing abiding.”20 Not only does the absence of abiding existence preclude consciousness of either the transitions of thought constitutive of direct associations or the frequency and

The I of Nature 281 constancy of the occurrence and co-occurrence of external appearances requisite to form associative customs, it also means that “there can be no standing or abiding self in this flux of internal appearances.”21 This refers to the self as apprehended empirically in internal intuition but holds equally for the most constant representation of all, “the [pure] representation I, which accompanies and connects all representations, [since] we can never make out whether this I, a mere thought, is not just as fleeting [in internal intuition] as the thoughts that are linked to one another by it.”22 All this is just to say that no internal sense perception, and so too no empirically apprehensible consciousness of any kind, can have the least duration given purely aesthetic internal intuition alone, where each new instant of consciousness is in and of itself an absolute beginning and terminal end. For “if everything is in flux, then the flux itself cannot be perceived.”23 This, however, raises the question of how the empirical self, as object exclusively of internal intuition, can be cognized at all. The problem is this. Before any representation can be made into an object, it must first be converted from a purely aesthetic sense perception into a recognitive AUA-concept; for, as we have seen, transcendental judgment cannot directly incorporate purely aesthetic representations into the propositional network constituted by transcendental schematism. But how can the flux of internal sense yield such a concept if it cannot even be perceived, darkly or otherwise? The answer emerges as soon as we recognize what follows given that the object of self-cognition, alone among empirical objects, is experienced uniquely through internal sensations and through all indifferently (Chapter  8). External sensations represent different objects—a bird, a tree, a cloud, another human being, etc.—and so need to be empirically differentiated and the result synthesized into segregated groupings before they can be recognitively conceptualized and cognitively related to their particular objects via schemata. Internal sensations, by contrast, all relate to one and the same object, the empirical self, thereby obviating the need to empirically differentiate or discriminate between them, synthetically group them, or normatively order them under rules of synthesis before relating them to their empirical object. Since this is just to say that empirical psychology has no role to play in forming the representation of the empirical self, why then can we not make do with an AUA-concept of the flux of internal sense drawn from pure intuition instead? A  purely formal concept of the flux derived from pure intuition is able to represent the successive intuitions composing it just as indifferently and indiscriminately as the empirical concept of the flux we would derive from them if, per impossibile, the flux could be perceived. There is thus nothing to prevent the former from being subsumed under the transcendental schemata in transcendental judgment in place of the latter for purposes of empirical self-cognition.

282  Nature in Mind: Through Kant’s I’s Given an AUA-concept of the flux of internal sense, the next requisite for self-cognition is personal identity—the representation of something that persists identically in existence through physically real time: We can only notice change in that which persists. If everything is in flux, then the flux itself cannot be perceived. Experience of coming to be and ceasing to be is therefore only possible through that which persists.24 Yet because an empirical concept of something persisting through the flux can be no less impossible than an empirical concept of the flux itself, Kant saw no other way to meet the need for a persistent than via the transcendental schema of the category of substance. The schema is the representation of a physically real spatial something in general with “an inner necessity of persisting”25 possessed of various determinations (accidents), each of which is embedded in a no less physically real causal network that temporally connects it with other such determinations as its causes or effects. In the case of the empirical self, this means cognizing the flux of internal sense as successive changes in physically real spatial substances persisting through time. Moreover, thanks to the schema of cause and effect, the changes experienced in one’s persisting self are also cognized as effected (causally necessitated) physical changes. As such, each instant of consciousness in one’s life is set in objective, physically real transcendental time, where it occupies a completely determined, uniquely differentiated position within a single unbranching, unconverging, ever-changing, irreversible, physically real objective time-series encompassing all events in nature, with each individually inhering (as final-predicate accidents) in physically real spatial substrates. Thus, the subject can cognize itself empirically as a physically real object existing in nature and everywhere subject to natural law just like any other, and more particularly as the accident of a spatial substance comprising a succession of representational events (consciousnesses) that continues just so long as the physical basis for that succession persists. The physicalism affirmed in Kant’s psychologistic explication of selfcognition undercuts both Cartesian style substance dualism and property dualism. Since “in time no parts whatever are simultaneous, all are successive,”26 a purely temporal persistent is impossible. Because the self is an object experienced exclusively through internal sensations (Chapter  8), it is consequently impossible to cognize it as a substance in its own right but only as the accident of a persisting substantial substrate existing physically in space. This enabled Kant to deny the substantiality of the self on a priori psychological grounds alone, quite independently of empirical cognition of brains, bodies, and other physical objects, much less the existence of correlations between particular conscious and particular neural occurrences (neural correlates). Nor can

The I of Nature 283 property dualism get a grip here since Kant’s distinction between the self and its body is entirely a function of the distinction between the objects of external and internal sense. This is a purely psychological duality: diverse expressions of productive imagination—apprehension in external-spatial or internal-temporal intuition—no different in principle than differences between species of sensation (feeling vs. hearing vs. seeing, etc.). Clearly, no such duality can suffice even for conceiving, much less cognizing, the kind of metaphysical duality between mental and somatic properties that property dualism affirms and monism, in both its idealist and materialist guises, rejects. And as for the things in themselves underlying the two species of appearances—transcendental objects and transcendental subjects—Kant deemed the question of whether there is any fundamental distinction between them to be not merely unanswerable but unaskable.27 Because of its complete indifference to the particular sensations that happen to be present in it, Kant’s objectively enduring empirical self more closely approximates traditional notions of the soul than the self of contemporary analyses of personal identity. Suppose, for example, that just before your brain triggered your first (neural correlate of) consciousness in utero, it had been transplanted into a suitable housing—the skull of another fetus, a cyborg, a vat, etc. Suppose further that its new housing existed at a radically different place and time, so that most of the external sensations you have had would not be found in this new “you,” with the consequence that the internal sensations they arouse would also be different. With a completely different psychobiography, even a different biology (or machinery rather than biology), would you still be yourself or someone entirely different? Here is where Kant drew the line between the objective empirical self and the subjective. The latter is the self whose individual identity is determined a posteriori by the particular external and internal sensations perceived in it, in precisely the same manner we individuate spaces subjectively as the blue sky above the green pasture or times by the painful fractures and contusions suffered after falling down a hillside. However, so far as Kant was concerned, neither individually nor collectively are these subjective criteria of personhood the objectively real empirical you. For however subjectively unfamiliar the experiences you might have had you been transplanted in utero, the conscious life of that brain would still, in every objective sense, be the life of your mind and its demise your death. By contrast, had another fetal brain taken your vacated place and circumstances been contrived so that there was no subjective difference between the life of its mind and the life you have actually lived, that brain would still not be you, and its death would have no more consequence for you than the demise of any of the countless other humans unknown to you. And when the self is explained so objectively that all subjective criteria of personhood are excluded, the result is indeed most properly described as the mortal soul.

284  Nature in Mind: Through Kant’s I’s There still remains the question of why, once the schema of substance has performed its role in making empirical self-consciousness possible, it cannot simply be dropped from the concept of the latter as an objective, physically real existent. Is there any further reason it is necessary to conceive the mortal soul as the accident of a substance with “an inner necessity of persisting,” which consequently must be conceived as existing permanently (never beginning, never ending), in addition to being conceived as the accident of a brain, human body, or any other persisting, yet ephemeral physical object? The answer is the unity of time on which Kant held the possibility of both customary association and objective experience of the self or anything else to depend. If there is to be a single linear time in which all possible sense perceptions occur and co-occur with objectively varying frequency and constancy, that time cannot be represented as limited to the life and experiences of one’s own or any other mortal soul. It must instead be indexed to something external, existing physically in space, in relation to which the beginning, duration, and end of souls can be represented as causally effected changes occurring in something that persists before, during, and after the changes. Similarly, if the beginning and end of a storm, a tree, a drought, and any number of other particular physical things are not to be represented as absolute beginnings and ends, there must be some substrate in relation to which all count not as persisting substances but as ephemeral accidents, whose beginnings and ends are merely events in the continuing existence of what was there before, during, and after. Early humans may well have regarded mountains and other features of the surrounding landscape as well as the earth beneath and the sun, moon, and stars above as permanent—but only in a comparative sense, not by any inner necessity of persisting. For the representation of the unbounded linear time of transcendental schematism was presumably an integral part of the association-based cognitive experience of our species from its inception, if only in dark consciousness. This would have enabled our ancestors to recognize that mountains et al. are things whose beginning, duration, and end are cognizable in principle, and might actually have been cognized if souls existed long enough to have the requisite sense perceptions. Indeed, using their transcendental schema of permanent substance, there is nothing in experience that they (or we) could not represent as an ephemeral occurrence in a time of sufficient scope to encompass all events, past, present, and future, however transient (e.g. virtual particles) or protracted (e.g. the still ongoing Big Bang expansion).28 And all this our ancestors could do perfectly naturally, by psychologistic means alone, without need for language, culture, or social conventions of any kind, much less platonists’ mind-independent eternal representations.

The I of Nature 285

Transcendental and Empirical Cosmology What about the physical universe in its totality? With regard to today’s prevailing cosmological conception, for example, some speculate that the Big Bang may very well be an absolute beginning that consequently cannot be cognized as an event (effected change) in a time extending further into the past, and that the same may also be true at the opposite temporal extreme if the Big Bang universe has an absolute end. If this is correct and the Big Bang’s beginning has no temporally antecedent cause and/or its end no posterior effect, is that compatible with there being permanent substances with “an inner necessity of persisting”? It may seem not. But here Kant would distinguish two distinct objective representations of the physical universe, only one of which is incompatible with the universe having an absolute beginning and/or end. I shall henceforth refer to the one that is compatible as “the universe” and the one that is incompatible as “the Universe.” The Big Bang universe is not a Universe because there is no saying it constitutes the totality of physical reality. As presently understood, it is impossible to preclude that there may have been a prior cosmological era, perhaps even an infinite number of them, or that there may be a successor era, perhaps also infinite in number.29 Nor can any number of conceptions of the Big Bang universe as part of a vastly greater multiverse be ruled out. In the former case the Big Bang universe would not coincide with the unity of time; in the latter it would not coincide with the unity of space, and so could not be equated with the Universe. Even if the Big Bang universe has no predecessor or successor eras and there is no multiverse, it lacks the sort of inner necessity to be represented cosmologically as the absolute totality of physical reality. The Big Bang universe is cognized through sense perceptions of human brains. Humans are creatures constituted entirely of electromagnetic forces operating at the atomic and molecular scales within and between cells, and at the macro scale between larger bodies, inanimate as well as animate, existing, like us, in three spatial dimensions. The subatomic scale weak and strong forces can be detected only insofar as the three-dimensional electromagnetic apparatuses we build and can interact with exhibit properties that can confidently be attributed to non-electromagnetic causes. Gravity too, consisting as it does in the curvature of spacetime by massive bodies, can be detected only via its effects on the motions and collisions of electromagnetically constituted bodies. If there are other forces and dimensions out there, their direct and indirect effects on three-dimensional electromagnetic human bodies have yet to be detected. It may be that they have none, that their force is exerted exclusively at humanly inaccessible scales, in humanly inaccessible dimensions and/or on humanly inaccessible kinds of entities. Why consider them physical in that case? So far as

286  Nature in Mind: Through Kant’s I’s Kant was concerned, any reality that exists, acts, and is acted on in space, at any scale in any dimension, is ipso facto a physical reality. Thus, even if physicists do someday come up with conclusive proof that the Big Bang universe is, quite literally, all there is, the claim would still be limited to the three-dimensional, four-force30 universe humans are able to detect, not to the physical Universe in its cosmological completeness. The foregoing should make clear that the universe/multiverse of physics is a parochially human, merely quasi-cosmological conception, whereas the Universe is a transcendental, genuinely cosmological one. Kant’s concern, as transcendental philosopher, was primarily with the latter. He readily acknowledged that there can never be empirical knowledge of the Universe’s existence, starting with the empirical impossibility of discovering anything that owes its permanence to an inner necessity of persisting. So, even if one were to grant that intrinsically permanent substances were essential to a genuinely cosmological conception of nature, they have to be validated transcendentally (i.e. a priori psychologistically) or not at all. In particular, their warrant rests exclusively on the cosmological demand for unity of time and space, a requirement that ultimately depends on the objective unity of apperception, and is satisfiable only by subsuming the matter of the appearances that exhibit sensations in sensibility under schematized categories in transcendental—not scientific—judgments. So far I have focused mainly on the role of the categories of substance/ accident, cause/effect, and their correlated modal categories in the constitution of the objective unity of space and time. But here mention must be made of a third relational category of particular importance cosmologically: the community formed by causally interacting substances. A permanent substance that persists by inner necessity may seem sufficient for the unity of time. But what if there were two such substances and nothing to connect them? Given that objective time, as Kant analyzed it, is entirely a function of the causal (irreversible ground-consequent) relations connecting the accidents (final predicates) of spatial substances, the absence of causal connections between them would make the two substances denizens of distinct, incommensurable unities of time. What was objectively before or after in relation to the one would be neither in relation to the other, since events and their times in the one would be completely undifferentiated and indeterminate in relation to the events and times in the other. Similarly, there would be no differentiated, determinate place of the one substance in relation to the other, much less determinate routes of determinate distances, directions, and dimensions. The case would be similar to the way different universes in the multiverse are sometimes portrayed: causally completely cut off from one another, with no greater, all-encompassing spatial or temporal network to unify them. The schematized category of community precludes this possibility in the Universe, if not the universe/multiverse. Without delving into details, the subsumption of appearances under the schema of community

The I of Nature 287 in transcendental judgment effects the Universal unity of space and time by ensuring that there can never be a time when some accident of every spatial substance is not in direct or indirect causal connection with some accident of every other.31 Physicists sometimes talk of attaining a universal (lower-case ‘u’) “theory of everything.” Yet, the more seriously we take the notion, the less plausible it seems. Suppose fifty years from now physicists agree that they have conclusive evidence that their theory is a theory of everything. This would mean that no physical realities are excluded from its scope or left unexplained, and that another fifty years, five hundred, fifty million, or any number of years of continued research, observation, and experiment could never turn up anything inexplicable by the theory. It would further mean that genetically engineered humans, cyborgs, or whatever else might be devised that vastly exceeded current human intellectual and technological capacities could also never turn up anything inexplicable by the theory, no matter how long such advancement was to proceed (e.g. humans engineering super-humans who in turn engineer creatures superior to themselves and so on in every generation for  millions of years). And it would mean that there could not be intelligent creatures so radically different in nature from us and anything we could become or construct that their physics of spatiotemporal reality could contain things both unimagined and unimaginable in our puny human “theory of everything” (e.g. because they exist at scales far below the Planck minimums that today seem to limit human physics, or far above the scales of the largest multiverse inferable by human physicists, or in other dimensions, or are for other reasons beyond even the most super of super-human ken). It thus seems safe to conclude that the chances of any envisageable universal “theory of everything” truly living up to its cosmological billing are vanishingly slim. The parochial four fundamental force universal community of impermanent “substances” of empirical human cognition must always be carefully distinguished from the Universal community of genuine substances of transcendental (i.e. a priori psychologistic) cognition that persist through their own internal necessity. Perhaps the key difference between the genuinely cosmological notion of a Universe and the quasi-cosmological notion of a universe/multiverse is that the former, as transcendental, is not moored in our electromagnetic constitution or the particular material nature of any sensate soul, possible or actual. Each physical being brings with it constitutional limits to its interaction with other physical beings. Of course, we can neither rule in nor rule out the possibility of souls whose sensations have radically different physical correlates than ours, whether organic but nonneural, inorganic, non-electromagnetic, far beneath the Planck scale minimums, far beyond the scale of any multiverse humanly hypothesizable, non-three-dimensional, or anything else. But Kant’s transcendental psychologism implies that our minds are equipped to experience not only

288  Nature in Mind: Through Kant’s I’s a universe/multiverse through sensations moored in the electromagnetic constitution of the human body, but a Universal community of permanent substances freed of these and all other creaturely moorings. This Universal community is (1) spatiotemporally unified by the same inner necessity whereby each substance individually persists (thus precluding a capital‘M’ Multiverse) and (2) accessible through sensational consciousness irrespective of its physical correlate and its intrinsic physical limitations. As such, the Universal community encompasses not only the universe/ multiverse accessible directly or indirectly through human sense organs, but also any realities existing in space beyond human observational and experimental ken, now and possibly forever, yet accessible through actual or possible sensations moored in physical correlates whose constitution may not only be unknown to us but is quite possibly inconceivable by human brains. Yet, insofar as the very same external sensations that give us access to the universe of ordinary and scientific empirical cognition simultaneously provide access to the Universe of transcendental cosmological cognition, empirical and transcendental cognition relate to one and the same objective physical reality. The upshot is that whereas the Universe bottoms out on a spatially and temporally united community of permanent substances that persist by inner necessity, the universe/multiverse neither does nor can. The latter is limited to mathematical representations of physical reality as experienced (via observation and experiment) through human sense organs. The Universe, by contrast, is not an empirico-mathematical but an a priori psychologistic structure formed by the a priori subsumption in transcendental judgments of all possible sensation-based experiences under the schematized categories, and so regardless of the physical constitution correlated with them a posteriori. Cognition of the Universe is, to be sure, very limited in comparison to the universe/multiverse since the Universe must be known transcendentally, by a priori psychologistic means, or not at all. Nevertheless, because of its independence of everything specific to the limitations of human or any other creature’s physical nature, genuine cognition of the Universe is objective to a degree no human science can approach. Thus, rather than claimants to the same territory, transcendental philosophy and humanly accessible physics should be regarded as non-overlapping, mutually complementary magisteria.

Nature’s Amazing Mathematizability Ordinary understanding of nature, without the guidance of mathematics, is largely limited to association-based recognition, and this is never truer than when it comes to the universe itself, where the scientifically illiterate are largely confined to cosmogony. In Kant’s view, the only way to supplant cosmological mythologizing with true scientific understanding of nature is through the application of mathematics (“there can be only as

The I of Nature 289 much genuine science in it as there is mathematics”).32 Nor can anyone today doubt that relativity theory, quantum theory, and any foreseeable successor theories represent nature only insofar as the mathematics requisite to do so has been invented and scientists have discovered how to apply it. It is not just that the most fundamental laws known to us take mathematical form, but that they never could have been hypothesized or the hypotheses confirmed if scientists had not been able to supplant ordinary imagistic and linguistic representation with mathematics. Since at least Galileo, scientists, philosophers, and other cognoscenti have marveled at the fact that nature’s laws, its constants, and virtually everything else about it not only is amenable to being expressed mathematically but cannot even be so much as conjectured to exist unless and until all the relevant phenomena have been so expressed. Why does nature seem always to exist and operate on purely mathematical principles? Why should it be amenable to mathematical expression at all, much less in all its aspects? There is no universally accepted explanation why this should be so, much less how it can or must be so. But that it certainly seems to be so has convinced some, such as the physicist Max Tegmark, that the universe itself is a work of mathematics—self-constructing, selfactualizing calculi. Kant, who was as struck as anyone by the seemingly limitless mathematizability of physical reality, deemed it one of the singular virtues of his philosophy that it can explain how this state of affairs is both possible and necessary. The reason lies in his account of the possibility of mathematics in terms of the transcendental space and time that result when the manifolds of purely aesthetic, formally intuited space and time are converted into a propositional network and determined conformably to pure concepts of the understanding via transcendental synthesis (Chapter 9). Since the sole and entire function of transcendental schematism is to extend transcendental synthesis from the form to the matter of appearances, any physical reality whose representation incorporates transcendental schemata is ipso facto mathematizable. Given that the Universe is constituted by subsuming the totality of empirically apprehensible physical realities under transcendental schemata in transcendental judgments, it follows that nothing in it can fail to be mathematizable. What about the universe/multiverse? First and foremost, it is an empirical construct, and so rooted in experience. Conscious experience, if considered simply in terms of the individual isolated psyche—i.e. apart from language and everything else bound up with uniquely human sociality and lifeways—is primarily a function of customary association. Customary association depends on consciousness of the frequency and constancy of the occurrence and co-occurrence of sense perceptions, which in turn presupposes the representation of sense perceptions as events occurring in an ever-changing, irreversible, linear time in which each moment is objectively individuated (i.e. uniquely differentiated and completely

290  Nature in Mind: Through Kant’s I’s determined in relation to all the others). The representation of time that makes customary association possible is, according to Kant, none other than the Universal time of transcendental schematism. This, as we have seen, is ineluctably bound up with the Universal space defined by the transcendental community of permanent substances, each with “an inner necessity of persisting.” So, if, as Kant held, transcendental space and time also make mathematics possible, the implication is clear: the mathematizability of the objects we experience is built into the very possibility of that experience. This, to be sure, is merely ordinary experience, in which the potential to represent nature mathematically may be realized only in the most primitive fashion if at all. In the case of our species at any rate, language was key to tapping it, although still not enough. It was not until the advent of writing as a means of keeping records of what people owned, what they owed, and the extent to which they had met their economic obligations that the full mathematical potential that transcendental representations build into ordinary experience could be unleashed. Language, plus the use of writing to augment its representational power, has proven extremely successful over the millennia in enabling humans to develop and apply mathematical techniques in the cognition of nature. Even so, it is an imperfect instrument, and along with everything else that limits humans in their pursuit of scientific understanding, can get us only so far. Nature—its laws, composition, and actual workings—owes no more to human language and the technique of writing than it does to any other bodies of social conventions and the technologies that subserve them. How then can our feeble, parochial, all too human means of communicating with one another provide us with the key to unlocking nature’s deepest secrets? Science cannot now and may never be able to provide any more insight into its own ability to penetrate the secrets of nature mathematically than it can explain the mathematizability of nature itself. Kant’s way of meeting this challenge should now be evident. Science owes the incredible advances it has made by following the path of mathematization not to its social and social-psychological underpinnings, much less to mystical platonistic powers of insight, but to the nature of objective experience in the individual, isolated human psyche. Insofar as scientific representation, like linguistic representation generally, incorporates the logical forms of judgment (Chapter 7), it is no less capable of incorporating the non-linguistic propositional network constitutive of the physically real transcendental space and time that results from the subsumption of empirical synthesis under the transcendental schemata derived from these forms. So, if, as Kant maintained, these schemata make not only ordinary cognitive experience and its objects possible but mathematics itself, then the mathematical path to the understanding of nature is open to any human with the socially acquired competencies requisite to tread it. For there can then be nothing truly objective in experience—i.e.

The I of Nature 291 independent of sensation with respect to both existence and individuality— that can fail to be fully amenable to mathematization, provided only observation and experiment are able to reveal it and the requisite mathematical techniques have been devised to represent it. Put another way, the linguistically represented universe/multiverse necessarily conforms to, and is encompassed within, the inherently mathematizable Universe represented by purely a priori psychologistic means. All mathematical laws, constants, etc. that characterize the former are therefore founded on, and folded into, the transcendental laws constitutive of the latter. As transcendental laws are that which first makes mathematizable nature itself possible, Kant saw fit to single them out as the most fundamental of all laws of nature and accordingly denominated them principles. And because they are founded neither on empirical nor on mathematical concepts but solely on transcendental pure concepts of the understanding (schematized categories), he termed them principles of pure understanding. These transcendental laws are what most directly underpin Kant’s Copernican thesis that nature must conform to pure understanding just as surely as empirical understanding must conform to nature: [H]owever exaggerated and absurd it sounds to say the understanding is itself the source of the laws of nature, and thence of the formal unity of nature, it is also true and proportionate to the object, namely experience.  .  .  . There are thus certain laws that first make nature possible, and indeed do so a priori. Empirical laws can occur and be found only through experience, and indeed only in consequence of those original laws according to which experience itself first becomes possible.33 Insofar as I am an empirical I, or soul, I am a physical object like any other, subject in every way to laws of nature, and no more a free agent than a boulder tumbling down a mountainside to the valley below after being dislodged by an earthquake. But insofar as I  am a no less real (sum) transcendental I (cogito) responsible for subjecting all apprehensible appearances to a twofold transcendental synthesis constitutive of nature itself, nature and everything in it are subject to the transcendental laws my understanding prescribes.34 Again, none of this is to gainsay the impossibility of deducing or in other way extracting empirically grounded mathematical laws of nature from transcendental principles. The only way empirically grounded mathematical laws of nature could be extracted from principles of pure understanding is if the Universe were not just a necessary but also a sufficient condition for the existence of the universe/multiverse, which Kant decidedly did not claim. Mathematical propositions have always to be constructed, scientific ones discovered. By contrast, the transcendental laws governing physical substances, causality, and simultaneity are

292  Nature in Mind: Through Kant’s I’s completely indeterminate empirically as well as mathematically, and so cannot help being compatible with any system of empirically grounded mathematical laws of nature, past present or future. The equations of relativity and quantum theory, in particular, are just one set among the infinitely many mathematizations of nature that principles of pure understanding make possible. Yet, that is all the Kantian requires to maintain that the laws of nature mandated by the equations of relativity and quantum theory are merely “particular determinations of the pure laws of the understanding under which, and according to whose norm, those laws are possible in the first place.”35

The Subjective Core of Objectivity We have seen that Kant’s psychologism is founded on (1) the representation ‘I think,’ from which he, like Descartes before him, deemed it to derive a certainty immune to even the most radical skepticism. With this representation as its supreme principle, and more particularly the psychological reality it confers on logical universality and the logical universe, Kant presumed his psychologism to extend with equal certainty to comprise several other representations: (2) the sensible, purely aesthetic formal intuitions that make the quintessentially intellectual representation ‘I think’ possible by producing original synthetic unity of apperception (unity of sensibility) (Chapter 8); (3) the logical forms of judgment that give representational force to AUA-concepts in propositions, comprehend the resulting propositions within an extended synthetic unity of apperception, and thereby broaden the scope of the universality of ‘I think’ to include concepts and propositions as well as sensations and intuitions (Chapter 7); (4) the pure concepts of the understanding derived from logical forms of judgment by representing restrictions on the logical freedom inherent in those forms, thereby introducing extra-logical necessities into propositional representation (Chapter 7); (5) the objective unity of apperception yielded by transcendental syntheses in conformity with pure concepts of the understanding from propositions composed of AUA-concepts derived from the manifolds of formal intuitions (Chapter 9); (6) the infinite propositional network yielded by transcendental schemata from the matter of appearances that makes possible nature itself, understood as both “the sum-total of all appearances (nature considered materially)” and their “necessary conformity to law (nature considered formally)”36 (the early sections of the present chapter); and (7) the transcendental principles of pure understanding that subsume association, recognition, and propositional representation (linguistic included) under the schematized pure concepts of the understanding

The I of Nature 293 (categories), thereby defining the fundamental laws of nature whereby alone rules of subjective, non-cognitive experience can be transformed into laws of genuinely cognitive objective experience. I have deliberately omitted from this enumeration of the representations that Kant utilized psychologism to explicate the specific formal intuitions, forms of judgments, pure concepts of the understanding, transcendental schemata, and principles of pure understanding that Kant attributed to minds constituted like ours. This is not because space, time, categorical judgment, hypothetical judgment, substance/accident, cause/effect, number, et al. do not belong on a complete list. Instead, it is because Kant expressly refused to deny that formal intuitions other than space and time may exist that are equally capable of producing original synthetic unity of apperception, and perhaps do exist in other creatures. Nor did he preclude the possibility of creatures with other forms of judgment and pure concepts of the understanding that are no less capable of giving representational force to their AUA-concepts in propositions, extending the scope of their ‘I think’ to include propositions, and effecting transcendental synthesis. This is not to say that Kant affirmed the phylogeny dependence of these forms. He simply insisted that we are not, and can never be, in a position to preclude it. But even if other forms are for some unknowable reason impossible, space, time, categorical judgment, et al. still cannot be set on the same level of generality or fundamentality as the representations listed in (1)–(7). For none but the latter are capable of being determined by transcendental psychologism to be absolutely essential to any sensation-dependent cognitive understanding as such, regardless of the specifics of its physical (e.g. electromagnetic with terrestrial phylogeny) and mental constitution (e.g. reliance on customary association by resemblance and contiguity). If this distinction of levels is accepted, several things follow. First and foremost, nature, whether in the guise of the Universe or the universe/ multiverse, is in itself nothing other than unity of apperception in its final, fully objectified, physically realized guise: when we consider that this nature is in itself nothing but a sum-total of appearances, and so . . . merely an assemblage of the mind’s representations, then we cannot wonder at seeing it merely in the capacity at the root of all our cognition, namely in transcendental apperception, in that unity for which reason alone it can be called object of all possible experience, i.e. nature.37 Thus does the career of the ‘I’ in Kant’s transcendental psychologism culminate in its becoming the I of nature. Second, the kind of maximalist epistemology that rejects the distinction between understanding and knowledge seems to be predicated on

294  Nature in Mind: Through Kant’s I’s an assumption regarding the nature of consciousness that is seldom acknowledged or defended. This is the idea that consciousness is essentially a matter of not being asleep, or at least not being in a dreamless sleep. If this is indeed how it is viewed, then one will hardly be inclined to give much thought to whether and how consciousness might figure in equally essential ways in higher functions of mind, particularly those on which logic, mathematics, and science depend. It is this mindset that Kant’s transcendental psychologism challenges head on when, in the guise of (1)–(7), it sets the ‘I’ of a priori subjectivity at the core of logic as well as mathematical and scientific objectivity. Thus, unless and until one can prove Kant to be mistaken, the proposition that logical, mathematical, and scientific knowledge are phylogeny- and consciousness-­ independently objective, and so necessarily and universally just as they are for human understanding, must at least be acknowledged to be debatable, and so to require defending. Third, if space, time, categorical judgment, substance-accident, and the other transcendental representations not mentioned in (1)–(7) are phylogenydependent subjective variables, or at least cannot be known not to be, then the same must be true of the mathematics they precede and make possible, in both its pure and scientifically applied varieties. This means that if creatures with different but equipollent forms of sensibility and/ or understanding exist or are at least possible, they would have just as objective an experience as we do, but instead of mathematizing spatial and temporal appearances like us—geometrically, arithmetically, algebraically, mathematical-logically, etc.—their mathematizations, and consequently their science as well, would be incommensurably different. Accordingly, it would be not just our ordinary understanding of the world that would differ radically from theirs, but even the understanding we attain through our most abstract, highly formalized, best established present and future mathematics and science. Again, we cannot know whether creatures do or can exist that instantiate (1)–(7) differently than we do. But the mere fact that Kant’s transcendental psychologism is able to pose the question obliges us to be open to the idea that creatures that equal or surpass us technologically could do so without sharing any of our scientific understanding at all, be it quantum mechanics and relativity theory or anything else. This is why I have throughout stressed the importance of not conflating knowledge with the potentially phylogenetically dependent understanding through which it is attained. For without for a moment questioning the truth or objectivity of human knowledge in both its ordinary and scientific forms, Kant’s psychologism implies that we cannot preclude the possibility of an understanding of nature that is at once completely incommensurable with ours and nevertheless able to yield knowledge that is no less true and objective. Fourth, what is true of science generally must likewise be true of neuroscience: its cognitions, howsoever true and objective, are dependent

The I of Nature 295 on a sensibility constituted to exhibit sensations as spatial and temporal appearances and an understanding constituted to individuate these spaces and times via networks of propositions formed in accordance with the logical forms of categorical judgment, hypothetical judgment, and so on. While this does not prevent it from having the potential to fully explain the empirical subject (self, soul), it puts the transcendental subject forever beyond its reach: [T]he subject of the categories can obtain no concept of itself as an object of the categories from the fact that it thinks. For in order to think these categories, its pure consciousness, which has yet to explained, has to underlie the thought. In the same way, the subject in which the representation of time originally has its ground cannot determine its own existence in time by means of that representation. And if the latter cannot take place, neither can the former, as a determination of itself (as thinking being in general) by means of the categories.38 How can the subject be cognized as any kind of existent in time if time itself exists only in and through the syntheses of apprehension that subject performs in productive imagination? How can its performance of these syntheses be cognized causally if the concept of cause and effect itself derives essential components of its content from the apperception of that same subject, and is possible only in and through its pure understanding? And how can the subject be cognized as either substance or accident if these and the other categories likewise derive essential components of their content from this subject’s pure self-consciousness and are possible only in and through it? If nature itself, both as Universe and universe/multiverse, is the creation of the subject, then that subject can no more be part of nature than the dreamer of a dream can have dreamed itself up. This is not to deny that one can seek and potentially find neural correlates of the ‘I think’ and other transcendental representations synthesized by the conscious psyche. But if the objects of neural science, like nature itself and its laws, are impossible apart from a consciousness equipped with all of (1)–(7), it does mean that we are utterly incapable of penetrating this transcendental dream to discover the first thing about its dreamer beyond the mere fact of its existence: cogito ergo sum.

Transcendental Idealism and Things in Themselves The distinction between understanding and knowledge goes to the heart of the doctrine whereby Kant distinguished his philosophy from all others: transcendental idealism. Everything subject to conditions (1)–(7), including physical objects and souls, is ipso facto dependent on the

296  Nature in Mind: Through Kant’s I’s particular constitution of the representing subject and therefore incapable of existing consciousness-independently in itself. But because these conditions are one and all a priori and moreover suffice to establish the independence of these objects vis à vis sensations in both existence and individuation, Kant’s transcendental (i.e. a priori psychologistic) idealism in no way compromises the objectivity and truth of ordinary, scientific, and mathematical cognition as the purely empirical idealisms favored by Berkeley, Hume, and their successors arguably do. The only a priori idealist alternative to Kant’s transcendental psychologism is the supposition that we have in our possession representations of objects that are subjectively completely unconditioned, i.e. platonic ideas. But in addition to all the debilities to which platonism is prey, Kant could see no way to reconcile it with the actual workings of human psychology. For if his transcendental psychologistic anatomy of the intelligent mind is correct, then the absence of (1)–(7) means there could not be intuitions/sense-divide transcending appearances, SUA/aesthetic individuality, the ‘I think’/logical universality, AUA-concepts/normative recognitive syntheses, logical forms of judgment/non-linguistic propositional thought, pure concepts of the understanding/metaphysical categories, transcendental synthesis/mathematics, transcendental schematism/ nature, or principles of pure understanding/laws. In other words, far from human consciousness being able to revel in the representation of a consciousness-independent objective reality, freedom from the constraints of a priori psychologistic idealism would reduce human consciousness to a level barely distinguishable from Locke’s oyster.39 The only consciousness-independent reality that transcendental idealism leaves in its wake is the completely unknowable “transcendental matter of all objects as things in themselves (facticity, reality) . . . that corresponds to sensation.”40 As noted earlier, in the absence of any content whereby to represent these objects, we can do no more than indicate them by an X, where the ‘X’ is not even so much as a placeholder, much less a variable, since it is impossible in principle to produce any representation to replace or instantiate it. This, however, raises a question: if Kant’s transcendental psychologism implies that things in themselves are unknowable, indeed completely unrepresentable, how could he nevertheless posit their existence as even so much as an X, much less affirm it? He could not propose them as causes of sensations since the psychologistic explication of the category of cause and effect shows it to be a subjectively conditioned concept, intrinsically bound up with transcendental apperception, that consequently has no more application to subjectively unconditioned objects than pleasure and fear do. Since the same is true of substance and accident, number, space, time, and all other transcendental representations that instantiate (1)–(7), how could Kant nevertheless suppose that sensations furnish the means to conceive the unrepresentable and affirm the unknowable?

The I of Nature 297 It is often supposed that Kant contradicted his own injunction against inferring the existence of anything not giveable in experience from something given in it when he posited and affirmed things in themselves as “corresponding” to sensations. In fact, he had no need for any such inference from sensation because transcendental idealism permits the concept of subjectively unconditioned transcendental reality to be derived directly from sensation itself, analytically, with no need for inference at all. To see how this works, recall Kant’s distinction between sensations and appearances from Chapter 8. Every other psychological philosophy identifies sensations with spatiotemporal appearances, thereby treating their occurrence and co-occurrence in consciousness as events in nature no different from physical events of every kind. So construed, sensations can provide no other concept of reality than physical. According to Kantian transcendental idealism, however, sensations themselves are neither spatial nor temporal, but instead are exhibited in time through non-sensational appearances in apprehending imagination, and so are only comprehended within the unity of sensibility constituted by the formal intuitions of space and time at second hand. This means that in sense itself (a priori synopsis), prior to and independently of apprehending imagination and apperception, sensations are as atemporal as they are aspatial, and, as such, are equally prior to and independent of the spatiotemporal objectivity of events and enduring objects comprehended within the propositional networks formed by transcendental synthesis and transcendental schematism. Thus, on Kant’s theory of sensibility uniquely, the concept of reality obtainable directly from sensations cannot be the same as the physical reality that pertains to the sense-divide transcending spatiotemporal objects existing in nature under subjective conditions (1)–(7). If not a concept of the physically-real-yet-transcendentally-ideal, then what is it? To see that the reality given with sensation is subjectively completely unconditioned, one has only to recognize that, on Kant’s analysis, sensations can be represented as subjectively conditioned in respect to only their quality, not their existence. This is easily done. We have no means to cognize existential conditionedness other than by conceiving something to be the effect and/or accident of something else. Since the psychologistic explications of these concepts reveal them to be subject to all of conditions (1)–(7), and so to be grounded on the transcendental subject, the latter cannot be represented either as cause (or effect) or as substantial substrate (or accident) with respect to anything else without vicious circularity, sensations included. Even if this contradiction is ignored, since sensations, in contrast to appearances, exist entirely independently of (1)–(7), they too can no more be represented as causally or substantially dependent on anything else, including the transcendental subject, than the latter can be represented as their cause or substrate. Thus, if sensations are to be represented as subjectively conditioned in

298  Nature in Mind: Through Kant’s I’s any manner at all, it can be with regard to only their qualitative constitution, not their existence. The following example suggests how. As presently constituted, the human visual system terminates in neural correlates of sensations of colors of varying luminousness. However, given a relevantly different phylogenetic history, or futuristic genetic engineering, the same neural processes might have terminated in neural correlates of sensational consciousness completely incommensurable in quality with colors or indeed any of the other sensations known to our species. Of course, transcendental idealism requires that we eliminate all reference to transcendentally synthesized objective physical entities like neural correlates and their causes (as subject to conditions (1)–(7)), and restrict our consideration to the mere existence of a sensation of this or that particular quality in consciousness (in the purely sensational, non-categorial senses of ‘existence’ and ‘quality’). In this context, the crucial point is that even if there is a subjective condition for the quality of a sensation, it is not of itself sufficient for a sensation of that quality actually to exist. If this is granted, it follows counterfactually that the annulment of its subjective qualitative condition cannot of itself annul the existence the sensation would mark had it not been annulled. This suffices to prove the independence of the condition of its existence from the subjective condition on which sensational quality depends. Since none of the other subjective conditions of representation—i.e. (1)–(7) and their instantiations in diversely constituted creatures—are presupposed by sensation or its conditions, the existence marked by sensation is ipso facto subjectively unconditioned. Thus, each and every sensation we have immediately (analytically) brings with it the means of conceiving and affirming the existence of a subjectively unconditioned transcendental reality, which, as subjectively unconditioned, is at the same time radically unrepresentable, i.e. a mere thing in itself or transcendental object = X.

The Humean Challenge Met Kant wrote the Critique of Pure Reason to answer the question: how are synthetic a priori judgments possible? “Synthetic a priori judgment” is a formula expressly crafted to take into account Hume’s challenge to prove the psychological representability of necessary connections between distinct existents and its extension by Kant to include mathematical relations (Chapter 5): ‘synthetic’ corresponds to Hume’s connection of the (logically) distinct; ‘a priori’ indicates that the connection of the distinct is (extra-logically) necessary; and ‘judgment’ signifies the combination of these three elements in a single act of (non-linguistic) propositional thought—precisely what Hume concluded to be impossible. But because Kant regarded mathematical cognitions as indubitable instances of synthetic a priori judgments, his question was not whether such judgments are possible but rather how.

The I of Nature 299 Convinced that neither metaphysics, mathematics, science, nor logic could answer it, Kant looked instead to egocentric transcendental psychologism. It is now clear what that answer was: 1. Pure concepts of the understanding contribute necessity of the kind requisite for synthetic a priori judgments, for although no different in content from logical forms, their necessity is extra-logical. 2. Transcendental synthesis initially incorporates this extra-logical necessity into their connection component in a purely intellectual manner by determining propositions composed of AUA-concepts derived from the manifolds of formal intuition conformably to the pure concepts of the understanding (synthesis intellectualis). 3. But since the only difference between these AUA-concepts and the manifolds from which they are derived is the contentless universal self-consciousness ‘I think,’ there is nothing to prevent the determinations of transcendental synthesis from carrying over to the contents of the purely aesthetic sensible syntheses responsible for producing formal intuitions and the infinitely iterable manifolds of appearances comprehended within them (synthesis speciosa). 4. The necessity of pure concepts is then extended from the formal side of intuition to its material side via transcendental schematism. Insofar as the matter of appearances exhibits sensations and sensations indicate existences, transcendental schematism thus brings the existence component of Hume’s challenge into play. 5. Principles of pure understanding are transcendental judgments that subsume empirical syntheses of the existences indicated by sensations under transcendental schemata. As such, they combine the extralogical necessity thought in pure concepts of the understanding, the two-stages of transcendental synthesis, and apprehended, associated, and conceptually recognized existences together in a single act of non-linguistic propositional thought, and so prove the purely psychological representability of precisely what Hume held to be psychologically unrepresentable. Q.E.D. 6. In addition, what Hume, using identical reasoning, could have claimed is impossible in the case of mathematics, Kant showed to be representable insofar as transcendental synthesis provides the objectively (sensation-independently) differentiated, determinate formal succession, juxtaposition, and permanence necessary and sufficient for synthetic a priori mathematical judgments. Besides showing how synthetic a priori judgments are possible, the Critique also provides epistemic warrant for the transcendental kind (mathematical synthetic a priori judgments having no need of philosophy to establish their validity). For if, as Kant claimed, ordinary experience, science, and even pure mathematics are possible only insofar as

300  Nature in Mind: Through Kant’s I’s they incorporate transcendentally synthesized propositional networks, then transcendental synthetic a priori judgments cannot have less validity than the propositions of mathematics or science. Nor is the case for their validity entirely parasitic: transcendental synthetic a priori judgments would have their own independent validity if, as Kant claimed of the principles of pure understanding, they are both constitutive of nature itself and indispensable to transforming subjective experiences into objective ones. Whether these claims are true is for each to judge. For Kant, however, it all came down to the certainty of his philosophy’s supreme principle, cogito ergo sum: since everything asserted in transcendental psychologism is directly derivable from it, if its truth is certain, then the truth of the rest must be conceded as well.

Notes 1. See CPR A143/B182 and AA 18 §6314, both cited later. 2. CPR A156–7/B195–6. See also B146–7 and A239/B298. 3. CPR A112. 4. “Beasts lacking understanding do indeed have something similar to what we term representations (because it agrees in its effects with what in human beings are representations), though [it] may perhaps be entirely different— but no cognition of things” (Anthropology, p. 397). 5. CPR A20/B34. 6. CPR B128 and A126–7. 7. See, for example, CPR A114, A127, B163–4, and PFM 320. Kant’s comparison with Copernicus is at Bxvi–xviii. 8. If it is objected that thoughts are instances of consciousness and so indicate existence in their own right, independently of sensations—e.g.  cogito ergo sum—Kant’s reply is that they can do so concretely, and so in a manner capable of being objectified via transcendental synthesis, only in conjunction with internal sensations. In the absence of the latter, they can only be represented as expressions solely of the “spontaneity” of the mind, which is not a source of representable content at all, and so leaves us entirely incapable of singling out any one existence from any other, either subjectively (which requires sensation) or objectively (which requires contentful inputs from pure sensibility and their determination conformably to pure concepts of the understanding): see CPR B157–8n and KAIM 120. 9. Readers sometimes equate the matter of appearances with sensation, but this is mistaken. In those texts where Kant is most precise, it is clear that he took the matter of appearances to correspond to sensation rather than being sensation itself. Nor could it be otherwise: insofar as appearances are non-sensational products of productive imagination, their matter must be understood likewise. See KAIM part II, especially chapter 3-B. 10. CPR A143/B182. Also: “That (in representation) which is related to the object of the senses in itself is sensation” (AA 18 §6314 [1790–1]). The distinction between physical and transcendental realities should become clear as the chapter progresses. Here it suffices to remark that Kant regarded them as one and the same reality/existence, confining their difference to the kind of objectivity each has: the objectivity of physical realities is constituted by the categories making them, in the final analysis, consciousness-dependent, whereas the objectivity of transcendental objects (things in themselves) is

The I of Nature 301 completely beyond our comprehension and so can be understood only negatively as subjectively unconditioned and so consciousness-independent. 11. The textual case for distinguishing a two-stage transcendental synthesis that starts with the formal side of appearances and, in the guise of transcendental schematism, proceeds to the material side is presented in the introduction and section A  of KAIM chapter  15 (see also chapter  16-B). Most of the differences between transcendental schematism and transcendental synthesis derive from the purely aesthetic representations of possibility and existence contributed by the matter of appearances mentioned earlier. See KAIM chapter 16-C for a schema-by-schema consideration of these concepts. 12. See note §8 earlier. 13. CPR B127, A126–7 and AA 23 E LI (at A125–6). See also CPR A114 and B163–5. 14. “To be sure, empirical laws can never as such derive their origin from the pure understanding, no more than the immeasurable manifoldness of appearances can be fully conceived from the pure form of sensible intuition. But all empirical laws are only particular determinations of the pure laws of the understanding under which, and according to whose norm, those laws are possible in the first place, and appearances assume lawful form just as they all must nevertheless always be in conformity with conditions of the pure form of sensibility irrespective of the distinctness of their empirical form” (CPR A127–8). 15. Anthropology, p. 135. See also CPR B414–5n and PFM 306–7. 16. CPR A121. See KAIM chapter 13-B. 17. CPR A103–4. There is no necessity that these darkly perceived steps be individually discernible and capable of being accessed by attention. 18. CPR Bxiv. 19. See CPR A112–3 and A121–2, discussed in KAIM chapters 13-D and 17-E. As for Berkeley and Hume, association for Kant requires consciousness of the transition to and from associates as well as of the associates themselves. Non-conscious associative processes are not here at issue. 20. CPR B291, A183/B226, and A381. 21. CPR A107. 22. CPR A107 and A364. 23. AA 18 §5871 (early 1780s, possibly late 1770s). For the same reason, Hume regarded consciousness as confined to the present instant: see chapter 8. 24. AA 18 §5871. Note that the representation of the persistent is as instantaneous as any and all other representations, for “the representation of something permanent . . . is not the same as a permanent representation” (CPR Bxli n.). 25. CPR A185/B229. 26. CPR A183/B226. 27. Since no representation can be used to conceive things in themselves, and since “in experience no question is ever asked in regard to them” (CPR A30/ B45), there are neither means nor occasion to inquire into the differences and commonalities among transcendental realities, including the transcendental subject and transcendental object (which may or may not be one and the same: see A358–9 and A383–91, discussed in KAIM ch 8-F). 28. Even Lee Smolin’s evolving succession of universes, of which our Big Bang universe (including its laws and constants) is only a temporary phase, is not represented as existing by any inner necessity. 29. In addition to Smolin, many others, including Roger Penrose and Paul Steinhardt, postulate eras that preceded and will succeed the Big Bang universe. 30. See chapter 9 note §2.

302  Nature in Mind: Through Kant’s I’s 31. As such, the community principle is an essential requisite for the all-­ encompassing individuality of the objective synthetic unity of apperception to be realized physically, and so too for the physical realization of the unrestricted universality of the analytic unity of the ‘I think’—i.e. the I of nature in beings constituted like us (i.e. with our set of pure concepts). The derivation of the pure concept of community from the logical form of disjunctive judgment is analyzed in KAIM chapters 10-D and 11-C, where the associated modal concept of necessity-contingency is also considered. Their roles in transcendental synthesis and schematism are examined in KAIM chapters 15-D and 16-C. The community-derived principle of pure understanding, its significance for objective unity of apperception, and its bifurcation at CPR A214–5/B261–2 into commercium (= the Universe) and communio (= the universe) are discussed in KAIM chapters  17-I and 18-B, while the necessity-contingency derived principle—the third postulate of empirical thought—is analyzed in chapter 18-D. 32. MFPNS 470. 33. CPR A114 and A216/B263. Also: “Unity of nature in time and space and unity of the experience possible for us are the same” (CJ AA 20 209). See PFM 318–20. 34. CPR A126–7 and B128. Also: “the laws of nature really have their origin in the understanding and are just as little to be encountered outside it than space and time are [to be encountered outside of sensibility]” (AA 23 E LI at A125–6). 35. CPR A127–8. To keep this book as straightforward and compact as possible, I will not be discussing any of the principles of pure understanding individually (those interested in my take should see KAIM part V). Suffice it to say that the nonbifurcating, nonconverging linear paths mandated by transcendental synthesis apply only at the bottommost level of substantial final subjects and their accident predicates. No such restrictions hold at higher levels. For example, the mathematical constructions that yield the splitting causal histories distinctive of the many worlds interpretation of quantum mechanics are perfectly compatible with a foundation in transcendental synthesis. 36. CPR B163–5. See also PFM 295, 318, and AA 17 §5406. 37. CPR A114. Since the principles of pure understanding concern transcendental schemata, which relate to the matter and not just the form of appearances, they serve to confer physical reality on the objective I, which, as such, constitutes nature materially as well as formally (regarding this last distinction see the texts cited in the preceding note). Thus, the formal representations of sensibility and understanding in beings constituted like ourselves yield principles of substantial permanence, causal succession, and interactive simultaneity that together “exhibit the unity of nature in the interconnection of all appearances under certain exponents which express nothing else than the relationship of time—insofar as all existence is comprehended within it—to the unity of apperception” (A216/B263). See also A177–8/B220. 38. CPR B422. See also A346/B404, A362–3, and A401–2. 39. Kant did allow for the problematic concept of an intuitive understanding (intellectual intuition). Having no need for sensibility to intuit its objects, it would be free of the sensible condition for concepts—original synthetic unity of apperception—and consequently have no need for the (AUA-) concepts it makes possible. This of course would make it radically incommensurable with our own sensibly-conditioned variety, starting with the fact that, in the absence of that original synthetic unity, it could have no ‘I think’. More importantly, to say that intellect is sufficient by itself to intuit objects is just

The I of Nature 303 to say that those objects would be given—exist, be created—in the very act of being understood. Since this would make intuitive understanding creative in a sense no sensibly conditioned understanding could ever be, it is no surprise that Kant held that any such understanding would count as divine—or that he regarded intellectual intuition not only as unaffirmable but as radically incomprehensible as well. 0. CPR A143/B182. 4

Conclusion After Kant

A philosophical refutation, Hegel said, “must penetrate the opponent’s stronghold and meet him on his own ground; no advantage is gained by attacking him somewhere else and defeating him where he is not.”1 Kant’s stronghold was egocentric transcendental psychologism. While taking its start from Descartes’s cogito ergo sum, Kant’s a priori psychologistic analysis of the ‘I think’ as universal self-consciousness was entirely his own. No less original is the theorizing founded on this “supreme principle,” including his psychologistic accounts of the sensible conditions of the possibility of universal self-consciousness, how this pure self-consciousness serves to make possible the logical forms requisite for propositional thought, how from these forms pure concepts of the understanding may be derived, and, finally, how, with transcendental synthesis in conformity to these concepts, the ‘I’ of the ‘I think’ raises itself to an ‘I’ of nature. The transcendental psychologism expounded in the Transcendental Aesthetic and Analytic of the Critique of Pure Reason as an alternative to both platonism and empiricism/conventionalism is unique to Kant. And he used it as a basis for building a comprehensive philosophical system, starting with a critique of metaphysical speculation that transgresses the ‘I’-demarcated bounds of a priori psychologism (the Transcendental Dialectic of the Critique of Pure Reason), proceeding to elaborate from this ‘I’ a wholly new idea of freedom to re-ground moral philosophy (the Critique of Practical Reason), and concluding with the analysis of judgments that are founded on transcendental principles of a priori psychologism but directly involve neither the categories nor the idea of freedom (the analyses of the beautiful, the sublime, and teleology in the Critique of Judgment).2 To attempt to refute any of this without superseding or undermining Kant’s transcendental psychologism in the process is, accordingly, a pointless exercise precisely because it means “defeating him where he is not.”

Platonistic Refutations of Kant Since there are only two genuine alternatives to Kantian transcendental psychologism, platonism and empiricism/conventionalism, it should

Conclusion 305 come as no surprise that attacks on Kant’s stronghold have come from one or the other of these quarters. Platonistic refutations focus on showing that the ‘I think,’ while important, is by no means the supreme philosophical principle Kant took it to be but is instead, in the final analysis, just one concept among others. The first wave of platonistic refutations of Kant coalesced in a group of philosophers known as German Idealists and culminated in Hegel’s analysis of the ‘I think’ in his Science of Logic (1812–16). The Science of Logic unfolds as a succession of concepts starting from simple being and ending with the Absolute Idea. The Absolute Idea is the ultimate result of innumerable dialectical transformations of the concept of being through quantitative and qualitative categories, segueing into relational and modal categories like substance, causality, and necessity, until arriving at the concept of the concept itself. It was in regard to the latter that Hegel lauded Kant’s analysis of the ‘I think’ for breaking wholly new philosophical ground.3 Nevertheless, Hegel’s treatment of the topic is at the same time a refutation of Kant’s thesis that the analysis of the ‘I think’ is the culmination of the analysis of the concept. For while Hegel accepted that the existence (sum) of the concept as an ‘I’ (cogito) is a crucial stage in its logic, it is by no means the last. From the reciprocal synthetic and analytic unities identified by Kant, the concept must then be reconceived as judgment, next as inference, followed by a sequence of objective guises—mechanical, chemical, and teleological— that culminate in the concept as an idea in the full platonist sense, first as the idea of Life, then that of the True, the Good, and finally the Absolute Idea of Divinity. I do not pretend to understand Hegel’s logic well enough to assess it. But this much can be said of its relation to Kant’s transcendental psychologism: it purports to show that the a priori synthetic and analytic unities of the manifold into which Kant resolved the logic of the concept have no essential dependence on sensations, intuitions, self-consciousness, or anything else ineluctably psychological in nature. If true, this would genuinely refute Kant because it would set the logic of the concept free from its psychological moorings and convict Kant’s psychologization of logic of being an instance of what has come to be known as the “psychologistic fallacy”: the misbegotten endeavor to explicate the non-psychological psychologically. Could Kant have responded with a counter-refutation? The issue centers on the manifold to which both unities, analytic and synthetic, originally relate. Kant and Hegel agreed in denying that the manifold consists of sensations and affirming that it is purely relational in character. Yet, whereas Kant held that it is the manifold of sensible appearances intuited by means of purely aesthetic formal intuitions of sensibility, Hegel deemed it a manifold immanent to the concept itself, independently of sensibility and the psychological generally. Against Hegel’s platonism, Kant could have argued that Hegel’s logic is a nonstarter because its

306  Conclusion various manifolds would not be consciously representable were it not for the manifold of purely formal individuals that result from transcendental synthesis of formal intuitions of sensibility such as pure space and time. Hegelian logic can therefore no more take place outside of pure consciousness than Aristotelian logic, mathematics, or post-Fregean mathematical logic can, and, like them, not only can but must be explained psychologistically. If so, then there is no fallacy of psychologism here. On the contrary, the real fallacy would be Hegel’s platonistic endeavor to explicate the inherently psychological non-psychologically. I do not know how to choose between Hegel’s refutation and its Kantian counter-refutation. But one point in Kant’s favor is surely that whereas platonism, Hegelian or any other, seems to render our minds evolutionarily inexplicable, not only is Kant’s psychologism consistent with evolution, it actually spotlights suitably primitive forms of empirical consciousness that would gain adaptational advantages from a priori consciousness. Most basically, formal intuitions can easily be conceived to be of use to minds grappling with the challenge of combining external sensations into a single, unified external sense capable of providing immediate access to sense-divide transcending objects. Of even greater adaptive benefit is the structure of cognition built on these intuitions, starting with customary association and recognitive AUA-concepts all the way up to science and mathematics, without at any point requiring evolutionarily inexplicable platonistic inputs. Indeed, from a Kantian perspective, the fact that Hegel’s Logic could be grasped by at least one member of our species already of itself suffices to show that it can contain nothing that does not arise from and conform to the same, naturally selected psychology to which everything else accessible to human understanding is subject. A less ornate platonistic refutation of Kant emerges from twentiethcentury analytic philosophy. Oddly, its most influential presentation, Peter Strawson’s The Bounds of Sense (1966), offers up anti-­psychologistic platonism as Kant’s actual position. If the present book has shown anything, it is that nothing could be further from the truth. Nevertheless, it tempts one to ask whether the Kant of analytic philosophers like Strawson can refute the determinedly psychologistic Kant presented here—“Kant” vs. Kant. It is a platonism because it not only excludes psychology from the explanation of transcendental concepts on pain of psychologistic fallacy, but excludes conventionalist explanation as well. Hence, in the same way that platonistic mathematicians view mathematics as eternal truth, entirely independent of psychology, society, and all things human, the Kant of analytic philosophers regards logical forms, the categories, space and time, and even the ‘I think’ itself as in effect platonically real archetypes, ectypes of which are innately hardwired into our minds. The original twist that distinguishes the platonism of analytic philosophy’s Kant from others is the idea that platonically real representations

Conclusion 307 constitute a single, coherent conceptual framework in and through which alone objective representation is possible. While its inclusion of a thinking, knowing subject is not essential (much less its “supreme principle” as the historical Kant regarded it), it serves to provide its users with a way of anchoring the conceptual framework of objectivity in the closely aligned epistemological framework of belief, evidence, proof, testimony, credibility, confidence levels, and the like. As such, analytic philosophy’s extreme makeover of Kant into a platonist purports to yield a refutation of radical skepticism, be it Descartes’s, Hume’s, or any other. For while it is perfectly fine to doubt this or that particular knowledge claim, or even a particular body of knowledge claims (e.g. astrology), how can one presume to doubt the very conceptual frameworks that make objectivity and knowledge possible in the first place? Since doubting can no more take place outside them than knowing or believing can, it is quite literally nonsensical to pretend to be able to doubt them. Yet, as many have pointed out, this supposedly Kantian platonistic refutation of skepticism is easily countered. For even if conceptual frameworks that exist independently of both psychology and vernacular conventions are indispensable to our thought of objects, necessities of thought are not the same as necessities of things. Our conceptual frameworks, be they eternal or conventional, can never, in and of themselves, constitute the least ground for thinking that the objects they are used to represent are anything like the way we represent them as being, and it does not make the slightest difference if we cannot help representing them that way and are constitutionally incapable of representing them otherwise. So much for the refutation of skepticism that Strawson and others claim to find in the platonism they ascribe to Kant.4 There remains the question of whether there is a Kantian counterrefutation to the platonistic refutation of his psychologism. This platonism is held to render the entire edifice of transcendental psychologism redundant: space and time have no need of formal intuitions or productive imagination; logical universality has no need of universal selfconsciousness; logical form is completely independent of modes for combining AUA-concepts in a single act of thought; categories have no need for pure concepts of the understanding derived from such modes; mathematics has no need for transcendental synthesis of the manifold of formal intuitions conformably to pure concepts of the understanding; and nature, along with empirical cognition, scientific no less than ordinary, is completely independent of transcendental synthesis of the matter of appearances conformably to the schematized versions of pure concepts of the understanding and their subsumption of empirical psychology in transcendental judgments.5 But, again, one of the features that makes this edifice special is its presentation of intelligent mind as something that could conceivably have evolved from minds confined to sensation, emotion, and simpler forms of association, with no capacity for

308  Conclusion propositional representation at all. Can the same be said of analytic philosophers’ wholly non-psychological, non-conventional platonistic conceptual frameworks?6 Stripped of all foundation in the psyche as well as all dependence on linguistic convention and its human-societal context, how, short of a hugely improbable saltation, could our simian ancestors have evolved minds equipped to represent and deploy conceptual frameworks that have no antecedents in their evolutionarily determined psyches? Indeed, how, deracinated from psychology and social convention, could their contents be about anything at all, i.e. mean, refer, or otherwise do the work of representations for us? And how are we, creatures crafted by natural selection, to imagine what thoughts are in the absence of conscious thinkers and conventional practices, or why they should be called ‘thoughts’ at all?

Conventionalist Refutations of Kant None of the foregoing considerations mean that analytic philosophers— or Hegel—are wrong. But they do show that platonistic refutations of Kant leave much to be desired qua refutations. There is, however, another no less anti-psychologistic way of refuting Kant that avoids the pitfalls of platonism: conventionalism. The basic idea is that it is not the individual isolated psyche but collective human society that bestows meaning on categories, logical forms, space and time, and the ‘I,’ so that any purely psychologistic explication of these representations is ipso facto fallacious since there can be no psychological explanation of the irremeably sociological. Though there are many distinguished philosophies in this vein, the two that seem to me most compelling as sources for refutations of Kant in the fastness of his psychological stronghold are Martin Heidegger’s version of existentialism in Being and Time (1927) and Ludwig Wittgenstein’s conventionalism of humanly instituted language games existing within the weave of uniquely human forms of life in the Philosophical Investigations (left unfinished at his death in 1951). Heidegger contrasted categories with existentialia, the one employed in relation to things present at hand, the other things ready to hand. They differ not in the things to which they are applied but in how they apply to them, the former for signifying how things factually are and the latter how they are for us, i.e. in relation to human purposes and lifeways. Existentialia get their meaning from Dasein, which for present purposes can be understood as existence in the context of human life in its totality, not least its intricate social fabric, but most particularly time as measured by human mortality. From a Kantian perspective, categories have priority because they originate a priori in the individual isolated psyche, making them completely independent of the human and every other empirical context. But from Heidegger’s standpoint, categories, like concepts, propositions, and representations of the present at hand

Conclusion 309 generally (including science and mathematics), have to be understood, in the first instance, instrumentally, as ready-to-hand implements for meeting socially grounded human needs, and so through existentialia on the foundation of Dasein. If, as Heidegger held, existentialia have primacy over categories, then any attempt to comprehend the former through the latter becomes just another variant of psychologism: explaining the intrinsically human social context in which our lives are led in terms of the individual psyche considered a priori in isolation from that context, which can have no other result than nonsense. The Heideggerian refutation of Kantian psychologism is therefore this: we can entertain no hope of understanding anything going on in consciousness if we sever it from the broader context of human purposes and lifeways, i.e. from Dasein. The ‘I’ is not only a case in point but Heidegger’s prime exhibit against Kant. Heidegger interprets Kant as defining the ‘I’ as a subject set over against its objects, whose own being therefore cannot be understood through any of the categories proper to those objects, and, given the epistemological primacy of the categories, cannot be understood positively in any way at all.7 But this is hard to credit. Isn’t Heidegger leaving out of account the richly detailed positive story of the ‘I think’ in Kant’s Critique of Pure Reason? The logical I that grounds the categories, the aesthetic I that makes understanding itself possible, the objective I on which categorially grounded knowledge depends, and the transcendental I constitutive of nature itself—i.e. everything discussed in Part II—together result in as systematic and unified a positive conception of the ‘I’ as one could reasonably ask for—sufficient certainly for the ‘I’ to be carried over into moral and other kinds of philosophy in Kant’s subsequent writings. Once this is recognized, it is easy to see that his doctrine of the categories needs to be accepted for the chapter in transcendental psychologism that it is. Far from being social implements whose meaning is their accepted use, the categories (and space and time) need to be understood as products of the human psyche that arise naturally from the same phylogeny that human sensations, desires, and species-wide lifeways do. Their true foundation is thus not the human present, much less Heidegger’s ­de-psychologized version of it, but the human psyche’s evolutionary past. If Heidegger discounted the results of Kant’s psychologistic focus, it is probably because he, like analytic philosophers and others of his era, interpreted Kant as a steadfast opponent of psychologism rather than its foremost champion. Doing so, however, leads to a serious underestimation of the explanatory power of transcendental philosophy. For example, unlike Hegel, Heidegger saw in Kant’s logical ‘I’ only a ghost-like external assembler of preexisting concepts into propositions,8 seeming to overlook its positive logical contribution of the universality through which both the logical universe of concepts and the logical forms from which categories derive their content can alone be understood. So too the ‘I’ of nature: far from being isolated from the world, this ‘I’ quite

310  Conclusion literally comprehensively encompasses it, just as that same world comprehensively encompasses the empirical ‘I’ (the mortal soul). Translated into Heideggerese, the world’s being, for Kant, is being-in-the-I, the soul’s being-in-the-world, and the being of the I (sum) as transcendental subject  =  X, like that of the ‘in itself’ generally, is not even a posable question. That Heidegger was thus led to underestimate the explanatory potency of Kant’s transcendental psychologistic stronghold best explains why he saw no real difference between Kant’s notion of the ‘I’ and that of a platonist like Descartes, and so extended to the former his criticism of the latter’s dualism for both detaching the I from the external world and setting them on the same ontological level—both equally present at hand, each severed from the other in a duality of metaphysical kinds, immaterial vs. material. Since Kant’s a priori psychologism focuses exclusively on the isolated, individual psyche, viewed purely as a product of nature (understood today in terms of physiology, ecology, ontogeny, phylogeny, abiogenesis, chemistry, and physics), it is completely neutral in all matters premised on our being-there (Da-sein) as mortal conscious creatures thrown into the world existing along with everything else, particularly our fellow humans. If Kant would have directly opposed Heidegger on any point, it would most likely have been the latter’s notion that the ‘I think’ can and should be explained phenomenologically. For Kant, that is no more possible than it is to explicate Dasein transcendentally (i.e. psychologistically). This is because the sphere of Dasein can be deemed a priori only in a relative sense inasmuch as it presupposes an acquaintance with the full panoply of human life, and death, which is only possible empirically— and even then is closed to any creature not endowed with all that a uniquely Homo sapiens phylogeny brings with it, culture included. As such, Dasein is not the sole possession of the phenomenological philosopher and may perhaps be better understood from the vantages of the natural scientist, historian, linguist, social scientist, humanist, novelist, artist, and, more generally, the wisest among us. By contrast, the ‘I think’ belongs exclusively to the province of the transcendental psychologistic philosopher. And if one concedes reality to the ‘I think’—the sum that seems inseparable from the cogito—then transcendental philosophy may be accounted as much a science of its corner of reality as mathematics is the science of quantity and the natural sciences are sciences of physical reality. The refutation of Kant that has always seemed to me the most original and compelling derives from Wittgenstein’s conventionalist critique of logic in the Philosophical Investigations. Wittgenstein questioned whether propositional representation can ever be disentangled from language; and after presenting a compelling philosophical case that it cannot, proceeded to assimilate propositional logic to the logic of language. What separates the later Wittgenstein from his previous incarnation in

Conclusion 311 the Tractatus Logico-Philosophicus is that ‘language’ in the Philosophical Investigations signifies not a platonic ideal of language but simply ordinary vernacular speech and writing, rooted in established conventions, and ineluctably bound up with culture, society, and uniquely human lifeways. Wittgenstein never attempted to refute Kantian transcendental psychologism directly, so one can only surmise how he might have done so. My best guess was presented in this book’s second part. In a nutshell, it is the charge that the propositional thought that features in Kantian transcendental psychologism is no different than wordless thought generally, and so is vulnerable to Wittgenstein’s proofs that a private language is a radical impossibility. In reply, I argued that the role of propositional thought in transcendental psychologism is nothing like that of a language, and indeed is so far distant from it that neither Kant himself nor anyone conversant with his psychologism could be under any illusion to the contrary. In the first place, far from being mental counterparts to words, Kantian AUA-concepts do nothing more than confer on the deliverances of sensibility and empirical psychology the AUA-universality requisite to combine them in propositions according to the logical forms of judgment. And that is all: their conversion does not give them any further logical, mathematical-logical, or grammatical character, does not imply their incorporation into private mental language games, and decidedly does not weave them into some kind of holistic epistemic structure à la W. V. Quine’s “web of belief.” AUA-concepts and propositions formed from them are simply too obviously psychological to be confounded with even the most minimalist and primitive of languages, private or otherwise. Second, Kant’s derivation of the categories from the logical forms of judgment is so simple and straightforward that he cannot possibly be suspected of surreptitiously smuggling in content from language, platonic heaven, or anywhere else. Being in and of themselves no more than pure concepts of the understanding, the categories represent nothing more than restrictions on the logical freedom implicit in each such form. Nor can a similar suspicion be entertained in connection with the purely aesthetic space and time of formal intuition. There can be no concern that their undifferentiated, indeterminate manifolds correspond, or were intended by Kant to correspond, to any linguistic notion of space or time, be it in ordinary vernacular, mathematical, or scientific usage. Indeed, so entirely primitive and psychological are they that Kant supposed them to apply even at the level of apprehension, a synthesis so elementary that no one before him—or after for that matter—ever suspected it might involve imagination in addition to sense (synopsis). Clearly, such a synthesis, one that ostensibly precedes and makes possible all reproduction of spatial and temporal appearances in memory and association, and, a fortiori, all recognition in AUA-concepts as well, cannot be thought to be anything other than purely psychological, and certainly not linguistic.

312  Conclusion Finally, transcendental syntheses of the form and matter of sensible appearances are supposed by Kant to yield objective, physically real transcendental space and time. Although it is a propositional structure, Kant gave us absolutely no reason to view it as any kind of language and every reason to identify it with the space and time of ordinary and scientific experience. At its core, it is nothing more or less than the psychological structure Kant deemed necessary for the cognition of sense-divide transcending objects through external and internal sensations. It is true that his psychologism identifies it with nature itself. But since there is nothing to indicate that Kant regarded nature as a language in even the most attenuated sense, if nature is indeed the propositional structure constituted by transcendental synthesis, then that should be seen as proof that that structure too is not linguistic but 100% psychological. But even if it is conceded that Kant did not transgress the boundary separating non-linguistic propositional representation from private language that so many others have, this does not yet tell us what position he might have adopted vis à vis Wittgensteinian conventionalism. Here it may be helpful to think back to a time before anything remotely like modern language existed. One would perhaps not have to go back even 100,000 years, but just to be on the safe side let’s imagine people who lived more than a million years ago. Are we to suppose that because they lacked language they ipso facto were wholly incapable of propositional thought? Indeed, are we to suppose that if we too had been raised without language, we would be incapable of propositional thought? Imagine that instead of human parents, we were raised by futuristic androids that never used language but otherwise trained us to function similarly to how we actually do (with all the necessary safeguards to prevent us from harming ourselves): what buttons to press to operate the lights, the dishwasher, the car, and the music system (all instrumental music and wordless chant); how to dress, adorn, and take care of our bodies; how to participate in sports, games, and other quotidian peer activities; how to play musical instruments; painting, carving, weaving, sewing; marching and dancing; pratfalls and other forms of non-verbal joking; etc. Is it conceivable that our minds would be complete blanks when it came to everything we normally use linguistic propositions to represent? Or would our non-linguistic thoughts and actions go on much as they currently do with language? This is not the same as asking would they still be linguistic thoughts and actions, only private rather than public. As I understand him, Kant could agree that thoughts and actions would not qualify as linguistic in even the most attenuated sense in the absence of the social infrastructure requisite to support linguistic conventions. Instead, the relevant question is whether psychology can provide for logically structured propositional representation even when that infrastructure is lacking. If so, we would

Conclusion 313 still be capable of forming propositional representations employable in judging and reasoning. If not, then our minds would be reduced to the representational level of creatures altogether incapable of giving logical form to their representations, and consequently be wholly dependent on association and similar such blind (i.e. non-recognitive) psychological operations. Were it not for Kant, platonism would be the only way logically structured thought could still be presumed possible in the absence of vernacular language. But platonism has no remotely convincing evolutionary tale to tell of how logically structured propositional representation can have sprung, like Athena from the head of Zeus, fully developed from the minds/brains of our pre-logical ancestors. The story told by the Wittgensteinian conventionalist is, by contrast, fully consonant with evolutionary theory; and if Kant had not provided a further alternative, no one who abjures mystical metaphysics can avoid preferring conventionalism to platonism. For, however implausible aspects of Kantian transcendental psychologism may strike one, there is no disputing that it represents a way of understanding non-linguistic propositional thought compatibly with evolution. Indeed, it improves on conventionalism by enabling us to understand the emergence of propositional discourse as a gradual, step-by-step process from non-linguistic yet still propositional representation (including the objective sort involving transcendental judgment), rather than having to regard it as a saltation from non-propositional representation devoid of all logical structure. Thus, from the standpoint of transcendental psychologism, the original inventors of language would already have been skilled practitioners of propositional representation, in both thought and action. And rather than an entirely new, uniquely cultural species of representation that emerged ex nihilo atop what, in a non-linguistic state of nature, is a wholly non-logical, non-propositional repertoire of conscious representation, propositional discourse would be simply an additional set of linguistic techniques for communicating private mental representations alongside those used for communicating non-linguistic, non-propositional mentation such as emotion, conviction, attitude, desire, and volition. One way to understand this is by contrasting linguistic expression of the non-linguistic with translation from one language into another. The Kantian can agree with the conventionalist that the English proposition “the sky is blue” is no more a translation of a proposition in the private language of mentalese than “ouch!” is the translation of my cry of pain when I stub my toe, or my intonation in saying “Absolutely!” is the translation of my conviction. All alike are instead best regarded as linguistic expressions of non-linguistic mental representations, with the “the sky is blue” expressing a non-linguistic proposition, “ouch!” a non-linguistic affect, and the tone in which “Absolutely!” is pronounced non-linguistic belief (e.g. Humean vivacity affect). The only cases where there might be

314  Conclusion at least an attenuated sense of translation (psycho-linguistic overlap) are the twelve logical forms of judgment that Kant deemed integral to both linguistic and non-linguistic propositional representation. Otherwise, he could have fully endorsed the central conventionalist thesis that linguistic meaning and grammar derive entirely from publicly established norms of usage and in no measure from private mental representations. Are there reasons for preferring transcendental psychologism to conventionalism? For one thing, it frees propositional representation from any essential relation to human sociality and lifeways without requiring recourse to platonism. All creatures may be deemed capable of such representation insofar as their psychology equips them to form the representation ‘I think,’ to utilize it to produce AUA-concepts, and to combine AUA-concepts in a single act of thought by means of logical forms. Some may in addition have something akin to human languages whereby they communicate with their fellows. But others might be completely asocial and use propositional representation solely to the end of cognizing sense-divide transcending objects in accordance with principles of pure understanding. But however it is used, propositional representation is one thing, propositional discourse quite another. So, even if it is conceded that if lions could speak we could not understand them, if their psyches equipped them to form the representation ‘I think,’ then, on Kant’s theory, they would be capable of non-linguistic propositional representation on all fours with ours. This point is important because if conventionalism has an Achilles’ heel, it is its difficulty in providing a persuasive account of objective representation. Conventions are artifices that enable humans to coordinate their actions in ways they otherwise could not, and so are essentially bound up with all the ways humans have evolved to interact with conspecifics. If propositional representation is supposed to be purely linguistic and language is inherently conventional, then anything that is impossible apart from propositional representation is ipso facto inseparable from human sociality and lifeways. Since the propositions of mathematics and natural science seem impossible to represent by anyone lacking a mastery of spoken and written language, the implication is that they too must be deemed intrinsically bound up with the social infrastructure requisite for convention. So, how can they be “objective,” “true,” “supported by evidence,” or “proven” in a sense any less conventional and ineluctably bound up with human sociality and lifeways than the application of such terms to the propositions that my king is in check, that you were impolite to your guests, or that the Dodgers blew it again? The conventionalist seems to leave us no alternative but to set mathematics and natural science on a par, both epistemologically and ontologically, with such other human language games as making up stories, play acting, singing catches, joking, asking, thanking, cursing, sermonizing, and translating from one language into another. Moreover, given the amazing mathematizability

Conclusion 315 of nature and all things in it, doesn’t conventionalism commit us to the absurdity that nature itself is, at least in the mathematical propositional form taken by its laws, no less bound up with human convention than versifying or religious ceremonial? The apparent inability of conventionalism to do justice to the objectivity of mathematics and natural science has led many to look for an alternative. Although the one generally favored is platonism, this may simply be because it is so seldom appreciated that there exists another with none of the drawbacks of platonism: transcendental psychologism. Without denying that the propositions of mathematics and natural science are inaccessible to humans except through language, the overlap between language and transcendental psychologism instanced by the logical forms of judgments suffices to ground language generally and mathematics and science in particular in transcendental syntheses—non-linguistic propositional representations of manifolds of fully individuated objects that are as completely independent of convention and human sociality generally as they are independent of sensations. This makes objective discourse a hybrid of conventional techniques and transcendental, purely psychological representations that counts as “objective” insofar as it allows of being construed as relating to these manifolds exactly as non-linguistic propositions do. Since this is truest in the case of the propositions of mathematics and natural science, they qualify as “objective,” “true,” “provable,” etc. to a degree no other discourse can approach—so much so that this difference in degree is, for all cognitive intents and purposes, a difference in kind. The ability of linguistic propositions to blend with non-linguistic is, in its way, no less amazing than the mathematizability of nature. But isn’t this exactly what we should expect if language was originally crafted by creatures already fully conversant in the use of non-linguistic propositional representation, transcendental judgments included?

Kant’s Transcendental Legacy to the Sciences of Mind Psychological theories of the nature and workings of propositional thought are thin on the ground. Psychologists still do not know whether to treat propositional thought as essentially linguistic, and so as a cultural artifact belonging more to social science than psychology, or whether at least some propositional thought can be understood purely psychologically, in terms of the individual isolated mind/brain, without needing to draw on social science and social psychology at all. Wittgenstein was the greatest philosophical champion of the first position, and psychologists like Michael Tomasello are well aware that it behooves all mind scientists to incorporate Wittgenstein into their conceptual landscape. Readers of this book now know that Kant was the leading exponent of the second position, and I hope to have convinced at least some contemporary mind theorists that Kant’s non-platonist, I-centered transcendental

316  Conclusion psychologism has an equal claim to be included in the conceptual landscape of the mind sciences. What are the most important mind-science relevant takeaways from Kant’s transcendental psychologism? If you are looking for an example of a theory of mentation that coheres with evolutionary science and cultural development that yet is powerful enough to take into account propositional representation, you need look no further than Kant. Whether or not the theory is judged to succeed, it contains abundant food for thought. For example, if we use its lens to view the brain as housing logic itself, we discover a whole new way, over and above sensational consciousness and other non-logically structured varieties of conscious representation, in which that remarkable organ differs from the liver, spleen, and the rest. More particularly, in showing how, independently of language and everything unique to humans, our brains can be organs of intelligent consciousness, transcendental psychologism at the same time shows how consciousness may be intrinsic to intelligence. And while this would not rule out AI, it does preclude AI without an AC (artificial consciousness) that includes a no less artificially generated cogito. Equally importantly, Kant obliges scientists and philosophers to consider whether understanding the psychology of space and time requires that they trace them back to something as subjective, undifferentiated, and indeterminate as pure formal intuitions that earn their evolutionary keep by enabling creatures to represent physically real sense-divide transcending spatiotemporal objects. Indeed, Kant’s theory of pure sensibility offers us an opportunity to think about sensing and perceiving in hitherto neglected ways, and do so moreover in a manner that, with the overlay of transcendental synthesis, dovetails perfectly with ordinary and scientific cognitive experience of such objects. Above all, Kant’s transcendental psychologism provides unique insights into one of the greatest of all psychological mysteries: consciousness of objectivity. How do creatures snuffling about on the ground, swinging from branch to branch, darting about in seas, or otherwise preoccupied with their immediate sensory surroundings manage to so transcend them as to be able to represent objects independently of their sensationdefined perspectives? How is it possible ever to transcend one’s sensationally determined subjective standpoint to represent anything, oneself included, objectively? Since philosophers are nowadays no more likely than mind scientists to suppose that the key to objectivity lies in the subjectivity of the ‘I think,’ the psychology of objective representation tends to be regarded as ancillary or altogether irrelevant to the question of representational objectivity per se, so that anti-psychologistic platonism and conventionalism become the only options ever considered. But one has to look no further than Kant’s psychologism for a profound, fully elaborated account of objectivity as constituted entirely by the psyche.

Conclusion 317 Its upshot is that genuinely objective representation is possible only in creatures capable of the universal self-consciousness ‘I think’ and nonlinguistic propositional combination of AUA-concepts, so that, without them, even the most sophisticated conventional techniques for communicating with conspecifics or the most generous platonistic endowments could never yield standpoint-transcending objective representation. That, at any rate, is the Kantian psychologistic challenge to anyone, scientist or philosopher, who believes differently.

Notes 1. Science of Logic, translated by A.V. Miller, London: George Allen & Unwin Ltd., 1969, p. 581. 2. Kant’s critical philosophy also has adjuncts in the philosophy of natural science, morals and justice, religion, politics, and society. 3. Hegel deemed it “one of the profoundest and truest insights to be found in the Critique of Pure Reason that the unity which constitutes the essence of the concept is recognized as the original synthetic unity of apperception, as unity of the I think, or of self-consciousness” (Science of Logic, p. 584, translation slightly modified). 4. If it is objected that the psychologistic Kant fares no better because objects do not care how our psyches are constituted to represent them, the reply is that this is true of only things in themselves, not physical objects or nature, which are constituted by transcendental schemata and subject to the law of principles of pure understanding. Nor do things in themselves offer the least succor to the skeptic: being quite literally impossible to represent, and since “in experience no question is ever asked in regard to [them]” (A30/B45), it is a feature of Kant’s “Copernican revolution” in philosophy, not a bug, that they do not in any regard conform to our representations. 5. A perfect example of the kind of interpretation of Kant discussed in the introduction that ends up leaving most of the philosophical components he provided unutilized. 6. By contrast with Hegel, whose understanding of Kant was unsurpassed, analytic philosophers seldom if ever appreciate the significance of the fact that Kant conceived the ‘I think’ as a universal self-consciousness in and through which alone logic itself is possible (e.g. CPR B131 and B133–4n). But whether they were aware of it or not, it is the view that matters here, i.e. the question of whether psychological explications of the logical are psychologistic fallacy or, at least in Kant’s case, the antithesis of fallacious reasoning. 7. “When Kant uses the term ‘existence’ he has in mind the being-present-athand of consciousness just as much as the being-present-at-hand of things. . . . After the primordial phenomenon of being-in-the-world has been shattered, the isolated subject is all that remains, and this becomes the basis on which it gets joined together with a ‘world’ . . . [T]he construct of an isolated subject . . . remains indefinite in its being and is best characterized negatively as ‘un-Thing-like’ ” (Being and Time, H. 203–6, John Macquarrie & Edward Robinson, New York: Harper & Row Publishers, 1962). 8. “Kant tries to establish the phenomenal content of the ‘I’ as res cogitans [i.e. the Cartesian dualist notion of substantial thinking being]. If in doing so he calls this ‘I’ a ‘logical subject’, that does not mean that the ‘I’ in general is

318  Conclusion a concept obtained merely by way of logic. The ‘I’ is rather the subject of logical behavior, of binding together. . . . All binding together is an ‘I bind together.’ In any taking together or relating, the ‘I’ always underlies—the hypokeimenon. The subjectum is therefore ‘consciousness in itself’, not a representation but rather the ‘form’ of representation. . . . To define the ‘I’ ontologically as ‘subject’ means to regard it as something always present-athand. The being of the ‘I’ is understood as the reality of the res cogitans” (Being and Time, H. 367).

Glossary of Kantian Terms

These definitions apply only to the use of terms in this book, including terms used in the analysis of Kant’s philosophy, and are not intended to comprehensively capture their (often multiple) meanings in Kant’s oeuvre. a priori/a posteriori  if a representation involves the presence of sensations but no consciousness of particular sensational content, then it is a priori, or pure; if it includes consciousness of particular sensational content as well, as is required if sensations are to be distinguished from one another as regards quality, pattern of occurrence, the different existents their presence in consciousness indicates, or anything else, then the representation is a posteriori, or empirical. a priori psychology  the investigation of the nature and origin of a priori conscious representations; it can but need not take the form of a priori psychologism. aesthetic  the theory of sensibility. Sensibility includes all representations lacking the logical features that would suit them for inclusion in propositional representations: sensations, their reproduction in memory and fantasy, associations, and the pure formal intuitions generated in productive imagination. Transcendental aesthetic is the branch of a priori psychologism that focuses on pure sensibility: formal intuitions, the non-sensational appearances they make possible, and the two a priori features of sensations, manifoldness and the indication of a physically real existence. analytic/synthetic  a subject-predicate proposition is analytic if the predicate has to be incorporated into the subject concept merely in order to think it preliminary to predicating anything of it, e.g. “Gold is metallic”; otherwise, the proposition is synthetic, e.g.  “Gold is ideal for jewelry.” appearance  the non-sensational sense-divide transcending object synthesized in productive imagination to exhibit heterogeneous sensations in intuition. The form of appearances exhibits the manifoldness of sensations, their matter the existence indicated by the presence of sensations.

320  Glossary of Kantian Terms apperception  Kant’s preferred term for self-consciousness and the supreme principle of his transcendental philosophy (i.e. a priori psychologism). Apperception in its pure form has two sides, each analytically entailing the other: the synthetic unity of apperception, which is the unity of many representations in a single consciousness, and the analytic unity of apperception (a.k.a. the ‘I think’), which is the representation of the identity of that consciousness in respect to all the representations united in it. The structure of the former is that of a many-in-one individual, that of the latter a one-in-many universal (commonality, generality). Pure apperception can also be termed transcendental insofar as it specifically relates to the possibility of objective cognition. See also original vs. derived representation. apprehension  see synthesis of apprehension in intuition. association  in Hume, any transition in thought distinguished by a feeling of ease, or facility, in making it; in Kant, preferential reproduction of past representations (see synthesis of reproduction in imagining). Associative relations can be strengthened by becoming customary. AUA  analytic unity of apperception; see apperception. categories  concepts of objects so fundamental that they presuppose no other such concepts while all others presuppose them; psychologistically explicated by Kant as none other than the pure concepts of the understanding that derive their content entirely from logical forms of judgment insofar as they prove to be essential ingredients of principles of pure understanding. cognition  conceptual representations formed by synthesis of recognition that incorporate transcendental conditions of objectivity (schemata) and so can be used to represent sense-divide transcending objects via pure or empirical intuition. See also experience and objects. concept  any representation that incorporates analytic unity of apperception and thereby acquires formal logical universality (generality, commonality), i.e. an AUA-concept. consciousness, dark, bright/access  whenever even the barest minimum sensation is present, or sensation is reproduced in a dream, there is consciousness. If we are conscious of something without discerning or attending to it, then our consciousness of it is dark. When a content present in consciousness is discerned from others and singled out for attention, consciousness is bright and counts as access consciousness. If every constituent of such a content is sufficiently bright that it becomes impossible to mistake it for anything else, then our consciousness of it counts as clear. conventionalism  asserts that representations of a certain kind are impossible apart from human social convention. If that includes language, then anything that requires language to represent it is impossible apart from human sociality and lifeways, including objective discourse of the kind exemplified by science and mathematics.

Glossary of Kantian Terms 321 Conventionalism is an alternative to both psychologism and platonism insofar as these treat objective representations as independent of convention and its supporting social infrastructure. Copernican revolution in philosophy  the effect of Kant’s psychologistic proofs that mathematical and empirical objectivity, including nature itself, is entirely the product of our minds, so that instead of representations having always to conform to objects, objects necessarily conform to those of our representations that are constitutive of their objectivity. See also transcendental idealism. customary association  an association experienced with sufficient frequency and constancy to instill a habit of mind, thereby strengthening the association to the point of irresistibility. definition  a method of explicating meaning that excludes everything extraneous, leaving all and only those marks necessary and sufficient to single out the definiens from all other notions, especially those so similar to it that there is a real risk of confusing them when definition is lacking. Exponents of the method commonly take their explications to demonstrate that the notion is not merely verbally but objectively valid, and so in principle accessible to psyches and instantiable in reality independently of linguistic convention. empiricism  the view that nothing can be given to consciousness, whether for cognitive or other representational purposes, except through consciousness of particular sensations. It is opposed to both the platonist claim that some contents can be given by purely intellectual means and the psychologistic contention that some contents do not require consciousness of any particular sensation(s) in distinction from others. experience  subjective experience results from association or association plus synthesis of recognition in a concept, objective experience from the subsumption of recognitive concepts under transcendental schemata in transcendental judgments. (Interchangeable with cognition and not to be confused with perception.) form/matter distinctions  abstract distinctions derived from Aristotle, where ‘matter’ signifies a determinable and ‘form’ its constitutive determination(s). formal intuition  a pure intuition that comprehends the totality of possible appearances of a particular aesthetic form within it, e.g. pure space and pure time. Formal intuitions originate through productive imagination’s pure synthesis of apprehension. forms of judgment  see logical forms of judgment. Humean skepticism, Hume’s challenge  skepticism regarding the possibility of consciousness of a necessary connection between distinct existents, with the implication that such connections can be represented only by means of language, and so on the basis of social convention rather than purely psychologically. As extended by Kant

322  Glossary of Kantian Terms from cause and effect to other concepts classed as categories in traditional metaphysics as well as to mathematics (the necessary relation of distinct quantitative determinations), its definitive formulation became the question: how are synthetic a priori judgments—i.e. the necessary (a priori) relation (synthesis) of the distinct in a single act of conscious representation (judgment)—possible? More generally, Hume’s challenge requires one to show that a representation familiar from language is possible purely psychologically, i.e. independently of convention and its social infrastructure. I think (cogito), I  am (sum)  pure apperception, especially the analytic unity of apperception or universal self-consciousness. See apperception. imagination  the faculty of sensibility responsible for representing objects even in their absence from sensation. Imagination originally produces the appearances apprehended in intuition that it subsequently reproduces and associates. It is productive rather than merely reproductive if the contents are non-sensational and, at the same time, owe nothing to pure understanding, e.g. purely aesthetic formally intuited space and time. The understanding (especially the ‘I think’ and logical forms of judgment) must supplement imagination before its otherwise purely sensible representations can figure in propositional thought. intuition  the many-in-one structured representational consciousness in which manifolds of non-sensational appearances are originally apprehended, prior to being reproduced or associated. The doing of imagination as well as sense, intuition is a purely aesthetic notion with no logical (conceptual or propositional) component, so that it is vital never to confound Kant’s use of the term with the intellectualist sense in which other philosophers and mathematicians typically employ it. Intuition is empirical insofar as it involves consciousness of particular sensations and pure otherwise. See also formal intuition. judgment  see propositional thought. laws of nature  transcendental principles of pure understanding together with the laws posited in physics and other empirical sciences that Kant regarded as particular determinations of these principles. logic  the science of understanding considered independently of its relation to other faculties of representation, it relates exclusively to the logical universe demarcated by the universal self-consciousness ‘I think.’ Its scope is consequently limited to AUA-concepts, their synthesis to form propositions by means of logical forms of judgment, and the synthesis of the resulting propositions in more complex propositions by means of other such forms. So construed, logic excludes mathematical logic, which instead falls within the province of mathematics. logical forms of judgment  ways of synthesizing AUA-concepts or propositions formed from them so as to unite them in a single act of

Glossary of Kantian Terms 323 propositional thought, they alone enable the ‘I think’ to be extended from aesthetic to logically formed representations, and are essential to its constituting a genuinely universal self-consciousness. Though purely psychological in nature, like the logical universality of the ‘I think,’ they can and must be integrated into language as well. logical universe  demarcated by the representation ‘I think’ insofar as it defines both the sphere of representations possible for me (any the ‘I think’ cannot accompany, and so fails to be common to, are impossible, or at least nothing for me), and that which alone is capable of transforming non-concepts into AUA-concepts amenable to logical (rather than merely aesthetic) combination. Incorporating the ‘I think’ into a particular representational content to make it an AUAconcept ipso facto bifurcates the I-demarcated logical universe into those representations that incorporate the content and those that do not; and it is solely on the basis of that distinctively logical bifurcation that the logical forms of judgment are able to perform their function of combining such contents to yield propositions. The logical universe is moreover unique among universes—the universe of numbers, the universe of sets, the corporeal universe, etc.—by virtue of being the only one that is entirely unrestricted as to content and in no way dependent on the peculiar constitution of the human mind (i.e. our particular forms of intuition and/or forms of judgment and/ or sensory capacities and/or psychological capacities, including the human capacity for language). manifold  a multiplicity of sensations and/or the appearances that exhibit them that is otherwise completely undifferentiated and undetermined. In the synopsis of sense, there is only the manifold of sensations and no single consciousness to contain them, just heterogeneous sensory fields without any encompassing unity. In the imagination’s synthesis of apprehension, by contrast, the manifold of homogeneous appearances is united in a single intuitive consciousness as its manifold, even if the appearances therein are all scattered and single. Subjective organization of manifolds occurs through reproductive and recognitive synthesis. Objective organization requires transcendental synthesis, transcendental schematism, and transcendental judgment as well. mathematics  the science of quantity made possible by the relation of understanding to pure sensibility effected through transcendental synthesis. Its brand of cognition is just as synthetic as empirical cognition, but since the necessity and universality of its syntheses can be demonstrated entirely a priori, independently of all empirical considerations, it counts as pure cognition. nature  the sum total of physical realities existing and individuated in the objective transcendental space and time constituted by transcendental schematism (nature considered materially); and the system of natural laws to which these physical realities are subject insofar as

324  Glossary of Kantian Terms they are subsumed under transcendental schemata in principles of pure understanding (nature considered formally). objectivity  representations that are independent of sensation as regards both existence and individuality; may also refer to subjectively unconditioned reality, which for Kant is restricted to unrepresentable transcendental objects = X, i.e. things in themselves. objects  the empirically apprehensible sense-divide transcending spatiotemporal physical realities generated by transcendental schematism. Experience of objects (objective cognition) results when the threefold empirical synthesis of apprehension-reproduction-recognition is subsumed under transcendental schemata in principles of pure understanding. Mathematical objects are the a priori products of ostensive and symbolic mathematical construction made possible by transcendental synthesis. Transcendental objects are things in themselves. original vs. derived representation  a representation is original if it presupposes no other of its kind while all others of that kind presuppose it, and derivative otherwise. Example: the original synthetic unity of apperception constituted by purely aesthetic formal intuitions (unity of sensibility) vs. the derived synthetic unities of apperception constituted by the logical forms of judgment, transcendental synthesis, and transcendental schematism. perception  sensation of which there is consciousness. Insofar as there is a spatial or temporal component in the consciousness (i.e. nonsensational appearance), imagination’s synthesis of apprehension is as essential to perception as sensation, and ‘perception’ becomes interchangeable with ‘empirical intuition.’ (Not to be confused with experience, which involves not only perceptions but their association, transformation into AUA-concepts, and, insofar as experience is objective, the subsumption of AUA-concepts under transcendental schemata in principles of pure understanding.) platonism  alternative to both psychologism and conventionalism that accords a meaning to representations that transcends both purely psychological and convention-dependent representation. In its original Platonic form, it accords mind-independent reality to representational archetypes that are then instantiated both by real ectypes (things having the nature represented in the archetype) and by mental ones (representations of such things). The Aristotelian variant denies autonomous mind-independent existence to archetypal forms, insisting instead they can exist only in combination with a material substrate determinable by them. Because platonistic meanings can be accessed neither purely psychologically nor via convention-dependent representation, either they are supposed to be accessed in other ways—be it directly, through some form of intellectual illumination (e.g. Malebranche), or indirectly, through ectypes innately implanted in our minds (e.g. Descartes)—or the question of access is avoided.

Glossary of Kantian Terms 325 principles of pure understanding  transcendental judgments formed by the subsumption of physical realities represented via empirical synthesis (apprehension-reproduction-recognition) under transcendental schemata, they constitute laws of nature so fundamental that all empirically based laws of nature necessarily incorporate them. propositional thought  logically structured representations that, on Kant’s analysis, are not limited to sentences of a language but may also exist completely independently of language, viz. the logical combination of AUA-concepts made possible by logical forms of judgment. psychological philosophy, a.k.a. “theory of ideas”  investigates the origins of representations in the mind, including a priori representations. It becomes psychologism when it additionally seeks to determine whether consciousness contributes ingredients essential to their content. psychologism  a method of concept explication distinct from definition that proceeds by inquiring whether consciousness contributes ingredients essential to the concept’s meaning. If it is found to do so, then even if the concept is not overtly psychological, any attempt to apply it in contexts that exclude consciousness can result only in nonsense—no less so than in cases of overtly psychological concepts like pain, excitement, or grief. The method is useful because it is otherwise easy to be misled by the absence of anything overtly psychological about a concept into applying it in contexts that exclude consciousness, notable cases (for Kant) including cause and effect, substance, space, time, number, and logical universality. pure/empirical  Kant’s preferred terms for a priori/a posteriori. pure concepts of the understanding  AUA-concepts that derive their content entirely from logical forms of judgment and are therefore purely intellectual and a priori. They are formed by representing restrictions on the logical freedom permitted by logical forms of judgment, thereby introducing an extra-logical element of necessity into the relation of logically distinct representations in propositional thought. As essential ingredients of principles of pure understanding, they are central to Kant’s psychologistic explications of the categories of traditional metaphysics. representation  Kant’s replacement for Locke’s term ‘idea’ and Hume’s term ‘perception,’ encompassing all contents that can ever be present in consciousness as well as consciousness itself when it in turn is represented; contrasts with the unrepresentable thing in itself, i.e. reality insofar as it is present to consciousness neither directly (in sensation or intuition) nor indirectly (through concepts and propositions), and is therefore in the most literal sense nothing to us. rule of synthesis  see synthesis of recognition in a concept. schemata, schematism  transcendental schemata are pure concepts of the understanding insofar as transcendental synthesis has conferred

326  Glossary of Kantian Terms a sensible (e.g. spatial and temporal) significance on them, and transcendental schematism is the synthesis whereby the matter of appearances is determined conformably to transcendental schemata. The product of schematism is physically realized objective transcendental space and time. self-consciousness  in its pure form, an amalgam of the synthetic and analytic unity of apperception; in its empirical form, an empirical object experienced exclusively through internal sensations. See apperception and soul. sensations  the manifold given to consciousness in the synopsis of sense prior to and independently of imagination and understanding. External sensations include the data of vision, touch, smell, hearing, and taste. Internal sensations are self-affections, i.e. products of a sensitivity to one’s own mental activity together with the passions, emotions, desires, and volitions incidental to such activity. The presence of sensations is the condition for all other wakeful representation, a priori included, while the reproduction of past sensations is the condition for dreaming. sense  the sum total of a creature’s faculties of external and internal sensation. See synopsis. sensibility  faculty comprising both sense and imagination but exclusive of logical form, which has its source in the understanding. Pure sensibility consists of sense (insofar as consciousness of sensations is possible a priori) and productive imagination, and is the source of the unity of sensibility (original synthetic unity of apperception). soul  the internal sense-divide transcending empirical object that results when synthesis of recognition in AUA-concepts is applied to the temporal appearances that exhibit internal sensations and subsumed under transcendental schemata in transcendental judgments. SUA  synthetic unity of apperception; see apperception and original vs. derived representation. subject, subjectivity  in the strict sense, only the pure I of cogito ergo sum counts as subject, and all and only that bound up with its development from aesthetic I to logical I to objective I to I of nature counts as subjective. In a looser, purely epistemological sense, a representation counts as subjective insofar as it is inseparable from a posteriori consciousness of particular sensations either in existence or in individuation. synopsis  the most primitive level of sensible mentation where a manifold of sensations (e.g.  the contents of the visual field) is provided but without any consciousness to unite it with other manifolds of sensations (the contents of other sensory fields). Sensations can be united in consciousness only indirectly, insofar as they are exhibited in intuition as a homogeneous manifold of non-sensational appearances via the imagination’s synthesis of apprehension.

Glossary of Kantian Terms 327 synthesis  any purely psychological (non-linguistic) joining together of distinct representations in a single, unified conscious representation. For Kant, it is always the act of imagination or understanding and never the passive doing of sense. Its principal varieties are: —synthesis of apprehension in intuition  the purely aesthetic psychological process whereby heterogeneous sensations are exhibited by a manifold of formally homogeneous non-sensational appearances in a single intuitive consciousness—a prerequisite for the representation of sense-divide transcending objects. In beings constituted like us, appearances take spatial or temporal form, and the intuitive consciousness in which they are united is the formally intuited space or time synthesized by productive imagination. Appearances in apprehension are scattered and single, lacking all order and relation, subjective no less than objective. —synthesis of reproduction in imagining  the reproduction in thought of contents previously encountered in sensation and/or apprehended in intuition. It is associative insofar as particular representations are reproduced in preference to others owing to such causes as empirically ingrained customs of thought. Customary association is the principal means of ordering and relating apprehended appearances, though such ordering is always subjective, never objective. —synthesis of recognition in a concept  concepts produced by incorporating AUA into non-intellectual representations, particularly as a way of uniting the successive steps of an associative synthesis into a single representation that can thereafter serve as a rule of synthesis. —propositional synthesis  the combining of AUA-concepts in a single representation via logical forms of judgment; a second sort involves the combination of the resulting propositional representations in more complex propositions via other logical forms of judgment. —transcendental synthesis  the procedure whereby propositional networks are formed from the manifolds of pure formal intuition conformably to the pure concepts of the understanding; responsible for objectifying purely aesthetic formally intuited space and time and transforming pure concepts of the understanding into concepts suitable for transcendental schematism. —synthesis intellectualis, synthesis speciosa  the former concerns the incorporation of pure concepts of the understanding into the propositional networks formed from AUA-concepts obtained by attaching the ‘I think’ to the synthesis of appearances; the latter concerns the extension of these transcendental determinations to the synthesis of appearances (apprehension in intuition) itself. synthetic a priori judgment  see Humean skepticism. thing in itself (transcendental object  =  X, transcendental subject  =  X)  reality insofar as it is present to consciousness neither directly (in intuition) nor indirectly (through concepts), which is therefore, in

328  Glossary of Kantian Terms the most radical sense, unrepresentable, and therefore nothing to us. Nevertheless, the existence of things in themselves can be known either via analysis of sensation (the fact of its presence in us) or the sum of cogito ergo sum. transcendental idealism  the doctrine that space, time, causality, number, logical universality, and other representations fundamental to objective representation are subjectively conditioned, and so have no application to mind-independent reality. While leaving the validity of human knowledge unaffected, it prevents us from precluding the possibility of creatures with different formal intuitions and/or logical forms of judgment whose knowledge may rival or excel ours in, e.g., technological prowess despite being representationally entirely incommensurable with human understanding. See also Copernican revolution in philosophy and psychologism. transcendental judgment  see principles of pure understanding. transcendental philosophy  within the remit of this book, ‘transcendental philosophy’ is interchangeable with a priori psychologism, but, in keeping with the cognitive emphasis Kant gave to ‘transcendental,’ its particular focus is the psychologistic explication of objectivity. transcendental subject  see thing in itself. understanding  in its pure form, it is first and foremost the faculty of apperception, but secondarily the faculty of forming pure concepts, propositions, and inferences; in its empirical form, its primary business is to fashion the concepts most useful for rationally navigating the world disclosed by experience.

Bibliography

Ackerman, Jennifer. The Genius of Birds. Melbourne: Scribe, 2016. Baggott, Jim. Farewell to Reality. London: Constable & Robinson Ltd., 2013. Barron, Andrew, and Klein, Colin. “What insects can tell us about the origins of consciousness,” PNAS, 2015. Barron, Andrew, and Klein, Colin. “Insects have the capacity for subjective experience,” Animal Sentience, 2016: 100. Craig, A.D. How Do You Feel? An Interoceptive Moment with Your Neurobiological Self. Princeton: Princeton University Press, 2014. de Waal, Frans. Are We Smart Enough to Know How Smart Animals Are? New York: W. W. Norton & Company, 2016. Deutsch, David. The Beginning of Infinity: Explanations That Transform the World. New York: Viking, 2011. Gefter, Amanda. “The evolutionary argument against reality,” Quantum ­E-Magazine, April 21, 2016. Hegel, G.W.F. Science of Logic. Translated by A.V. Miller. London: George Allen & Unwin Ltd., 1969. Heidegger, Martin. Being and Time. Translated by John Macquarrie and Edward Robinson. New York: Harper & Row Publishers, 1962. Hoffman, Donald D. “Conscious realism and the mind-body problem,” Mind & Matter, Vol. 6, 2008. Hossenfelder, Sabine. How Beauty Leads Physics Astray: Lost in Math. New York: Basic Books, 2018. Krauss, Lawrence M. Hiding in the Mirror: The Quest for Alternate Realities, from Plato to String Theory (by way of Alice in Wonderland, Einstein, and The Twilight Zone). New York: Viking Adult, 2005. Linden, David. Touch: The Science of Hand, Heart, and Mind. New York: Viking, 2015. Mallatt, Jon, and Feinberg, Todd. The Ancient Origins of Consciousness. Cambridge, MA: MIT Press, 2016. Musser, George. Spooky Action at a Distance. New York: Scientific American/ Farrar, Straus and Giroux, 2015. Pinker, Steven. The Blank Slate: The Modern Denial of Human Nature. New York: Viking, 2002. Pross, Addy. What Is Life? How Chemistry Becomes Biology. Oxford: Oxford University Press, 2012.

330  Bibliography Sterling, Peter, and Laughlin, Simon. Principles of Neural Design. Cambridge, MA: MIT Press, 2011. Tomasello, Michael. Constructing a Language: A Usage-Based Theory of Language Acquisition. Cambridge, MA: Harvard University Press, 2003. Tomasello, Michael, Carpenter, Malinda, Call, Josep, Behne, Tanya, and Moll, Henrike. “Understanding and sharing intentions: The origins of cultural cognition,” in Behavioral and Brain Sciences, Cambridge: Cambridge University Press, 2005. Verdral, Vlatko. “Living in a quantum world,” Scientific American, June 2011: 24. Wilan, Naomi, McCarthy, Rosaleen, and Brewer, Bill, editors. Spatial Representation. Oxford: Oxford University Press, 1993. Wilson, Edward O. The Social Conquest of Earth. New York: Liveright Publishing Corporation, 2012. Wittgenstein, Ludwig. Philosophical Investigations. Translated by G.E.M. Anscombe. Oxford: Blackwell Publishers Ltd., 1953. Wittgenstein, Ludwig. Wittgenstein’s Lectures: Cambridge 1930–2. From the Notes of John Kind and Desmond Lee. Edited by Desmond Lee. Chicago: University of Chicago Press, 1980.

Index

aesthetic see transcendental, aesthetic AI (artificial intelligence) 12, 316 algebra 180, 232, 236 – 7, 240, 251, 294 analogy see association, associative relation analytic philosophy 12, 23 – 7, 31n11, 224n2, 306 – 8 anti-physicalism 18 – 23 appearance 8, 47 – 50, 70 – 2, 76 – 7, 91n7, 103 – 7, 112, 115, 117, 124, 151 – 2, 156, 195n12, 203 – 8, 210, 212 – 16, 218 – 22, 225n9, 225n13, 226n25, 227n27, 231, 235 – 6, 239, 243, 250, 254 – 5, 258, 273, 275, 278 – 81, 283, 291, 293 – 7, 299, 301n11, 301n14, 302n37, 305, 311 – 12 matter of 266 – 72, 286, 289, 292, 300n9, 301n11, 307, 312 apperception, self-consciousness analytic unity of (AUA) as universal self-consciousness 22, 157 – 65, 168 – 77 AUA presupposes SUA 198 – 201, 223 concepts constituted by 160, 165n9, 166n19, 168 – 72 spandrel of original synthetic unity of apperception 223 intellectual character of 157 – 9, 164 as Kant’s supreme principle 7, 23, 165, 199, 222, 227n34, 241, 292, 300, 303, 307 objective unity 255 – 9, 286, 292, 302n31 pure 163, 166n18 self-consciousness as (analytically) both SUA and AUA 223 – 4, 228n36

subjectivity of 162 – 3 (see also subjectivity, the conscious subject) synthetic unity of (SUA) 22, 28, 174 – 7, 182 – 3, 207 – 8, 227n34, 229, 256 – 7, 292, 302n31 formal individuality of (intuitional) 198 – 200, 205 – 8, 214, 218, 220 – 3, 227n34, 229, 243, 245, 264 – 5, 296, 302n31 original 220 – 4, 241, 268, 293, 302n39, 317n3 transcendental 163 – 4, 293, 296 see also ‘I’, the ‘I think’ (cogito), the ‘I am’ (sum) apprehension, synthesis of 201 – 22, 225n6, 225n10, 226n25, 231 – 2, 234 – 6, 242 – 3, 245, 249, 260n9, 266 – 7, 271, 275, 278 – 9, 281, 283, 289, 291, 295, 297, 299, 311; see also imagination; perception a priori 4 – 6, 9, 11, 16n5, 16n7, 26 – 30, 31n7, 31n10, 143, 147 – 8, 163 – 4, 166n18, 166n21, 168, 206 – 7, 214 – 17, 225n13, 235, 239, 248, 257, 266, 276, 278 – 80, 282, 286, 288, 297 – 300, 308 – 10 Aristotle, Aristotelian 24, 140 – 1, 145n13, 166n21, 204 logic of see logic arithmetic 24, 180, 182, 294 association, associative relation 35 – 6, 84 – 90, 93 – 126, 126n2, 170 – 2, 177, 184, 187, 190 – 4, 194n5, 200 – 7, 211 – 12, 219 – 20, 226n21, 227n30, 239, 264, 266, 274 – 81, 284, 288 – 90,

332 Index 292 – 3, 301n19, 306 – 7, 311, 313 analogical 112 – 13, 117 basis of belief in real existence 94 – 101 and the body 118 and cause and effect 103 – 26, 127n20 cognitive application of 109 – 16 scientific and other “philosophical” relations 124 – 5 constituted by facility affect 94 – 7, 103 by contiguity in space and time 103 – 9, 113, 115, 117, 124 – 5, 193, 245 – 6, 249, 275, 293 and contrariety 115 – 16 customary 3, 36 – 7, 105 – 6, 127n20, 101 – 9, 112 – 15, 119 – 25, 139, 143, 168, 184, 219, 280, 284, 289 – 90, 293, 306 generality 106, 115 – 16, 156, 170 – 1 experience a cause of 95 – 6 and external objects, external world 117 – 22 and general causal laws 113 – 16 principles of 103 by resemblance 103 – 10, 112 – 13, 115 – 17, 124 – 5, 126n8, 133, 156, 164, 168, 170, 275, 293 and the self (personal identity) 118 – 22, 125 – 6 attention 40 – 5, 52 – 9, 74, 78, 81, 86 – 9, 100, 107 – 8, 111 – 12, 115, 203 – 4, 206 – 7, 218 – 19, 221, 273 – 4, 301n17 Baggott, Jim 145n13 Barron, Andrew 63n11 Bayesian analysis 110 beauty 23, 140, 304 belief (in matters of fact and real existence) 95 – 101, 105 – 7, 110 experience a cause of but not basis of 95 – 6 follows facility feeling 95 – 7 as force and vivacity feeling 98 – 101 see also association, associative relation Berkeley, George 6, 9 – 10, 15, 20, 33 – 5, 39, 55, 62, 63n14, 65 – 94, 105, 140, 143, 150 – 1, 168, 177, 185, 201 – 2, 204,

216, 227n30, 235, 296, 301n19 blank slate 28 – 9, 31n12, 51, 57 – 9, 77 – 8, 84, 87, 160 blindsight 41, 92n20 Block, Ned 62n6 body (human-animal) 12, 15, 68, 70 – 5, 78, 80, 91nn7 – 8, 108, 118, 121 – 2, 127n21, 148, 161, 201, 211, 218, 262 – 3, 270, 273, 283 – 4, 288 tactual bodily spatiality see space; touch, tactual sensation categorical form see logical forms of judgment categories 22, 30n6, 137, 149 – 50, 152, 164, 167, 183 – 5, 188 – 94, 194n7, 197nn32 – 3, 200, 208, 228n37, 239, 241, 247, 261n17, 262, 265, 278, 286, 288, 291, 293, 295 – 6, 300n10, 304 – 9, 311 grounded on apperception 166n20, 224n3 qualitative 185, 192, 258, 305 quantitative 192, 305 cause and effect (irreversible groundconsequent) 4, 6, 13, 26, 35 – 7, 49 – 50, 57 – 9, 103 – 16, 135 – 7, 166n19, 188 – 93, 241, 244 – 51, 256 – 7, 270, 279, 282, 285 – 6, 293, 295 – 8 cognition 14, 18, 30n5, 35, 45, 62, 65, 84 – 90, 93, 102, 109 – 16, 123, 125, 137 – 8, 143, 155, 163 – 4, 166n18, 166n21, 184 – 7, 194n5, 219, 231, 233 – 4, 240 – 1, 263 – 4, 275 – 98, 300n4, 306 – 7, 312, 314 – 16 community (category) 192, 258, 302n31 concepts, AUA-concepts 156 – 7, 160, 165n9, 166n19, 168 – 72 of the understanding 7 – 8, 31n9, 148 – 52, 166n20, 167, 179 – 80, 184, 187 – 94, 196n30, 197n33, 200, 208, 210, 217, 219, 224n3, 227n26, 229, 232, 234 – 6, 238 – 42, 244, 246 – 7, 250, 252, 255, 257 – 9, 261n20, 261n22, 261n26, 262, 265, 268 – 9, 272 – 4, 277 – 8, 289, 291 – 3, 296, 299, 300n8, 304, 307, 311

Index  333 explicate categories 183, 239, 265, 278 not innate 196n30 see also schematism, schemata; synthesis, synthetic unity consciousness 1 – 2, 8 – 16, 17n9, 17n13, 18 – 23, 26 – 30, 30n5, 31n11, 39 – 60, 62, 62n6, 63nn11 – 14, 64n16, 65, 67 – 70, 72 – 8, 81, 84, 86 – 90, 92n20, 93 – 4, 96 – 102, 104 – 10, 113, 116 – 22, 124 – 6, 129 – 36, 140, 143, 144n3, 157 – 64, 165n7, 165n9, 165n15, 166n19, 167 – 79, 181, 183, 186 – 7, 191, 193, 194n3, 195n9, 195n15, 196n30, 198 – 224, 225n13, 226n21, 227nn26 – 7, 227n34, 232 – 5, 237 – 9, 242 – 4, 246, 252, 257 – 8, 262 – 7, 269 – 83, 288 – 9, 294 – 8, 300n8, 301n19, 301n23, 306, 309, 316 access (bright) 45 – 52 (see also attention; discernment) dark 39 – 45, 52 – 8, 67, 73 – 4, 76 – 7, 87 – 8, 93, 100, 104, 107, 112 – 13, 116 – 17, 120 – 1, 124, 159 – 61, 169, 171 – 2, 181, 186, 190, 198 – 9, 203, 206, 210 – 11, 218 – 19, 221, 232, 266, 273 – 6, 278 – 81, 284, 301n17 hard problem 51 – 2 neural correlate of 13 – 14, 17n13, 28, 49 – 53, 55 – 7, 59 – 62, 63n14, 71, 73 – 4, 77 – 8, 84, 87 – 90, 94, 96, 99 – 100, 104, 110, 116 – 17, 129 – 30, 132, 143, 160, 163, 172, 177 – 8, 186, 190, 201, 203 – 4, 209 – 12, 214, 218, 221, 225n10, 262 – 3, 282 – 3, 295, 298 self- see apperception, selfconsciousness; ‘I’, the ‘I think’ (cogito), the ‘I am’ (sum); self sensation as criterion of 28, 39 as spandrel 54 convention 81, 113, 123, 126n4, 129 – 34, 138 – 9, 142, 143n1, 157, 161, 178 – 80, 185, 187, 192, 237, 252, 314, 316 conventionalism 3, 27, 36, 129 – 43, 147, 188, 192, 217, 304, 306 – 17 Copernican complex 61

Copernican revolution (Kant’s) see Kant, Immanuel cosmology see universe definition, method of 10, 17nn8 – 9, 31n11, 134 – 7 Descartes, René 6 – 7, 9 – 10, 17n8, 20, 22, 27 – 8, 31n11, 34, 41, 61, 103, 134, 140 – 1, 148, 165n10, 186 – 7, 199, 233, 263, 279, 292, 304, 307, 310, 324 importance for Kant 148, 155 – 62 Deutsch, David 145n13, 260n5 de Waal, Frans 33, 120 discernment (discrimination) 40 – 5, 53 – 8; see also attention; consciousness, access dualism 19 – 21, 52, 126, 282 property 21, 283 substance 20 – 1, 282, 310, 317n8 Edelman, Gerald 15, 62n6, 225n10 electromagnetism see fundamental forces empiricism 7 – 9, 12, 28, 138, 164, 188, 233, 304 epistemology 3 – 4, 7, 9, 11, 15, 16n4, 27, 96, 126n4, 155, 163, 196n29, 227n27, 277, 307, 309, 314 maxmimalism 60 – 2, 139, 293 evolution 2, 12 – 14, 17n14, 30, 46, 52, 54, 57 – 62, 64n16, 78, 96, 98 – 9, 110, 113, 142 – 3, 172, 177 – 8, 201, 203 – 4, 223, 239 – 40, 273 – 5, 306, 308 – 9, 313, 316 existence (non-categorial) 266 – 7; see also belief existence-non-existence (irreversible assertoric) 189 – 90, 192, 196n27, 251, 257, 267 experience 7 – 8, 11, 19 – 21, 23, 28 – 9, 39, 42 – 4, 46, 49, 51, 55, 59, 63n11, 66 – 7, 78 – 80, 84 – 90, 90n1, 92n14, 95 – 6, 99 – 100, 104 – 5, 107 – 17, 121, 123, 127n20, 131 – 2, 137 – 8, 140 – 1, 143n1, 155 – 6, 166n18, 184, 192, 196n30, 204, 206, 213 – 17, 219, 226n14, 231 – 2, 235, 240 – 1, 254 – 5, 262, 265, 273, 275 – 6, 277 – 84, 287 – 94, 297,

334 Index 299 – 300, 301n27, 302n33, 312, 316 objective 8, 96, 277, 284, 290, 293 facility feeling see association force and vivacity see belief Frege, Gottlob 15, 23 – 4, 26, 30n7, 36, 140, 178, 180, 185, 194n2, 238, 306 function 237 fundamental forces (electromagnetism, gravity, weak, strong) 25, 45, 47, 229 – 31, 255, 259n2, 285, 287 – 8, 293 Galileo, Galilei 1, 289 geometry (Euclidean/Non-Euclidean) 2, 4, 11, 24 – 5, 31n10, 151, 180, 186, 208 – 9, 230, 232 – 3, 235 – 7, 240, 249 – 51, 259, 259n5, 294 God 159 gravity see fundamental forces ground-consequent propositions see cause and effect; logical form of hypothetical hearing 40 – 1, 70, 72 – 3, 85 language 81 – 2 Hegel, G.W.F. 9, 16n7, 140, 226n14, 304 – 6, 317n3, 317n6 Heidegger, Martin 308 – 10 heterogeneity problem 150, 152, 167 – 8, 191 – 4, 240 – 7, 262 Hoffman, Donald D. 17n14 Hossenfelder, Sabine 63n7 Hume, David 3 – 6, 8, 10 – 11, 15, 16n4, 16n7, 17n9, 19, 21, 33 – 7, 39, 55, 59, 63n11, 63n14, 64, 85, 87, 90, 93 – 143, 144n3, 147, 149, 153, 156 – 61, 164, 167 – 8, 170 – 1, 177, 184, 186 – 7, 192 – 3, 203, 205, 211 – 17, 219 – 20, 225n8, 226n22, 227n31, 233, 235, 241, 249, 262, 274 – 5, 280, 296, 298 – 9, 301n19, 301n23, 307, 313 on animal consciousness 93, 99, 101 – 3, 108 – 12, 119 – 20, 122 human nature vs. artifice 102 importance for Kant 6, 8, 93, 137 – 8, 147 – 8 missing shade of blue exception 127n20 quandary 125 – 6, 219 – 20

‘I’, the ‘I think’ (cogito), the ‘I am’ (sum) 7 – 8, 22 – 3, 27 – 8, 155 – 65, 198 – 200, 304 – 17 aesthetic 8, 150 – 1, 168, 198 – 224, 309 logical 8, 149 – 50, 167 – 94, 309 of nature 8, 152 – 3, 262 – 300, 309 objective 8, 151 – 2, 229 – 59, 309 subjectivity of 162 – 3 idealism 4, 9 – 10, 19 – 21, 52, 63n14, 90, 91n8, 126, 153, 262, 283, 305 transcendental 296 – 8 imagination 84 – 5, 155 – 6, 204 – 19, 226n25 apprehensive see apprehension, synthesis of associative see association, associative relation; imagination, reproductive cognitive role of 84, 89 locus of multi-sensory external sense 85 pure, productive see apprehension, synthesis of reproductive 151, 205, 217 – 18, 225n13, 250, 265, 275 – 6, 280, 311 see also apperception; self individuation see objectivity infants 90n1, 94, 116 – 17, 120 – 1, 127n17, 172, 225n10 information theory 15, 63n16 innate, innatism 4, 16n7, 26 – 30, 31nn11 – 12, 143, 149, 159, 178, 187, 190 – 1, 196n30, 205, 225n13, 265, 306 integrated information theory of consciousness 63n16 intelligence 18, 21 – 3, 28, 45, 52, 99, 102 – 3, 111, 129, 167, 171 – 2, 184, 187, 194n2, 224, 258, 276, 287, 296, 307, 316 introspection 13, 42, 44 – 5, 83, 203 intuition intellectual 186 – 9, 302n39 see also sense, sensibility James, William 30n7, 178, 196n21 judgment see propositions, propositional thought; transcendental Kandel, Eric 147 Kant, Immanuel

Index  335 on animal consciousness 30, 172, 177, 300n4 anti-psychologistic interpretations of 2 – 5 Copernican revolution of philosophy of 5 – 6, 11, 15 – 16, 265, 291, 317n4 on dark consciousness 44 – 5, 62n2, 219, 273 – 5 rejection of innate representations 29 – 30 supreme achievement 229 updating 2 – 3, 11 – 16 Kauffman, Stuart A. 30n3 Klein, Colin 63n11 knowledge, contrasted with understanding see understanding Koch, Christof 30n3, 63n16 Krauss, Lawrence M. 31n10 language 9, 11, 15 – 16, 17n9, 17n13, 19, 23 – 4, 26 – 7, 36 – 7, 41 – 3, 46, 56, 61, 75, 81 – 9, 90n1, 91n12, 93, 98, 102 – 4, 111, 113, 122, 125, 127n18, 129 – 35, 138 – 42, 142n1, 144n2, 144n6, 147 – 8, 157 – 9, 167, 169 – 70, 172, 177 – 87, 190, 192 – 3, 194n2, 194n6, 196n21, 208, 224n2, 234, 237 – 40, 256, 258, 272, 276, 278, 284, 289 – 90, 308, 310 – 16 Chomskyean linguistics 15 conventions of see convention private 183 – 4, 196n20, 238, 311 – 13 laws of nature see nature Leibniz, G.W.F. 4, 15, 16n7, 23, 27 – 8, 62n2, 134, 140 – 1, 187, 199, 221, 223, 233 Locke, John 3, 6, 9 – 10, 28 – 9, 31nn12 – 13, 33 – 4, 42 – 4, 52 – 7, 63n10, 63n14, 64n15, 65 – 9, 76, 90n1, 98, 103, 140, 144n3, 178, 199, 235, 263, 265 – 9, 272, 274 – 5, 277 logic 2 – 6, 8 – 12, 15 – 16, 18, 21 – 8, 30, 36, 42, 51, 53, 60, 82, 101, 106 – 7, 115, 126n4, 126n8, 127n20, 132 – 4, 136 – 9, 143, 148 – 50, 159, 160 – 2, 164, 165n10, 167 – 94, 194n2, 194nn4 – 5, 198 – 200, 203, 205 – 10, 217, 222 – 4, 224n3, 225n9, 232, 236, 238 – 9,

241 – 4, 246 – 52, 256 – 8, 290, 292, 294 – 6, 299, 302n31, 304 – 16, 317n6, 317n8 Aristotelian 4, 24, 173 – 5, 178 – 9, 181 – 2, 185, 195n9, 205, 255, 306 distinguished from mathematics and mathematical logic 24, 31n9, 178 – 83, 194n2, 195n14, 238 psychologization of 5, 11 – 12, 30, 30nn6 – 7, 160n20, 164, 166n20, 169 – 84, 195n10, 196n21, 240 logical forms of judgment 8, 22, 30n4, 172 – 7, 195nn9 – 16, 261n22 apodeictic 195n9 assertoric 176, 182, 188 – 90, 195n9, 195n13, 195n15, 196n27, 251 categorical (reversible subjectpredicate) 174 – 6, 182 – 3, 187 – 8, 195n9, 196n27, 238, 242 – 5, 248, 251, 293 – 5 disjunctive 195n14, 302n31 hypothetical (reversible groundconsequent) 176, 182 – 3, 188 – 90, 195n9, 195n15, 196n27, 242 – 5, 247 – 8, 251, 293, 295 problematic 176, 182, 188 – 90, 195n9, 195n13, 251 psychologization of 172 – 7 qualitative 174, 195n13 quantitative 174, 182, 195n9, 196n19 singular 195n9, 196n19 logical universe 169 – 78, 182, 190, 194n2, 198, 200, 205, 238, 292, 309 Malebranche, Nicolas 27, 141, 187, 324 Markopolou, Fotini 253 – 4 materialism (elimitivist) 52; see also physicalism mathematical logic 2, 4, 15, 20 – 1, 23 – 6, 178 – 80, 182, 191, 194n2, 195n14, 237 – 8, 240, 258, 294, 306, 311 mathematics 2 – 7, 9 – 13, 15 – 16, 18, 20 – 8, 30n7, 44, 55, 60 – 1, 62n2, 102, 123 – 5, 126n4, 130 – 1, 135, 137 – 43, 145n13, 147 – 8, 163 – 4, 166n21, 178 – 87, 191, 208 – 10, 215,

336 Index 217, 224, 225n4, 227n26, 232 – 41, 243, 249, 251, 255 – 9, 259n5, 260n12, 262, 275, 288 – 92, 294, 296, 298 – 300, 302n35, 306 – 11, 314 – 15 distinguished from logic 24, 31n9, 178 – 83, 194n2, 195n14, 238 see also algebra; arithmetic; geometry; number mathematizability of nature see nature maximalism see epistemology Mayr, Ernst 141 memory 19, 41 metaphysics (ontology) 11, 21, 52, 55, 163, 216 – 17, 230, 283, 310 of nature 166n21 mind-body problem 14 – 15, 51 – 4, 57 Molyneux, William 65 morality 23, 30n5, 93, 143, 161 – 2, 165n10, 191, 304, 309, 317n2 nativism see innate, innatism nature 1 – 5, 7 – 10, 12, 21 – 3, 25, 30n5, 61, 93, 97, 113 – 14, 116 – 17, 127n31, 130, 134, 262 – 99 laws of 7, 25, 60, 93, 113 – 14, 116, 122, 136, 140, 145n10, 149, 153, 196n30, 224, 229, 241, 262, 264 – 5, 270, 273, 276 – 7, 279 – 80, 282, 289 – 93, 295 – 6, 301n14, 301n28, 302n34, 315, 317n4 mathematizability of 288 – 92, 314 – 15 psychologized 272 – 3 necessary connection of the distinct 106 – 8, 135 – 7 necessity-contingency (irreversible apodeictic) 192, 258, 302n31 neural correlate of consciousness see consciousness neuroscience, neuroscientists 17n13 neutral monism 19, 21, 52 Newton, Newtonian 4, 11, 25, 90, 166n21, 259 number 26, 133, 185, 191, 209, 213, 233 – 7 objectivity 10 – 11, 60 – 2, 135 – 7, 215 – 17, 238 – 40, 262 – 6 objects 162

external 35, 65 – 90, 91n12, 103, 117 – 22, 127n27, 127n31, 213, 218, 241, 264, 284, 310 internal 98, 117 – 22, 127n31, 211, 213, 218, 241 sense-divide transcending 35, 44, 62, 86, 89, 104, 117, 121, 127n31, 151, 201 – 4, 206 – 7, 211, 213 – 15, 218, 224, 227n30, 241, 256, 264 – 5, 267, 270, 274 – 5, 278, 296 – 7, 306, 312, 314, 316 transcendental see thing in itself, transcendental object = X O’Keefe, John 31n10 ontology see metaphysics oyster 28, 34, 53 – 7, 63n11, 63n14, 64n16, 98, 143, 165n15, 168, 172, 213, 296 panpsychism 21, 28, 30n3, 39, 54, 63n12, 165n15, 199 Pearl, Judea 110 Penrose, Roger 15, 253, 301n29 perception 20, 39, 41 – 2, 44, 49, 53, 56 – 7, 89, 96, 117, 120 – 1, 124 – 5, 126n2, 157 – 8, 163, 166n18, 200, 204 – 5, 212, 215 – 16, 219, 226nn21 – 2, 232, 266, 280 – 1, 284 – 5, 289; see also apprehension, synthesis of phylogeny 12 – 15, 25, 34, 45 – 62, 79 – 81, 95, 103, 140 – 3, 173, 176, 178, 186 – 7, 195n10, 201, 204, 212 – 14, 238, 240, 258, 263, 266, 293 – 4, 298, 309 – 10 physicalism 18 – 23, 30n2, 90, 282 Pinker, Steven 31n12 Plato 27, 139 – 40 platonism 4, 16n7, 26 – 7, 36 – 7, 103, 139 – 43, 145n13, 147 – 8, 159, 178 – 80, 184 – 93, 199, 216, 233, 235, 240 – 1, 256, 258 – 9, 284, 290, 296, 304 – 8, 310 – 11, 313 – 17 possibility (purely aesthetic sense) 266 – 7 possibility-impossibility (irreversible problematic) 189, 244 – 5, 247 – 8, 251, 256 principles of pure understanding 7 – 8, 197n33, 273, 291 – 3, 296,

Index  337 299 – 300, 302n31, 302n35, 302n37, 314, 317n4 propositional functions 237 – 8 propositions, propositional thought 3, 8, 21 – 3, 28, 30n6, 51, 53 – 6, 60, 102, 129 – 30, 133, 149 – 51, 164, 167, 169, 172, 175 – 85, 188 – 94, 195n9, 195n11, 195nn13 – 15, 196n27, 196n30, 200, 203, 205, 207 – 11, 224, 232, 238 – 52, 254, 256 – 8, 263, 266, 268 – 9, 272 – 81, 289 – 90, 292 – 3, 295 – 300, 304, 308 – 17 non-linguistic 3, 129, 134, 158, 167 – 8, 175, 177 – 84, 194n2, 238 – 52, 254, 256 – 8, 272 – 81, 289 – 90, 295 – 300, 310 – 17 synthetic vs. analytic 4, 224n2 synthetic a priori 16n3, 138, 298 – 300 see also logic; schematism, schemata; synthesis, transcendental Pross, Addy 63n9 psychologism 8 – 11, 17n9, 26 – 7, 33 – 4, 93 – 126, 155 – 318 a posteriori (empirical) 6, 8, 30n7, 93 – 126, 147 a priori (transcendental) 1 – 2, 5 – 6, 18, 147 – 8, 164 – 5, 171, 177 – 84, 226n26 different levels in Kant’s 8, 172 – 4, 202, 215, 293 psychology 3 – 6, 9 – 11, 13, 15, 16n7, 26, 33 – 4, 37, 52, 63n10, 63n14, 96 – 7, 102, 116, 123, 129, 141 – 3, 150 – 1, 157 – 8, 164 – 5, 166n18, 166n20, 168, 170 – 2, 177 – 9, 187, 193, 194n5, 204, 219, 229, 239, 257, 265, 272, 278, 280 – 1, 296, 306 – 7, 311, 314 – 16 prior to ontology 20 – 3, 52, 283, 310 pure see a priori quale, qualia 18 – 21, 23, 31n11, 34, 39, 55 – 8, 63n11, 162 – 3, 191 quantum theory 2, 4, 12, 15, 25, 47, 49, 127n18, 152, 156, 166n21, 229 – 31, 243, 253 – 5, 258 – 9, 271, 289, 292, 294

many worlds interpretation 302n35 Rationalists, Rationalism 4, 9, 27 reality see belief; existence consciousness-independent (mind-, phylogeny-) 10 – 11, 14 – 15, 17n9, 26, 35, 46, 48, 59 – 60, 62, 67, 116, 135, 139 – 42, 176, 189 – 90, 227n26, 240, 258, 263, 274, 284, 296, 300n10 physical 262 – 95 recognition, synthesis of 45, 62n5, 243, 275 – 81, 288, 292, 296, 306, 311, 313 relation see association, associative relation relativity theory 2, 4, 12, 20 – 1, 49, 166n21, 212, 229, 259, 271 – 2, 289, 292, 294; see also spacetime representation 162 origin of conscious 9, 26 – 30, 31n11, 55, 61, 64n16, 125, 151, 165n15, 173, 178 – 9, 184, 190, 205, 210, 217, 232, 278, 301n14, 302n34, 308 (see also imagination; psychology; sense, sensibility) psychologism, platonism, conventionalism as original acquisition 29, 31n15, 196n30, 225n13 reproduction, synthesis of see imagination rules of synthesis 194n5, 272 – 81, 293 schematism, schemata 7 – 8, 152, 197n33, 260n11, 267 – 84, 286, 288 – 93, 296 – 7, 299, 301n11, 302n31, 302n37, 307, 317n4 science, scientists 2 – 4, 6, 8 – 16, 17n11, 17nn13 – 14, 18 – 30, 30n2, 30n7, 31n10, 34 – 5, 39, 41 – 3, 45 – 6, 51, 56, 60 – 1, 63n12, 64n16, 65, 80 – 1, 86, 89, 93, 102, 123 – 5, 130 – 1, 135, 138 – 43, 144n2, 145n13, 148, 150 – 1, 156, 161 – 2, 164, 179 – 80, 184 – 5, 187, 217, 227n26, 229 – 30, 232 – 3, 239 – 41, 243, 256 – 8, 259n5,

338 Index 261n25, 263, 265, 273, 275, 286, 288 – 91, 294 – 6, 299 – 300, 306 – 12, 314 – 17, 317n2 self (soul, person, empirical I) 15, 118 – 22, 125 – 6, 218 – 19, 280 – 4 sensation 13 – 14, 28 – 9, 34 – 5, 39, 41 – 62, 63nn11 – 13, 64nn15 – 16, 65 – 90, 90n1, 91n9, 94, 96 – 109, 111 – 13, 115 – 23, 127n20, 131, 133, 149 – 51, 163 – 4, 165n10, 168, 170 – 3, 175 – 8, 186, 191, 200 – 23, 225n8, 225n13, 226n14, 226nn16 – 17, 231 – 5, 239 – 43, 247, 250, 254 – 8, 262 – 8, 270 – 1, 274, 276 – 83, 286 – 8, 291 – 3, 295 – 9, 300nn8 – 10, 305 – 7, 309, 312, 315 – 16 and belief 98 as criterion of consciousness 28, 39 evolutionary development 57 – 8 external 59, 65 – 90, 210 – 11, 215, 218 – 22, 226n16, 226n22, 231, 234, 239, 242, 248, 262 – 4, 266, 270, 281, 283, 288, 306, 312 internal 151, 210 – 19, 225n8, 226nn16 – 17, 226n22, 231, 234, 239, 243, 248, 262, 271, 300n8, 312 presence presupposed by all other representations 163 simple 57, 63n15 as spandrel 57 see also hearing; sense, sensibility; synopsis; touch, tactual sensation; vision sense, sensibility 4, 21, 65 – 90, 198 – 224 -divide transcending see objects external 21, 23, 35, 65 – 90, 85 – 6, 89, 151, 201 – 4, 211 – 14, 218 – 19, 262 – 4, 283, 306 internal 21 – 3, 34, 44, 49, 138, 151, 160 – 1, 210 – 13, 215, 216 – 18, 220 – 2, 226nn16 – 17, 235, 260n9, 280 – 3 intuition 1, 201 – 8, 211, 214 – 15, 221, 225n4, 225n12, 250, 260n9, 267, 278, 283, 296, 305

a priori (pure formal) 2, 22, 28, 31n9, 152, 207 – 8, 210 – 11, 215 – 24, 226n14, 226n25, 227n34, 231 – 2, 234 – 50, 252, 255 – 7, 266, 274 – 5, 281, 293, 297, 299, 301n14, 305 – 7, 311, 316 empirical 163, 278, 281 unity of 220 – 2 sets, set theory 191, 237 sight see vision skepticism 3, 35 – 7, 61, 96, 123, 125, 134 – 43, 147, 149, 157 – 60, 167, 220, 233, 241, 249, 292, 298 – 300, 307 Smolin, Lee 145n10, 301n28 space 25, 75, 201 – 10, 229 – 95 adjacency 246, 249 – 51 dimension 250 – 1, 259n5 direction 250 – 1 distance 250 essential to mathematics 25, 232 – 8 and external sensation 65 – 92, 201 – 10 objective transcendental 209 – 10, 229 – 59 physically realized by schematism 262 – 95 purely aesthetic 22, 201 – 10 representation of requires acts of synthesis 44 – 5 routes through 250 see also sense, sensibility; touch, tactual sensation; vision spacetime (block universe) 20, 25, 185, 213, 217, 229 – 30, 271 – 2, 285 spandrel 54, 57, 177, 223 Spinoza, Benedict 27, 134, 140 spontaneity 210, 300n8 Standard Model see fundamental forces Steinhardt, Paul 301n29 Strawson, Peter 16n7, 306 – 8 subjectivity, the conscious subject 7, 14, 18 – 28, 30n2, 30nn4 – 5, 31n11, 42, 49, 56 – 7, 59, 63n11, 96, 122, 133, 138, 140, 143, 150 – 2, 162 – 3, 177 – 8, 191, 196n39, 200, 209 – 10, 213 – 17, 219, 226 – 7nn25 – 7, 229, 231, 235,

Index  339 240, 243, 257 – 8, 263, 266 – 8, 271, 277, 279 – 80, 282 – 3, 291 – 8, 300, 300n8, 301n27, 307, 309 – 10, 316, 317nn7 – 8 subjective deduction of the categories 30n6, 194n7, 197n32, 261n17 transcendental 162 – 3 substance and accident (irreversible subject-predicate) 137, 156, 159 – 60, 166n19, 185, 188, 191 – 2, 196n25, 199 – 200, 220, 244 – 6, 248 – 51, 254 – 7, 259, 261n24, 269 – 72, 279, 282, 284 – 8, 290 – 1, 293 – 6, 302n35, 305 synopsis 203, 205, 211, 214 – 15, 220, 225n6, 297, 311 synthesis, synthetic unity 4, 174 transcendental 7 – 8, 148, 152, 180, 183, 196n28, 210, 217, 219, 227n27, 234 – 6, 239 – 55, 257 – 9, 260n5, 260n11, 261n20, 261n22, 262, 265 – 6, 268 – 77, 289, 291, 293, 296 – 7, 299, 300n8, 301n11, 302n31, 302n35, 304, 306 – 7, 312, 316 see also apprehension, synthesis of; recognition, synthesis of; reproduction, synthesis of synthetic a priori judgments see propositions, propositional thought tactual spatiality see touch, tactual sensation Tegmark, Max 140, 289 thing in itself, transcendental object = X 11, 162 – 3, 210, 296 – 8, 317n4 thought 59 – 62 and belief 98 – 100 objectivity of 60 – 2 possibility of 23 time (succession, temporality) 20 – 2, 48 – 9, 160 – 1, 211 – 95 essential to mathematics 232 – 8 perception of 48 – 9 objective transcendental 215 – 17, 229 – 59 physically realized by schematism 262 – 95

purely aesthetic 22, 213 – 20 and unity of sensibility 220 – 2, 227n33 Tomasello, Michael 143n2, 315 Tononi, Giulio 1, 15, 63n16 touch, tactual sensation 14, 20 – 1, 29, 35, 42 – 5, 48, 56, 64n15, 65 – 90, 90nn1 – 3, 91nn8 – 9, 91n12, 92n15, 92n20, 98, 111, 117 – 18, 120, 122, 150 – 1, 163, 201 – 4, 206 – 7, 209 – 12, 216, 218, 220, 231, 234, 236, 250, 254, 262 – 3 heterogeneity with visual sensation 68 – 9, 75 – 7 spatiality confined to for Berkeley 69, 73 – 7 transcendental 25, 163 – 4 aesthetic 200 – 1 judgment 197n13, 275 – 81, 287 – 9, 299, 307, 313, 315 see also a priori truth-functional logic see mathematical logic Turing, Alan 1 Turing machine 88 Turing test 161 understanding 3 – 4, 30n5, 35 – 6, 52, 65, 86, 89, 93, 102 – 3, 116, 123, 126, 131, 133, 140 – 3, 147 – 8, 150 – 2, 156 – 7, 164, 165n10, 167 – 8, 170, 173, 179 – 80, 187, 190, 194nn2 – 3, 194n6, 205, 207, 209, 217, 223, 224n2, 226n14, 227n34, 232, 256 – 7, 265, 273, 279, 288, 290 – 3, 295, 300n4, 301n14, 302n34, 302n37, 306, 309 contrasted with knowledge 59 – 62, 130, 139, 257 – 8, 293 – 4 intuitive 302n39 pure concepts of see concepts, AUA-concepts universality (logical), commonality, generality see apperception, self-consciousness; association, associative relation; concepts, AUA-concepts universe (multiverse), Universe 35, 47 – 8, 63n7, 111, 125, 129, 169, 213, 230, 254, 258,

340 Index 271 – 2, 285 – 9, 291, 293, 295, 301nn28 – 9, 302n31 Verdral, Vlatko 230 vision 28, 65 – 92 experiencing (vs. inferring) 78 – 9, 81, 85, 90n1 experiencing secondary meanings 78 – 81, 87 – 9 as language of tactual spatiality 80 – 7 and mien 77 – 81 non-spatial character of 67 – 70, 73 – 7

spatiality of dependent on association 70 – 1, 73 – 5 in three dimensions 29, 42 – 4 Welles, H.G. 100 Welles, Orson 100 – 1 Wilson, Edward O. 31n12 Wittgenstein, Ludwig 3, 24, 27, 31n8, 36, 130 – 43, 144n6, 147, 159, 180 – 5, 187, 196nn20 – 1, 238, 310 – 15 Wolff, Christian 4