Ontology and Analysis: Essays and Recollections about Gustav Bergmann 9783110327038, 9783110326666

Citation preview

Laird Addis, Greg Jesson, Erwin Tegtmeier (Eds.) Ontology and Analysis Essays and Recollections about Gustav Bergmann

Philosophische Analyse Philosophical Analysis Herausgegeben von / Edited by Herbert Hochberg • Rafael Hüntelmann • Christian Kanzian Richard Schantz • Erwin Tegtmeier Band 19 / Volume 19

Laird Addis „ Greg Jesson Erwin Tegtmeier (Eds.)

Ontology and Analysis Essays and Recollections about Gustav Bergmann

ontos verlag Frankfurt I Paris I Ebikon I Lancaster I New Brunswick

Bibliographic information published by Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliographie; detailed bibliographic data is available in the Internet at http://dnb.ddb.de

North and South America by Transaction Books Rutgers University Piscataway, NJ 08854-8042 [email protected]

United Kingdom, Eire, Iceland, Turkey, Malta, Portugal by Gazelle Books Services Limited White Cross Mills Hightown LANCASTER, LA1 4XS [email protected]

Livraison pour la France et la Belgique: Librairie Philosophique J.Vrin 6, place de la Sorbonne; F-75005 PARIS Tel. +33 (0)1 43 54 03 47; Fax +33 (0)1 43 54 48 18 www.vrin.fr

2007 ontos verlag P.O. Box 15 41, D-63133 Heusenstamm www.ontosverlag.com ISBN 13: 978-3-938793-69-5

2007 No part of this book may be reproduced, stored in retrieval systems or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use of the purchaser of the work

Printed on acid-free paper FSC-certified (Forest Stewardship Council) This hardcover binding meets the International Library standard Printed in Germany by buch bücher dd ag

CONTENTS INTRODUCTION

7

PHILOSOPHICAL PAPERS The Development of Bergmann’s Metaphysics ERWIN TEGTMEIER

13

The First Station of Gustav Bergmann’s Odyssey GUIDO BONINO

31

Bergmann as Historian ROBERT BAKER

51

Bergmann on the Synthetic A Priori Truth: Nothing Can Have Two Colors All Over at Once DONALD SIEVERT

59

Gustav Bergmann’s Quest for the Ontology of Knowing: From Phenomenalism towards Realism GREG JESSON

79

Time for Bergmann’s Bare Particulars ERNÂNI MAGALHÃES

123

Bradley’s Regress: Meinong versus Bergmann FRANCESCO ORILIA

133

Bergmann’s Thinkable Inexpressibles WILLIAM HEALD

165

Placing Bergmann FRED WILSON

185

REMINISCENCES Reminiscence of Gustav Bergmann CARL LEIDEN

277

Reminiscence of Gustav Bergmann (1987) REINHARDT GROSSMANN

283

Reminiscences of Bergmann’s Last Student L. NATHAN OAKLANDER

287

AFTERWORD

299

BIBLIOGRAPHY OF THE WRITINGS OF GUSTAV BERGMANN

303

Introduction

O

n 19-20 May 2006, the University of Iowa was the site of the first of three international conferences within a year to be devoted exclusively to the life and work of Gustav Bergmann (1906-1987). The other two conferences were held at the Université de Provence on 9-11 December 2006, and at the Università Roma Tré on 18-19 May 2007. The Iowa City conference, with more than one hundred people in attendance at a dinner and reception that included a display of photographs and other memorabilia of the life of Gustav Bergmann, brought together philosophers from Canada, Italy, Germany, and the United States to remember, to honor, and to discuss the life and work of the person many regard as the most important ontologist of the last half century. Whatever one may think of the answers he proposed, no one—in that period or any other—has pursued the basic questions of first philosophy as deeply as Bergmann, a fact to which many of the papers in this book are themselves testimony. The papers read at the conference consisted of nine philosophical studies (two with significant material on their authors’ associations with Bergmann) and two reminiscences. All of those papers, with the exception of one of the reminiscences, are included in this volume. Also included are the reminiscences of two of Bergmann’s students who were unable to attend the conference. The volume closes with an Afterword and a complete bibliography of Bergmann’s published writings. The book opens with Erwin Tegtmeier’s critical survey of most of the main themes of Bergmann’s philosophy by way of distinguishing the early, middle, and late periods of his thought. There follow seven papers on particular aspects of Bergmann’s thought. Guido Bonino discusses some of Bergmann’s earliest attempts, especially in relation to Carnap, at understanding the aboutness of language and thought, an issue that was to become, under the heading of intentionality, one of his consuming interests of later years. Bergmann’s unique way of doing history of philosophy, characterized by him as “structural history,” is discussed by Robert Baker,

8 along with reminiscence of his years at Iowa as a colleague of Bergmann. Donald Sievert, while remembering what it was like to be a student of Bergmann’s, engages in a wide-ranging discussion of Bergmann’s solution to the problem of what the ontological ground is of those truths that are neither analytic nor straightforwardly empirical, especially as treated by Bergmann in one of his most important papers, “Synthetic A Priori.” Gustav Bergmann’s long journey from the phenomenalism of his logical positivist beginnings to the robust realism of his later years is recounted by Greg Jesson, especially in connection with “the ontology of the knowing situation,” as Bergmann characterized epistemology. Ernâni Magalhães proposes to retain most of the features of Bergmann’s notion of the bare particular except for that of being momentary, treating this crucial feature of Bergmann’s ontology in the context of the philosophy of time. The different solutions of Bergmann and Meinong to the problems posed by Bradley’s regress are explored in great detail by Francesco Orilia, partly by way of comparing them to other proposals. William Heald discusses one of the most difficult features of Bergmann’s very late ontology: that of what can be thought but not, in an important way, expressed. The final philosophical paper, by Fred Wilson, is a very extended argument, with many detailed discussions of various of Bergmann’s ideas and comparisons of his philosophical methodology with that of some other major thinkers, that Bergmann’s philosophy always retained enough of its earliest features to be regarded as positivist. The three reminiscences contribute memories of different aspects and different times of Bergmann’s life and personality. Carl Leiden, who was a student in political science, took several classes with Bergmann in the late 1940’s, and recalls both his academic and his personal associations with Bergmann. Reinhardt Grossmann, whose contribution was written shortly after Bergmann’s death in 1987, wrote his doctoral dissertation under Bergmann in the late 1950’s; among other things, he tells of both his first and his last meeting with Bergmann. Nathan Oaklander, the last person to write a dissertation with Bergmann (in 1974), recounts, among other matters, what it was like to be a doctoral student under Bergmann. Expenses for the conference in Iowa City and for the editing of this book were covered by a generous grant, for which we express our gratitude, from the Carver Charitable Trust of Muscatine, Iowa. Also, in

9 conjunction with the conference and with donations from relatives, students, colleagues, and friends of Gustav Bergmann, a bench in his honor and memory was installed on a walkway along the Iowa River, a place where he often strolled when he needed a respite from his desk. A biography of Gustav Bergmann can be found in American National Biography, Oxford University Press, 1999, pp. 639-641. More material on him is available on the Bergmann website available through the University of Iowa Department of Philosophy website: http://www.uiowa.edu/~phil/.

Laird Addis

Philosophical Papers

The Development of Bergmann’s Metaphysics ERWIN TEGTMEIER Universität Mannheim

Bergmann’s Progress in Austria

I

n The Cambridge Dictionary of Philosophy, Gustav Bergmann is introduced as an Austrian philosopher. The author of the article: William Heald. But didn’t Bergmann become a US citizen in 1944? I suppose that Heald chose his words carefully. Presumably, the emphasis is on ‘philosopher.’ Did Bergmann never leave Austria philosophically? At least he changed sides in Austrian philosophy. From anti-metaphysics to metaphysics. From the Vienna Circle to the schools of Meinong and Brentano. The direction of this change was mainly due the influence of an American philosopher of German origin, Reinhardt Grossmann. I suppose one cannot call Grossmann an Austrian philosopher. Maybe Grossmann was more of an American philosopher than Bergmann. Both were system builders and in that respect continental philosophers. One can with good reason claim that at the end of the 19th and the beginning of the 20th century Austria was the center of ontology in the philosophical world. How did it become that center? Mainly due to the philosophical work of Brentano and Meinong, of course. But there was a favorable background. The homogenous Catholicism of Austria prevented a Kantian grip on philosophy like that in Germany (this is Bergmann’s own diagnosis). I think another not merely negative circumstance helped ontology a lot, that Johann Heinrich Herbart was made state philosopher of Austria, above all because of his philosophy of education, but it led to his other writings also being widely known in Austria. Herbart, a student of the arch-idealist Johann Gottlieb Fichte, later turned to realism and ontology taking up Parmenides whose fragments had been collected and republished at that time. Like the Eleatics, Herbart emphasizes the

14 importance of logic, and for some time the leading logic textbooks had been written by members of the Herbart school. Herbart was a key figure of the movement towards objectivity in the 19th century, as Passmore has called it; a movement away from idealism and towards epistemological realism. There was from the start an alliance between ontology and epistemological realism. Ontology was in need of renewal and vindication after being destroyed and distorted by Kant. Without overcoming idealism and promoting its antipodean epistemological realism, ontology would not have stood a chance of renewal. To keep the new ontology with the times and in accordance with its aim of creating a scientific philosophy, it is designed to reconcile empiricism and ontology. Before, from Descartes to Wolff, from the 16th to the 18th century, ontology was rationalist. Kant managed to use that against ontology and accused rationalist metaphysics of claims that cross the limits of human reason. Brentano gives the book that introduces his theory of intentionality the title Psychology from an Empirical Standpoint, and though he starts from Aristotle, he follows the French and British Empiricists in rejecting substances and universals. However, rejecting universals could also be derived from Aristotle himself. Brentano continues Aristotelian philosophical psychology, which was already empiricist, and he does not separate ontology from psychology, more precisely, from what he calls descriptive psychology. Meinong begins as a psychologist in the style of empiricism. He publishes papers on abstraction and relations, which he collects under the title of Hume Studies. Later he takes his turn to object theory and converts his earlier psychological results into contributions to object theory. Thus, the philosophical roots of the Brentano and Meinong Schools were not very different from that of the Vienna Circle. Nevertheless, there was the appearance of a radical difference; namely, that between metaphysics and anti-metaphysics. However, as you know, Bergmann has pointed out that the Vienna Circle did have a metaphysics. That should not be unexpected, taking into account that Wittgenstein’s Tractatus was the bible of the Vienna Circle. Well, they did not read the later parts of the book. But metaphysics is in it from the first page on. The contrast between the two Austrian schools of philosophy was rather that between metaphysics of better and lesser quality (that is surely not a contentious claim). Bergmann was molded by implicit metaphysics of

15 lesser quality and worked his way to the top-quality explicit metaphysics of the Brentano and Meinong School. The Reconstruction of Metaphysics Bergmann did not reject the central tenet of the Vienna Circle that metaphysical talk is nonsensical but added that such talk can be reconstructed. His method of reconstruction was a combination of the early Wittgenstein and Moore. He accepted Moore’s diagnosis that the metaphysical uses of key words are not commonsensical, giving it a new twist by proclaiming that it can be made meaningful by transforming it into talk about language, about an ideal language (IL), a language designed not for communication but only to represent the categorial structure of the world. Bergmann thus claimed that non-commonsensical metaphysical talk can be made meaningful by translating it into commonsensical talk about an ideal language. Instead of saying that substances and accidents exist, which would be non-commonsensical and meaningless, one should instead say that the ideal language has substance signs and accident signs, which would not be philosophical but commonsensical talk. Bergmann insists that all simple signs of the ideal language represent and criticizes Quine for holding that predicate symbols can be used without ontological commitment. Quine took, also in the tradition of the Vienna Circle, a linguistic approach to ontology. He is credited with being the first formal-analytic philosopher who vindicated the old ontological question. However, Bergmann pointed out that the question “what is there?” was not the old question and that Quine’s question could not but be answered with “everything” which showed that the philosophical use of ‘is’ is not the ordinary use as Quine presupposed. Bergmann distinguishes mainly two schools of ontology according to whether the basic tenet is either “to be is to be simple” or “to be is to be independent.” The simplicity school is characterized by Bergmann also as that of thing ontologies, the independence school also as that of fact ontologies. He considers substance ontologies as combining both criteria of being. A highlight of Bergmann’s application of the ideal-language method is without doubt to be found in his paper “Ineffability, Ontology, and

16 Method.” The fundamental difference between categorial and ordinary attributes is, in this paper, grounded linguistically by the futility of representing being an individual or being exemplified by a symbol. Bergmann contrasts that futility with the necessity of representing ordinary attributes such as being round or being carmine. Categorial attributes are represented iconically (in Peirce’s terms) by the shape of a symbol or by putting two symbols side by side. In Bergmann’s later philosophy and already in Realism, however, being an individual and being exemplified are symbolized. There are in the later philosophy even symbols for non-existents such as the connection of diversity. They are called diacritical symbols. To be sure, readers are cautioned not to take diacritical symbols as part of his proper ideal language, but I think that Bergmann had for some time thrown off the shackles of the ideal-language method, a positivist hangover. In Realism: A Critique of Brentano and Meinong, at the latest, he had abandoned the indirect linguistic approach to enities though he continues to mark philosophical uses in Moore’s sense by inverted commas. My claim is not at odds with continuing to use symbols. Symbols are there merely a means of conspicuous representation. They are mere illustrations. Bergmann tacitly gave up the view which was a concession to positivism that one can talk about the categorial in the world only indirectly by talking about language. In “Sketch of an Ontological Inventory” he explains that the IL is not indispensable for him. Nevertheless, the later Bergmann makes weighty metaphysical points by considering what is not sayable in an IL and by noting incongruence between what is sayable in an IL and what is thinkable. Comparing Realism and New Foundations of Ontology, it is obvious that there are more symbols in the later book; and this is not only due to its main topics: the forms of molecular facts, analyticity and the foundations of arithmetic. It seems to me that in the later Bergmann the mathematician again comes through. The IL looms large. He even introduces (as he says) more relaxed symbolisms to avoid for the moment the complications of the proper symbolizations of the IL. Moreover, the ontological assays tend to move off from the phenomena to be grounded (he himself speaks of “phenomenological distance”) and move towards mathematical elegance or mathematical custom. To that extent, Bergmann has become more of an ideal-language philosopher.

17 It seems to me also that Bergmann abandoned the simplicity and independence patterns to distinguish the philosophical uses of “is” and “exist” and returned to the ordinary meaning of being there, though he would still separate existence and the so-called “existential quantifier”: there is an x, such that… The ontology of the middle Bergmann has complexes (facts) as well as simples (things) and has turned into a fact ontology. He no longer distinguishes thing and fact ontologies by these patterns. Rather, he defines fact ontologies by the acceptance of nexus and of a distinction between things and facts. Reistic or thing ontologies are characterized by the rejection of nexus. Bergmann’s reconstruction of metaphysics was not only an attempt to arrive at a meaningful metaphysics for himself but also to reconstruct the history of metaphysics and to do what he termed structural history in contrast to scholarly history of philosophy. In his reconstruction of metaphysics Bergmann quite rightly reconstructed the mistake that many philosophers since Plato committed of taking Parmenides’ signs of being for criteria. That mistake was a fact of history and it was pivotal. However, if Parmenides himself would have introduced simplicity and independence as criteria defining being or being proper, he could not have done what he did; namely, argue and even try to prove that what exists has to be simple and independent (of time and change and other entities). Historically the early Bergmann was right that philosophers just indulged in extraordinary uses of key words. Too often, the question of ontologists was that of what is simple or what is independent rather than that of what exists, which is, of course, different from the question of what existence is, which Heidegger found absent in all European philosophy (though wrongly concerning the period when Avicenna was read in the west). An important methodological rule to which Bergmann submits his ontology is the principle of acquaintance, which he adopts from Russell. It requires granting ontological status only to what we are acquainted with. Now, the phenomenalist gives a different answer to the question what we are acquainted with than the realist. The early Bergmann gives a phenomenalist answer and the middle Bergmann an empiricist answer, the later Bergmann an intentionalist answer. The early Bergmann answers: “what we sense”; the middle Bergmann: “what we perceive,” and the late

18 Bergmann: “what we intend.” Thus the late Bergmann holds that we are acquainted with the intention of any kind, any species of act be it a perception, a memory, or a belief or even a doubting. The late Bergmann is no longer an empiricist. Therefore, the principle of acquaintance can no longer play the role it originally played; namely, that of regulating the control of other kinds of cognition by a privileged kind. That kind is privileged because it is considered to be immediate and particularly reliable. In the later Bergmann the principle of acquaintance is superseded by the principle of presentation, which says that whatever can be presented to a mental act of whatever kind exists and vice versa. It still implies some constraints on ontology, e.g., the modes assumed by Bergmann have to be presented by some kind of act though this may not be perception. Any kind of entity assumed must be taken to be presented to at least one species of act by at least one member of the kind. It is not enough that there is a phenomenological datum that can be ontologically grounded by the assumption of a certain entity. One has to defend the claim that we are acquainted with it by some act. That is not always easy as can be gathered from the case of modes of facts in Bergmann’s ontology. The difficulty arose because the claim of being acquainted with the modes of facts had to be compatible with the occurrence of error since the truth of mental acts depends according to Bergmann on them. Where we are presented with the modes of facts we could not be wrong. The middle Bergmann conceded that modes are not given to us while the later Bergmann arrived at the doctrine that the modes of simple analytic facts (i.e., facts that are represented by relatively short logically true or false sentences) are given to us. The middle and the later Bergmann appeals to phenomenological data or to phenomenological rock bottom, but that is not restricted to a particular kind of act. It refers to any feature of what is given, i.e., of what is the intention of any act of whatever kind or what is in the intention of any act of whatever kind.

19

Bare Particulars Methodologically, Bergmann continued from the early Wittgenstein but in his ontological categories he started from Russell. That start was partly positive, partly negative. Bergmann adopted universals and Russellian relations, but criticized Russell’s view of particulars as complexes of universals. Against those complexes he set his famous/infamous bare particulars. He introduces them as individuators of ordinary objects, pointing out that the diversity of two qualitatively undistinguishable ordinary objects cannot be grounded without bare particulars. Bergmann had good reason to speak of an aversion against bareness because the strength of the rejection is far greater than that of the arguments. The rejection of the bare particular arose partly from easily avoidable misunderstandings (such as Wilfrid Sellars taking it to be the individual that has no properties), partly from taking it for granted that we are acquainted with qualities only, and partly from alternative groundings of the diversity of qualitatively identical objects. However, those alternative groundings (namely, relational groundings) seem to presuppose the respective diversity already and hence are question-begging. Bergmann exposed the bare particular to new objections when he realized that categories and the membership in categories need ontological grounds. He started to assume that in each particular there is the subsistent particularity, which is what makes it a particular. That invited the objection that the particular is not bare but has an essence since it is of itself a particular. The point is well taken. However, it seems to be only terminological. One could clarify and specify that “bare” has to be taken to mean “having no ordinary properties of itself.” Obviously, the gist of the category of bare particular is to get away from understanding the phenomena of possession of properties and the standing in relations as inherence, rather than to oppose particular and universal and to analyze the phenomena of the possession of properties and the standing in relations as an external connection. Not in the least was the category of bare particular meant to eliminate those phenomena.

20 If the objection to the bare particular points to Bergmann’s intention to continue the empiricist tradition and to overcome Aristotelianism including Aristotelian essentialism, the reply could be that Bergmann’s categorial essence is still far away from an Aristotelian essence. Bergmann’s essences are residual essences. A more serious difficulty is the mereology of things with subsistents in them. Not only does Bergmann insist that things are nevertheless simple, his analysis seems to imply that categorial subsistents are in things but do not exhaust them, if only because there is always another subsistent in a thing; namely, existence. Moreover, it would also be out of the question to hold that the categorial subsistent and existence together form the thing and exhaust the thing. The difficulty of the half-baked, i.e., incomplete, mereology of the things in Realism is removed by the Two-in-Ones1 of this later ontology. The things, individuals or universals, are composed of an item (a mere individuator) and an ultimate sort (the categorial determination), which form a Two-in-One together, and of nothing else. In the later Bergmann the items are bare individuators while the particulars, which are Two-inOnes, contain a categorial essence. The items are the new bare particulars and they are the strictly bare particulars as they involve no essences. The categorial essences are, of course, the ultimate sorts. There was another semantic field in which Bergmann wanted to place his particulars by the term ‘bare particular;’ namely, the opposition between bare entities (particulars) and natured entities (universals). That is what he said. But what he meant was the opposition between particulars as individuators and universals as natures. What Bergmann implied by what he said was that universals needed no individuators or needed not to be individuators. He mixed up what W. E. Johnson called otherness on the one hand and difference on the other hand, otherness being synonymous with Bergmann’s diversity. That difference is not enough to individuate universals such as carmine can be pointed out easily. It suffices to ground the diversity of carmine and other shades of color but not to ground its diversity from sounds or smells or even from particulars such as the 1 By numerals with capitals Bergmann does not refer to ordinary natural numbers because his ontological assay of the number 2 is not applicable to the Twoness of Two-in-Ones. The latter are too fundamental.

21 carmine spot or rather from what individuates the carmine spot. The problem of the individuation of an entity e is the problem of grounding the diversity of e from all other entities. In New Foundations of Ontology, that mistake (the confusion of diversity and difference) is corrected. Not only particulars but also universals involve an item, a pure individuator. What is so admirable about Bergmann’s posthumous opus and about the later Bergmann in general is that he corrected so many deep errors and saw and addressed so many difficulties in his earlier views. He always continued to “dig deeper,” as he used to say, according to Panayot Butchvarov’s report. Intentionality Influenced by G.E. Moore and in opposition to the views of Hume and logical empiricism, including Wittgenstein, Bergmann reintroduced mental acts and intentionality. He began his ontological analysis of intentionality by categorizing it as a logical relation like the logical connectives and he held that statements of intentionality are analytic in the sense of being logically true. Bergmann thus extends the range of the analytic in a narrow sense, a move that has an important ontological advantage. Logical relations do not in general require the existence of all relata. Thus the problem of non-existent objects of knowledge in cases of error or imagination got solved. That was also the main point in Brentano’s theory of relations with respect to intentionality. Bergmann refers to the logical relation of disjunction. A disjunctive sentence is true even if one disjunct is false. Thus the logical relation of disjunction holds even if one relatum does not exist. The ontology of the intentional connection must allow for the non-existence of a term because of the phenomena of error, illusion, and imagination. Although Bergmann’s ontology of intentionality allows that, one of his reasons for his initial rejection of facts was that one would have to accept existing false facts as relata of the intentional relation. Even when he accepts facts, he still brings the intentional connection into the same category as the logical connectives, or rather as what he supposes these to represent. This is because he wants to do justice to the closeness of the connection between act and object, which he takes to be a

22 phenomenological datum. Bergmann grounds that closeness partly in the necessity of intentional facts; and for them to be analytically necessary facts, the connection in them has to be logical. The middle Bergmann holds that the intentional relation, the intentional nexus, always connects with facts rather than with things and that for all acts a fact exists to which the intentional nexus connects even if it is wrong. Thus the second relatum of the intentional relation/nexus is always a fact. The first relatum is according to Bergmann not the mental act as a whole but a universal involved in it, which he originally called “thought” and which plays a role similar to the Meinongian content of a mental act. In the later ontology of Bergmann, the intentional and the logical connections have parted company. Not only do they belong now to different categories; the intentional connection is no longer an existent. There is only the diad consisting of the content (now named “text”) and the intended fact. Bergmann is quite happy with getting rid of the intentional nexus since it is not only causally ineffective (which he takes to be incompatible with being a relational universal) but since it has a beginning in time (with the appearance of mind in the evolution of life). He reports that he was always reluctant to grant the intentional connection the same ontological status and membership in the same category as exemplification and the logical connectives. The intentional connection is pivotal to Bergmann’s epistemological realism. Not only does it allow him to advocate a correspondence theory of truth without any implication or suggestion of similarity between mental acts and their objects (this is definitely excluded by his assumption that the contents of mental acts are simple while their objects are complex), it furnishes a connection that is both very close (analytic) and specific to mental acts and their objects and thus averts skepticism and justifies the realist claim that we know the world. That demonstrates that the key to realism is ontological. Some of Bergmann’s statements sound as if he could gear and did gear his whole ontology to the aim of supporting epistemological realism. However, to develop and articulate a consistent and coherent ontological system is such a difficult and daunting task (the best evidence for that is Bergmann’s own unending toil) that it cannot be subordinated to another goal. I think the order has to be reversed. Support for epistemological realism is the reward for excellent ontological work and idealism the

23 punishment for bad. Idealism started in the Middle Ages in Ockham who dodged ontological problems and offered epistemological solutions for them. Aristotelian metaphysics even in the watered-down version of Ockham does not allow one to ground complexity; hence, he infers that the complexity given to us is a contribution of mind, that it is a concept without foundation in reality. What corresponds to complex and relational concepts are merely collections of things. From there the way is not long to the conclusion that all categories are mental functions and that the whole world is the result of mental unification and construction. The second scandal of Bergmann’s ontology, greater still than the bare particular, is the potential fact, i.e., the fact with the mode of potentiality. It plays an important role in has assay of intentionality, providing intentions for acts with false contents. However, this is not the only role it plays and that justifies its assumption. It is also designed to ground ontologically the distinction between sentences including false ones and nonsensical strings of signs of the ideal language. It is also designed to ground the well-formedness of sentences. Providing an intention for false contents, providing a second relatum for the intentional nexus in those cases of wrong acts, could not have been Bergmann’s reason to introduce potential facts. As was mentioned already, in the middle Bergmann the nexus was designed and categorized as it was in order to allow for cases where the second relatum (the intention) is non-existent, as we usually say. In the ontology of the late Bergmann, an existing intention is indeed always indispensable. Intentionality is grounded on a diad and diads cannot stand on one leg only. Bergmann characterizes diads as internal relations. However, potential facts are introduced by the middle Bergmann. The deepest and the main reason for the middle Bergmann to acknowledge potential facts is an insight about the connection between thinking in the generic sense and existence, which Bergmann attributes to Descartes. It is the insight that if we are presented with something we are presented with its existence and that the only reason for us to hold that something exists is that it is presented to us. This implies that what we think (in the generic sense encompassing all species of mental acts) exists. Since the intention exists even when we are wrong, error has to be accounted for by something else than existence. Bergmann accounts for

24 error by the mode of potentiality. He assumes that each fact has a mode, either actuality or potentiality. Modes are traditional categories that have to be distinguished from modalities with which they have been mixed up by Kant. Facts as Complexes When Bergmann finally accepted facts, he introduced them by arguments that he had applied previously to show that there must be a nexus of exemplification, namely that there has to be an entity creating unity and connectedness. The middle Bergmann assumes that in addition to the things and the nexus there is the circumstance that the nexus connects the things together. That is the fact, and the things and the exemplification nexus are its constituents. It seems inescapable to ground the individuation of complexes on their constituents. Bergmann therefore lays down what he calls the fundamental principle of ontology: “‘Two’ entities yielding literally the same assay are literally, or, as one says, numerically, one and not two.”2 Note that Bergmann does not use the word “constituent” in this principle. He does not formulate: “Two” entities having the same constituents are literally one and not two. There might be entities other than constituents involved in a complex. The fundamental principle says that the sameness and by implication also the diversity of complexes depend on all entities that are needed in their assays. An assay produces a list that represents, according to Bergmann, a collection of entities. He emphasizes as a crucial ontological insight that a collection is not an entity. A collection corresponds to what is customarily called a set. That is to a certain extent an ontological criticism of set theory but does not call it explicitly into question if only because Bergmann also has classes in his ontology, which are derived universals in the middle Bergmann. He does not consider grounding arithmetic ontologically on anything other than classes. Now, the role of collections has led critics to attack the whole enterprise of ontologically assaying facts. They suggest that assaying facts 2 Bergmann (1967, p. 22).

25 assimilates facts to sets as it produces a list and they argue in the holist way that facts cannot be analyzed, that analyzing them is to reduce them to sets, and that facts are irreducible. Bergmann would certainly agree that facts are irreducible. Moreover, his assay of facts is not meant to be a reduction to something else. Finally, if facts are proper complexes then their parts are present in them, i.e., they are not merely potential parts which are realized when the whole falls apart as in Aristotle’s view. And if the parts are present in the whole, then an ontologist has to indicate and characterize those parts. That seems necessary even if one does not want to ground the sameness and diversity of facts on their assays. The alternative of grounding the sameness and diversity of facts on their assays would be to ground it on a sameness and/or a diversity connection. The later Bergmann could draw on a diversity diad and its negation to individuate facts and on its negation to ground their sameness. By doing so he could have avoided the problem of order in facts, which he did discover in the 1970’s and which surfaced due to the increase in precision brought about by the fundamental principle. One could argue though that the problem of order is not avoidable in that way because it concerns not merely diversity but also difference in a respect; namely, with respect to order which an be regarded as a phenomenological datum. The view of complexity and of facts of the later Bergmann is very different from that of the middle Bergmann. Not only does the later Bergmann advocate at least two kinds of complexes (facts and circumstances) while the middle Bergmann accepts only facts as complexes, the structure of facts and of the other complexes is very different. The middle Bergmann conceives of complexes as connected collections. Complexes have unity by virtue of connectors called nexus. In the ontology of the later Bergmann, there are no connectors, either in facts or in circumstances. What then holds the facts at core there together? It is the internal relation of diversity. At the core of each atomic fact there is a diad of diversity. That seems paradoxical since we usually think of diversity as separating the diverse entities. Nevertheless, diversity is a connection insofar as it involves two entities. It separates and holds together “at the same time.” In atomic facts the diads of diversity hold together things (particulars and universals), in molecular facts it holds

26 together facts. The (Meinongian, not Fregean) functions (makers) formerly categorized as nexus by the middle Bergmann cling to the diads, as the later Bergmann says. What the metaphor of clinging represents is a dependence, a dependence of the function on the diad to which it clings and which it makes a fact (the diad without the function would be a circumstance). The fact does not build on the diad and the function as the diad builds on the things. The function and diad do not go together per se, not even to form a diad of diversity, since the function is not a determinate and only determinates are, according to the later Bergmann, independent enough to form a diad of diversity. The connection between function and diad is based on the unsaturatedness of the function, to use a Fregean notion although the function is not a Fregean function. Thus the complexity of facts involves and builds on the complexity of circumstances. It also involves another kind of complexity; namely, that of things as Two-in-Ones. Bergmann does not want to call them complexes though. He continues to insist that things are simple. Nevertheless, he takes each thing to be composed of an item and an ultimate sort. Likewise the middle Bergmann insisted on the simplicity of things although he assumed that a categorial subsistent and existence is “in” each thing. Actually, there is a serious impediment in the ontology of the later Bergmann to the conceiving of things as complexes. Items and ultimate sorts are not determinates and therefore do not form diads of diversity. They lack diversity. That is presumably also why Bergmann changed his mind concerning the question whether exemplification, negation, conjunction etc. are constituents of facts; that is why he later did not continue to count functions (the former subsistents) among constituents. In the ontology of the later Bergmann conjunction, disjunction and the conditional are, of course, no longer nexus. Rather, they are functions, which cling to what is already connected; namely, diads, diads of facts. They are not nexus and they do not need a nexus to be connected to the diads since they are relatively dependent entities. The connection between the constituents of a diad is not due to some kind of dependence though. All constituents of diads are determinates and determinates are relatively independent existents. However, for any two determinates there is, as Bergmann says, eo ipso their diad.

27 If one is dissatisfied with connections that ground on the dependence of functions or the eo ipso production of diads, one might propose to ground the connection between the constituents of a diad (e.g., the particular a and the universal u or the facts f and f' ) and also the connection of the function and the diad to which it clings (the conjunction function and the diad of the diad of the fact f and the diad of facts f and f', e.g., which has also to ground the order of the conjuncts) on the respective canons (as Bergmann calls them) that rule the connection of entities into further entities and the correspond to the syntactical rules of the ideal language. The proposal would be misplaced because canons are not existents. Canons are some kind of laws, and laws are assayed usually as general facts by the middle and the later Bergmann. However, canons cannot be general facts because these range over determinates only and functions are not determinates. But they would have to range over functions, of course, if the proposal would be realized. Canons cannot be facts for the simple reason that they determine which complexes are facts and which not. The canons serve Bergmann mainly to analyze the structure of molecular facts, which can be very complicated and rise to a high hierachical structure. Immediately each fact, molecular or atomic, consists of one diad to which a function clings, but the ultimate foundation, as Bergmann calls it, is a collection of particulars, universals and functions only. In the middle Bergmann an atomic facts builds, so to speak, on a collection including things and one nexus, and it is nothing but the connecting of these things by the nexus. In the later Bergmann the things connect themselves together by virtue of their diversity, and the exemplification function binds itself to the diad. Critics of the later Bergmann have asked why he kept exemplification if it is no longer a nexus and why he does not turn the universals into functions like Frege. The answer to be gathered from the published texts is that the exemplification function has to ground the difference between atomic facts and the various kinds of molecular facts. Moreover, Bergmann distances himself from Fregean functions and characterizes his functions as M-functions (derived from Meinongian functions). Hence the reversion of views in the later Bergmann is not as

28 great as it sounds when one remembers the opposition between fact and function ontologies and the rejection of function ontologies in Realism. In Realism already exemplification, conjunction etc., categorized as nexus, are unsaturated, dependent entities that served to ground their connection with things or facts. It was also designed to avoid Bradley’s regress. Bradley’s regress targets the view that complexes have constituents and are connected by connectors. Bradley argued that if a complex is connected by a connector, there has to be another connector to connect the connector. The middle Bergmann meets Bradley’s challenge in Realism by drawing on his category of subsistents, which is the third category besides that of things and facts. Things and facts are independent and need nexus to be connected to other things and facts. By contrast, subsistents including nexus are dependent entities and therefore need no nexus to be connected. They connect others and themselves by virtue of their dependence. In the ontology of the later Bergmann Bradley’s regress cannot start because there are no connectors in it. The Later Bergmann I agree with Panayot Butchvarov that the new phase Bergmann’s philosophy entered after the publication of Realism is still more valuable.3 However, Bergmann’s later philosophy has a hard time, not least among the friends of his middle and earlier philosophy. No wonder, since Bergmann abandoned several central views that he had advocated vehemently before. In the middle Bergmann the nexus is pivotal. As was mentioned already, the later Bergmann has dropped it completely. The middle Bergmann holds that sameness and diversity are primary, i.e., that they need no ontological ground apart from and in addition to what is diverse and what is the same. In the latter Bergmann there are diads of diversity. However, already in Realism Bergmann is on the way of giving up that view. He countenances identity as a subsistent (which he had earlier and later called Leibniz-Russell Identity), connecting, and hence as an existent in addition to, the identical things. He needs the subsistent identity for his assay of classes as derived universals. It seems to have 3 Butchvarov (1987).

29 been his assay of classes and the attempt to furnish a foundation of arithmetic which drove the later development of Bergmann’s ontology (see the Presidential Address titled “Diversity”). For that purpose he introduced for a short time a connecting entity of diversity until he arrived at the diads of diversity. He had to do something about his assay of classes as derived universals, I suppose, because he got doubts about derived universals in general due to a difficulty with such universals raised by Reinhardt Grossmann in a paper of 1972 and originally discovered by F. P. Ramsey.4 Therefore, Bergmann looked for a new assay of classes, and he expected it to involve identity. The question was whether to transform the subsistent identity into a function like the other so-called logical relations. It seems that he took the chance of clarifying the relationship between the subsistent identity and primary sameness. In the mature ontology of the later Bergmann there are only the diads of diversity, and sameness is the negation of diversity. The price Bergmann has to pay for grounding sameness and diversity on existents; namely, on a a circumstance (the diad) and a fact (the negation of the diad), is that the range of diversity gets restricted. Therefore, the individuation of only a part of the existents can be based on that entity diversity. Only the individuation of determinates can be based on it. However, what the medieval philosophers held is very plausible namely that each existent is individuated. Thus, the later Bergmann gets into the difficulty of individuating non-determinates. Most offence was taken at the Two-in-Ones and the circumstances in Bergmann’s later ontology. Some felt that the limits of ontological analysis had been crossed over by it.5 They object that the analogy of spatial wholes and parts which governs ontological analysis has been overstretched by the later Bergmann. It seems to me that they have too simple an idea of spatial complexes. Such complexes are viewed by them as sets or wholes in terms of the mereological calculus. Moreover, any analogy goes hand in hand with disanalogies. Hence, I think there is no difficulty with assuming that different entities are involved in different ways in a complex, that some entities form internally connected pairs in 4 Grossmann (1972). 5 Butchvarov (1974), “The Limits of Ontological Analysis.”

30 them, that some entities are connected by their independence to such pairs, and that some entities are extremely closely connected to the complex as a whole (the modes). Bergmann thus distinguishes different degrees and different kinds of connections. Mainly, I think the variety of entities and connections in the ontology of the later Bergmann is due to the phenomena he wants to ground and it is entailed by his ontological theory building rather than by some spatial analogy or by some preconceived method or approach such as logical atomism. To see that, look at the ontology of Reinhardt Grossmann, the great alternative. Grossmann continues as much the logical atomist tradition as Bergmann does. Nevertheless, his categorial properties are ordinary universals and any complexity is based on relations as connectors. And this is not as weak a position as some think who object that the necessity of categorial properties is not grounded and cannot be grounded in Grossmann’s ontology. Grossmann grounds all necessity on general facts which are laws. Facts which logically follow from some law or laws are taken to be necessary. Grossmann could therefore refer to laws that logically imply that an entity has the categorial property it has, e. g., the law that every entity has exactly one of the categorial properties. The reason why Grossmann does not have different kinds and degrees of connection is that he decided not to make the distinction between the form and the content of the world, a distinction that looms so large in Bergmann.

REFERENCES Bergmann, Gustav. 1967. Realism: A Critique of Brentano and Meinong. Madison: The University of Wisconsin Press. Bergmann, Gustav. 1992. New Foundations for Ontology. Madison: The University of Wisconsin Press Butchvarov, Panayot. 1974. “The Limits of Ontological Analysis” in M. Gram and E. Klemke (eds.) The Ontological Turn: Studies in the Philosophy of Gustav Bergmann, Iowa City: University of Iowa Press. Butchvarov, Panayot. 1987. “In Remembrance of Gustav Bergmann.” Grossmann, Reinhardt. 1972. “Russell’s Paradox and Complex Properties,” Nous, 6 Grossmann, Reinhardt. 1983. The Categorial Structure of the World. Bloomington: Indiana University Press

The First Station of Gustav Bergmann’s Odyssey GUIDO BONINO Università di Torino

A Dispute of Long Ago

T

he title of the paper makes an obvious reference to Herbert Hochberg’s essay “From Carnap’s Vienna to Meinong’s Graz: Gustav Bergmann’s Ontological Odyssey,”1 in which the development of Bergmann’s philosophical views is traced from its positivistic beginning to its ontological conclusion. My aim is that of throwing some light on the very first episode of such a development, an episode that announced Bergmann’s first significant departure from the views of Carnap and of orthodox positivism. The departure is that brought about by “Pure Semantics, Sentences, and Propositions”2 and “A Positivistic Metaphysics of Consciousness,”3 two papers that Bergmann published in Mind in 1944 and 1945 respectively, and that bear witness to Bergmann’s first attempts at escaping from his earlier views and at the same time caused irritation and hostility among his old positivistic friends. A rich anecdotage—partly fostered by Bergmann himself—flourished concerning Carnap’s irritated reactions to the articles in question. The history of the events that preceded and followed the publication of the articles can be reconstructed with the aid of the letters kept in the Bergmann archives of Special Collections in the University of Iowa Library. In 1942 Carnap’s Introduction to Semantics4 had been published. Bergmann wrote the first version of “Pure Semantics, Sentences, and Propositions” as a sort of critical commentary on some of the issues dealt Hochberg (1994). Bergmann (1944). 3 Bergmann (1945). 4 Carnap (1942). 1 2

32 with in that book, and in the fall of 1942 he sent it to Alonzo Church, who was then the editor of the Journal of Symbolic Logic. Whether out of prudence, or out of deference, in any case the title with which Bergmann sent the paper to Church was “Remarks Concerning the Terminology of Pure Semantics.” Such a deliberately minimalistic title—which alludes only to terminological questions—was probably intended not to displease Carnap, who was after all the target of some criticisms, but was also at the origin of a misunderstanding. In fact, not only was the title minimalistic, but in the text as well Bergmann put forth his argument as if it were purely or mainly terminological. In short, Bergmann proposed to substitute ‘sentence’ for ‘proposition’ in some passages of Carnap’s book. Since ‘sentence’ is usually used as referring to linguistic entities—and that is the way in which Bergmann used it too—, whereas ‘proposition’ was used as referring to non-linguistic entities, and more specifically to the entities referred to by sentences, Bergmann’s proposal must have seemed rather odd. At first Church simply thought that Bergmann must have misunderstood Carnap’s intentions. In a letter to Bergmann of October 31, 1942, he wrote: From a first superficial reading, I get the impression that you must somehow have misunderstood Carnap’s intention in the use of the word ‘proposition.’ As I understand him, ‘proposition’ does not refer to any sort of syntactical entities (in any language) but is rather approximately 5 synonymous with his older term ‘state of affairs.’

After some clarifications provided by Bergmann, Church came to understand that the proposal concerning the substitution of ‘sentence’ for ‘proposition’ was not simply due to a misunderstanding, but was rather conceived of as a deliberate revision of Carnap’s terminology: Your letter clarifies one point in regard to your paper, namely that where I had taken you as misunderstanding Carnap’s intended meaning of

Letter of Alonzo Church to Gustav Bergmann, October 31, 1942, Papers of Gustav Bergmann, Special Collections Department, University of Iowa Library, Iowa City, Iowa, IX A 1 – 25/1. 5

33 ‘proposition,’ you were rather stating a proposed revision of Carnap’s 6 usage.

Bergmann’s cautious wording in the article may at least in part explain Church’s false impression. But Bergmann had sent the article to Carnap as well, who apparently found it very difficult to understand: I spent a good deal of time in studying carefully your note on semantics. Many sentences I read three times or more without understanding them. The greater part of your discussion is still obscure to me and some sentences are outright Chinese to me. You understand, this does not mean that you have no point here or that the arguments for your point are wrong. What I say is only that I do not clearly see what you mean and 7 therefore cannot say whether I agree or not.

Anyone who has read “Pure Semantics, Sentences, and Propositions” might sympathize with Carnap: the difficulty and obscurity of Bergmann’s early articles have always been a commonplace. In any case, on December 11, 1942, Church sent Bergmann a letter that included the report of an anonymous referee, whose evaluation was unfavorable to the paper: Despite careful study of Bergmann’s paper, I find the central point which the author tries to make quite incomprehensible. It does seem to me, however, that a criticism of Carnap’s own formulation is intended, rather than—as you thought possible—of a proposed restatement of the latter...In my opinion, Bergmann’s paper is not suited for publication in 8 the Journal.

Church declared his agreement with the referee:

Letter of Alonzo Church to Gustav Bergmann, November 7, 1942, Papers of Gustav Bergmann, Special Collections Department, University of Iowa Library, Iowa City, Iowa, IX A 1 – 25/1. 7 Letter of Rudolf Carnap to Gustav Bergmann, November 6, 1942, Papers of Gustav Bergmann, Special Collections Department, University of Iowa Library, Iowa City, Iowa, IX A 1 – 25/1. 8 Referee’s report on “Remarks concerning the terminology of pure semantics” by Gustav Bergmann, Papers of Gustav Bergmann, Special Collections Department, University of Iowa Library, Iowa City, Iowa, IX A 1 – 25/1. 6

34 I have the enclosed report from the referee of your paper. Since receiving it I have read your paper again with care myself, and find that I 9 am forced to agree fully with the referee.

Maybe in the meanwhile some other letters were exchanged between Bergmann and Church, Bergmann and Carnap, and perhaps also between Church and Carnap. Carnap’s understanding of the paper was not improved by Bergmann’s elucidations, as is revealed by another letter, whose tone—to say the least—was not very tactful: From your manuscript and your various letters I cannot help getting the impression that you are not yet clear yourself about the conception you wish to explain. I think it would be more advisable, before you start writing another paper on the same question, to wait until you succeed in 10 making it clear to yourself.

Bergmann always considered the rejection of his paper to be the result of pressure put on Church by Carnap and his positivist clique, or of some hostile plot. Thus—with a shrewd political move—he decided to resort to a completely different quarter, i.e., Gilbert Ryle, then editor of Mind. Of course Bergmann’s views were as remote from Ryle’s as it is possible for philosophical views to be, but probably he thought that Ryle could be favorably disposed to his paper out of hostility toward Carnap. Be that as it may, the paper was accepted, and published in 1944. The following year Mind published “A Positivistic Metaphysics of Consciousness,” which in many aspects is a sort of second installment of the previous article. Three other papers are significant in connection with the whole issue. One of these is the second part of “The Extra-Linguistic Reference of Language,” titled “Designation of the Object Language,”11 by Everett W. Hall, who was then a colleague of Bergmann in the Department of Philosophy at the University of Iowa, published in Mind in 1944. In that Letter of Alonzo Church to Gustav Bergmann, December 11, 1942, Papers of Gustav Bergmann, Special Collections Department, University of Iowa Library, Iowa City, Iowa, IX A 1 – 25/1. 10 Letter of Rudolf Carnap to Gustav Bergmann, January 4, 1943, Papers of Gustav Bergmann, Special Collections Department, University of Iowa Library, Iowa City, Iowa, IX A 1 – 25/1. 11 Hall (1944). 9

35 article Hall raised some objections to Carnap’s conception of semantics, dealing with questions very close to those dealt with by Bergmann in his own critique. Then there is Carnap’s rejoinder to Hall and Bergmann’s criticisms, “Hall and Bergmann on Semantics,”12 also published in Mind, in 1945. And finally came an intervention by Max Black, who was in correspondence with Bergmann in those years and privately encouraged him in his dispute with Carnap. In fact, in 1945 Black published in Mind his own review13 of Carnap’s Introduction to Semantics; no reference is made in it to Bergmann’s criticisms, yet some objections are raised that can be likened to Bergmann’s. It is difficult to evaluate Bergmann’s views in 1944 and 1945 from a theoretical perspective, because of the intrinsically confused character of some of his remarks and the programmatic nature of his proposals, which were not fully developed. Therefore, what I will try to do is simply to disentangle some of the issues involved in the discussion between Bergmann and Carnap. Such a discussion has very seldom been subjected to close scrutiny. The one exception is the aforementioned paper by Hochberg and the first pages of his book The Positivist and the Ontologist;14 since some of the questions involved have already been dealt with by Hochberg, I am going to pay comparatively little attention to such questions, and devote the main part of my analysis to other questions. I do not aim at making Bergmann’s views in 1944-1945 truly clear—and least of all at assessing their value vis-à-vis those of Carnap—, which is possibly a hopeless task, but only at making them a little clearer, trying to make explicit the connections with the discussions current in those days. In fact it seems to me that the whole issue is interesting both in the light of Bergmann’s later philosophical development, and from a more general historical point of view, since it involves many questions widely debated in the context of late logical positivism.

Carnap (1945). Black (1945). 14 Hochberg (2001); cf. especially pp. 2-10. 12 13

36 Realism and Positivism Let us come now to the content of the two articles by Bergmann. As has already been hinted at with reference to the first one, they are rather difficult and obscure. Though never explicitly disavowing them, Bergmann came to confess that at least they had been written badly. One can find in them several different themes, closely intertwined, but the most prominent ones seem to be the following: (i) a criticism of Carnap’s views concerning semantics; (ii) a new analysis of the notion of aboutness, and more generally of consciousness; (iii) a tentative formulation of what would become the ideal-language method. The criticism of Carnap’s views is to be found in “Pure Semantics, Sentences, and Propositions,” the analysis of consciousness in “A New Metaphysics of Consciousness,” the tentative formulation of the method in both of them. Some preliminary remarks on (i) and (ii) are appropriate, though a more adequate examination of both questions is to be postponed until the third is taken up. In the light of Bergmann’s later philosophical views, his criticism of Carnap’s Introduction to Semantics may seem paradoxical, for he charged Carnap with realism. More exactly, he thinks that the idea of a relation of designation holding between singular terms, predicates and sentences on the one hand, and individuals, properties and propositions on the other commits Carnap to a realistic metaphysical position, and to the abandonment of the positivistic attitude. It is not easy to understand Bergmann’s criticism in detail, since it is concealed—as has already been said—under the appearance of a proposal for a terminological revision. Nor is it immediately clear what Bergmann means when he speaks of “realism” in connection with Carnap. Some suggestions may come from the then current debate on the issue of positivism and realism. Only two years later Roy Wood Sellars published an article, “Positivism and Materialism,”15 which is referred to by Bergmann in another paper,16 and which—notwithstanding the rather “unpositivistic” muddle-headedness of the author—can throw some light on the question. In this article Sellars argues for realism—of which ‘materialism’ is significantly a synonym, or 15 16

Sellars (1947). Bergmann (1947a).

37 of which at least it indicates a variety—, and criticizes positivism. Positivism, as opposed to realism, is simply conceived of as a sort of phenomenalism (or even idealism17), i.e., a position according to which the external existence of physical objects cannot be admitted in a literal sense. Realism, on the other hand, is the position that overtly recognizes such an external existence. Of course all that sounds muddle-headed to positivistic ears, and such a characterization cannot be attributed to Bergmann in its original form, if for no other reason because it is at least misleading and at worst it tries to say what cannot be said according to Wittgensteinian interdictions. Yet it provides a rough idea of what people—Bergmann included—had in mind in those years when they were speaking of positivism and realism. Such a suggestion is indirectly confirmed by a paper written by L.O. Pinsky in 1954, “Positivism and Realism.”18 Pinsky was a student of Bergmann and the paper I am referring to is one of his first writings. As is well known, Bergmann’s students generally were not invited to set forth opinions opposing those of Bergmann himself, so that some of the early papers of Bergmann’s students can be assumed as indirect evidence of Bergmann’s own opinions. Furthermore, Pinsky himself is eager to make it clear in a footnote that his views are indebted to those of his teacher. In 1954 Bergmann’s departure from Carnap’s views had already gone to a considerable length, immeasurably more than ten years before, but some of the criticisms do not seem to have changed very much. “Positivism and Realism” aims at a criticism of Schlick’s and Carnap’s “realistic” presuppositions; such presuppositions, which according to Pinsky (and Bergmann, whose claims in his own papers of those years are rather similar) are the source of several other errors, can ultimately be reduced to an uncritical acceptance of the existence of an external reality (the “real world,” spoken of in a physicalistic language), whose structure is shared by all the private experiences. The question of the relationship of this derogatory meaning of “realism” with the complimentary one used by Bergmann with reference to his own views is rather complex, and will not be investigated here. However, in 1944 Bergmann could not accept Sellars’ overtly metaphysical language, or—for 17 18

The profusion of isms is Sellars’, not mine. Pinsky (1954).

38 that matter—Pinsky’s, which is also Bergmann’s in 1954. In 1944 the whole discussion with Carnap is carried out by means of an analysis of the purpose and the correct method of pure semantics, and by making reference to Wittgenstein’s distinction between what can be said and what can only show itself. The whole question will be resumed in the context of the examination of the third point. In “A Positivistic Metaphysics of Consciousness,” Bergmann elaborated a suggestion put forth in the paper of the previous year, according to which the complex hierarchy of languages and metalanguages developed in order to deal with semantical questions can be further enriched and can be used to resolve some philosophical problems concerning the notions of intentionality and of the self. Bergmann’s account both of the problem and of the suggested solution is very convoluted, and it is not easy to understand the relationship between this proposal and his later views. In a sense—as in the case of the criticism of Carnap’s realism—the analysis of the notion of intentionality—or aboutness—put forth in “A Positivistic Metaphysics of Consciousness” seems rather ironic if it is seen in the light of Bergmann’s later views, characterized by the acceptance of a primitive notion of intentionality (the nexus M). In fact in the 1945 paper Bergmann rejected any notion of aboutness, which was regarded as an “illusion spawned by language,” to use Hochberg’s apt formulation.19 Aboutness is neither a “natural” relation—such as to the left of—nor a logical or linguistic relation (this point will be resumed later). Yet, there seems to be a sense in which the suggestions put forward in 1944 and 1945 prepared the way for Bergmann’s later views. In a footnote to “Intentionality,” the paper in which the idea of a basic intentional nexus is explicitly formulated for the first time, Bergmann mentioned his old papers as forerunners of such an idea: “The fundamental ideas of this essay are first stated, very badly, in two papers that appeared over a decade ago: ‘Pure Semantics, Sentences, and Propositions’ [and] ‘A Positivistic Metaphysics of Consciousness’.”20 How is that possible, if one of the claims of those papers was that of denying any reality to aboutness? It seems to me that a possible answer is 19 20

Cf. Hochberg (2001), p. 2. Bergmann (1955), p. 5, footnote.

39 that in those papers Bergmann tried to develop—though rather confusedly—the notion of act, as something distinct from content, and this idea is certainly fundamental for his later views. Some Misunderstandings But now time has come to move to the third question, i.e., the tentative formulation of the ideal-language method, which takes place in the context of a criticism of Carnap’s conception of semantics. In fact, as has already been said, the main aim of “Pure Semantics, Sentences, and Propositions” is that of criticizing Carnap’s conception of semantics as it is put forth in Introduction to Semantics. This aim is in part concealed by the fact that Bergmann presented his criticism as a proposal for terminological rectification. The following is what exactly happened. Bergmann reads Carnap’s Introduction to Semantics, which makes use of a certain terminology; such a terminology, if taken at its face value, commits Carnap to a certain view—let us call it A—, which Bergmann finds objectionable. Thus Bergmann—maybe out of prudence, or out of deference, or less probably because he sincerely believes that Carnap cannot really hold A— proposes a change in the terminology, which, if accepted, would turn A into a different view—let us call it B—; B is the view Bergmann himself regards as correct. Therefore what Bergmann puts forth as an interpretation of Carnap’s real intentions, hidden under a misleading terminology, can—and perhaps should—be read as a critical assessment of Carnap’s own views and a proposal for a different view. As has already been hinted, a criticism very similar to Bergmann’s is set forth by Everett W. Hall in “Designation of the Object Language.” But Hall’s criticism is characterized, if compared to Bergmann’s, by a curious twist. Whereas Bergmann finds A objectionable, and regards B as correct, Hall thinks the opposite way: A is correct, and B is incorrect. But then, since Carnap’s view seems to be A (Bergmann’s tendentious interpretation notwithstanding), why does Hall feel himself compelled to criticize him at all? The fact is that Hall thinks that Carnap’s real view is B, since that is the only view that fits with other important tenets of Carnap’s doctrine. Thus, even if Carnap’s phrasing may seem to support A, his real view must be B, and B is—according to Hall—incorrect. The result of these two

40 creative interpretations is that Hall and Bergmann, who hold opposite views, agree, up to a certain point, in their attitude toward Carnap. Both of them criticize him, though in different ways: Hall, from his own interpretative point of view, for substantial reasons, Bergmann—at least if his paper is taken at face value—for a misleading terminological choice. Moreover both of them attribute to Carnap the view B, even though in Bergmann’s case that seems to be merely a rhetorical device. The tangle is made still more inextricable by Carnap’s reply to Hall’s and Bergmann’s criticisms. In fact Carnap lumps the two criticisms together, accusing both his opponents of misunderstanding his own views: he protests that his assertions must be understood in a literal way, and that he is truly maintaining A, not B. That of course falls short of being an appropriate response to Bergmann’s remarks, since Bergmann’s real point does not concern the interpretation of Carnap’s writings, but the choice between A and B. All this may seem rather abstract, since nothing has been said until now about A and B themselves, but it has been necessary in order to clear the ground in such a tortuous debate. Before taking up the real issue, one more remark is appropriate. A is what Bergmann calls a realist view, and that is the reason why he cannot accept it. B is what Hall considers a view characterized by the “linguacentric predicament,” and that Bergmann would regard as a truly positivistic conception. What is Semantics? The first sections of “Pure Semantics, Sentences, and Propositions” are devoted to a tendentious exposition of Carnap’s conception of pure semantics. The exposition is tendentious in the sense that what Bergmann really expounds is his own conception of pure semantics. At the end of such an exposition, it is all too natural to propose some corrections to Carnap’s original terminology, which does not fit with Bergmann’s views. Holding to what he considers a positivistic attitude, Bergmann embraced Wittgenstein’s distinction between saying and showing, and the consequent views about ineffability. As was common among logical positivists, Bergmann found a way out of Wittgenstein’s silence thanks to a solution first suggested by Russell in his Introduction to the Tractatus

41 Logico-Philosophicus. What cannot be said—but only shows itself—in a language L can be said in its metalanguage L´; of course there are things which cannot be said in L´, but they can be said in L´´; and so on… In particular, each language cannot speak about its own semantics. For example, let us assume that in the language L there occurs the name ‘a’, which refers to Mont Blanc; in L ‘a’ refers to the Mont Blanc, but it cannot say that it does so. Our success in saying so depends on the fact that we did not use the language L, but a metalanguage—which in this case is English—to speak about it; and the symbol ‘ ‘a’ ’ which occurs in the metalinguistic assertion does not belong to L, but is a name, belonging to the metalanguage, of a symbol of L (of course the ‘a’ surrounded by two single quotes belongs to a meta-metalanguage). In short, we can speak of the interpretation of a language L only in its metalanguage M, but M cannot speak of its own interpretation. In order to convey this idea Bergmann has recourse to a metaphor according to which any interpretation is completely “extraneous” to the language of which it is an interpretation. In The Logical Syntax of Language21 Carnap had made use of such a device in order to speak about syntax. In the same writing he had also drawn the famous distinction between the material and the formal mode of speech. According to Carnap, S1 is called a sentence in the material mode of speech if it asserts the property of an object when there is a syntactical property that belongs to a designation of that object if and only if the original property belongs to the object in question. In the formal mode of speech, which takes place in a syntactical metalanguage, no reference is made to extra-symbolic objects and their properties. The substitution of the formal mode for the material one makes it possible to avoid annoying confusions and useless debates in philosophy. For instance, the apparent opposition between the two sentences “A thing is a complex of sense-data” and “A thing is a complex of atoms”—both of them formulated in the material mode—may give rise to a sterile debate between phenomenalists and realists. Their translation into the formal mode22 permits one to Carnap (1937). “Every sentence in which a thing-designation occurs is equipollent to a class of sentences in which no thing-designations but sense-data designations occurs”; “Every sentence in which a thing-designation occurs is equipollent to a sentence in which

21 22

42 understand that these two theses can be reconciled, since they simply concern the possibility of translating a thing-sentence into “equipollent” ones. In 1942, with Introduction to Semantics, Carnap enlarged his views by adding semantics to syntax. Some of the results of the earlier work had now to be set in a new semantical frame,23 but on the whole Carnap did not regard such an addition as in any way problematical. Bergmann took Carnap’s injunction about the formal mode of speech very seriously, and applied it to the case of pure semantics. According to Bergmann pure semantics ought to be conceived as “the study of semantical metalanguages formally considered.”24 Let us examine in some detail what that means. Assume an object language, L, and then construct a semantical metalanguage, M, which speaks about L. Syntactical metalanguages speak about the syntax of their object languages, and therefore they include predicates such as ‘sentence’, ‘name’, etc. Semantical metalanguages, since they have to speak about the semantics of their object languages, include predicates such as ‘true’ and ‘designate,’ which are not present in syntactical ones. Therefore in the metalanguage M there will occur sentences such as “’Fido’ designates Fido.” In this sentence the expression occurring to the left of the predicate (quotes included) is a name (in M) of a symbol of L (it designates a symbol of L); the expression occurring to the right of the predicate is a name (in M) of the same individual which in L is designated by ‘Fido’. Thus the metalinguistic sentence of M refers both to symbols of L and to objects designated by symbols of L. In this sense such a sentence is not “formal,” since reference is made to extra-symbolic objects. But that is not a problem in Bergmann’s opinion, since according to his definition semantics is the study of semantical metalanguages, and such a study is carried out in a meta-metalanguage (MM), by means of which we can speak of M (furthermore, the semantical metalanguage which is the subject matter of semantics is to be conceived of as uninterpreted, since in pure space-time co-ordinates and certain descriptive functors (of physics) occur”: cf. Carnap (1937), p. 301. 23 Cf. the Appendix of Carnap (1942). 24 Bergmann (1944), p. 154.

43 semantics we are not interested in actual interpretations). And that in fact is what we have just been doing, using (ordinary) English as MM. I spoke of examining “in some detail” Bergmann’s views, for they are really more complex than that. Bergmann devised a very baroque hierarchy of languages, with different levels of metalanguages and “intermediate” levels which are supposed to “link” in some way metalanguages belonging to different levels. Yet these complications can be safely ignored here, since the core idea underlying Bergmann’s conception of semantics is simply that it is a meta-metalinguistical study of (uninterpreted) semantical metalanguages. In MM no reference at all is made to extra-symbolic objects, so that semantics can preserve its formal character. As has already been said, Bergmann put forth this view of semantics as if it were Carnap’s own view. Therefore he could not accept Carnap’s statement according to which semantics deals—among other things—with individuals, attributes, and propositions, if with these terms reference is made not to linguistic expressions, but to extra-symbolic referents. In accordance with this cautious expository strategy, Bergmann limited himself to proposing a terminological change, i.e., the substitution of ‘sentence’ for ‘proposition’ (and of ‘name’ for ‘individual,’ and of ‘predicate’ for ‘attribute’), since the latter term, in each case, would have been—in his opinion—misleading.25 In his reply to Hall and Bergmann, Carnap immediately spotted the fundamental problem, as he said that whereas Bergmann conceives of semantics as the study of semantical metalangues formally considered, he himself characterizes it as the study of interpreted object languages. But then he dismissed the whole issue by attributing Bergmann’s criticisms to a mere misunderstanding, ultimately depending on a certain inability to distinguish clearly between use and mention. The point is, of course, that in Bergmann’s view, Carnap’s characterization of semantics cannot be accepted. In fact according to such a characterization semantics would not be a formal discipline; it would abandon the formal mode of speech and adopt instead the material one, with all the ensuing difficulties and confusions. Now it is possible to come For further observations concerning Carnap’s ambiguous use of the term “proposition” cf. Hochberg (2001), pp. 2-10.

25

44 back to the loose end concerning the charge of realism directed against Carnap. From Bergmann’s point of view, by claiming that semantics speaks directly about propositions, conceived of as facts or states of affairs, as well as about individuals and attributes, Carnap would make the same mistake he had previously criticized when banning the material mode of speech. If semantics is allowed to say that sentences designate propositions (states of affairs), that is not so different from saying that a thing is a complex of physical atoms or of sense data. By doing so, Carnap abandons all the advantages of the formal mode of speech, and he also forgets the real meaning of Wittgenstein’s distinction between saying and showing. Bergmann’s remarks concerning the last question are scanty indeed, and not very clear. The main point of Bergmann’s criticism seems to be that Carnap erroneously believes that the mere fact of using a metalanguage allows one to avoid the difficulties pointed out by Wittgenstein. But that of course is not true, since Wittgenstein’s distinction between what can be said and what only shows itself turns up again at each successive metalanguage. Carnap’s error probably depends also—according to Bergmann—on his collapsing two different distinctions, i.e., that between syntactical and semantical, and that between formal and non-formal. In Carnap’s view, semantics, having to do—in some sense—with interpretations, is to be regarded as non-formal. But that would mean, from Bergmann’s viewpoint, the abandonment of the positivistic attitude. It has already been noticed that, in the light of Bergmann’s later views, the charge of realism directed toward Carnap may seem ironical. Yet the same charge—and that also has already been noticed—was made again by Bergmann and by “authorized” students of his some years later, when realism had already been fully embraced by Bergmann himself. A full treatment of the question lies beyond the scope of this paper; yet something may perhaps be said. It seems to me that Carnap’s version of realism was—from Bergmann’s point of view—an “uncritical” one, in the sense that Carnap simply fell back to a pre-positivistic attitude, whereas Bergmann’s own version of realism was a “positivistic” one, forced on Bergmann by the fact that a non-realistic ideal language would not be capable of accounting for what is phenomenologically given. But at this

45 point the big problem of what exactly Bergmann’s realism amounts to creeps up, and that is decidedly outside my present task. The Negation of “Aboutness” Another suggestion is present in “Pure Semantics, Sentences, and Propositions.” Elaborating on his complex hierarchy of metalanguages, and taking a cue from Charles Morris’ classification of semiotic disciplines, Bergmann devised, in addition to semantic metalanguages, pragmatic ones. This suggestion is taken up in “A Positivistic Metaphysics of Consciousness,” where pragmatic metalanguages are put to work in dealing with the question of propositional attitudes. It must be noticed, however, that “A Positivistic Metaphysics of Consciousness” is mostly programmatic, in the sense that speculative suggestions are put forth, but no actual and detailed analyses are carried out. It has already been said that one of the conclusions Bergmann thought he could draw is that aboutness is a mere illusion. The justification for such a conclusion is not clear, but it seems that it has to do with the fact that any interpretation is “extraneous” to the language of which it is an interpretation. As Bergmann said, rather cryptically, “The successive languages do, formally considered, not communicate with each other.”26 This formula seems to mean that each metalanguage can speak of the interpretation of its object language, but that this remains something that has to do with the metalanguage itself, not with the object language. In “‘Chien’ bedeutet Hund,” ‘Chien’ is not a French word, it does not belong to the object language; rather it is a German word, it belongs to the metalanguage. German (the metalanguage) does not communicate with French (the object language). Of course the French word ‘chien’ is the referent of the German word ‘ ‘chien,’ ’ but that is something that has to do with the interpretation of German, and one can speak of such an interpretation only in a meta-metalanguage (which in this case is English). At each level of the hierarchy a language and its interpretation are reciprocally extraneous; their relationship can be stated (said) only at the successive level, but here the same problem turns up again, for the 26

Bergmann (1945), p. 218.

46 metalanguage as well is extraneous to its own interpretation. This rather confused argument seems to be what Bergmann’s “elimination” of aboutness amounts to, or at least that is what I have been able to extract from Bergmann’s allusive remarks. However, that is a position that Bergmann would abandon within few years, with the introduction of the intentional nexus. Propositional Attitudes Now it is possible to come back to a second loose end, concerning the sense in which—every appearance notwithstanding—the views put forth in these papers can be considered as a preparation for Bergmann’s later papers, such as “Intentionality.” It is well known that one of the main ingredients of realism—according to Bergmann’s views—is the distinction between acts and their objects (or intentions). If compared with such a distinction, that between physical and phenomenal is rather shallow: in a sense, a minimal realistic position could be secured even in a world in which physical objects are not admitted, provided only that acts and their contents (intentions) are not squeezed into a single level of reality, thus implicitly trying to reduce acts to contents.27 From an analytic point of view, that means that a realist must be able to distinguish clearly in his ideal language between what is said by “This is green” and “I know that this is green.” And there is a sense in which both idealists and materialists fail to do so, since their world is a one-level world.28 The importance of the distinction between acts and their intentions is something on which Bergmann drew explicitly attention at least since “Bodies, Minds, and Acts.”29 But it seems to me that attention to the same point is implicitly drawn (perhaps rather dimly) already in the papers of 1944-45, and that this is the reason that Bergmann thought that these papers were the forerunners of later developments.

The clearest statement of such a view can be found in Bergmann (1960). Cf. especially Bergmann (1967). 29 Bergmann (1954); but the paper was first read in 1952 at a meeting of the American Philosophical Association. 27 28

47 The main allegation made against logical positivists in “A Positivistic Metaphysics of Consciousness” is their neglect of propositional attitudes, that is to say, of acts. In fact, according to Bergmann, the positivistic analysts “must, first of all, take seriously the distinction between such linguistic events as ‘I see red’ and ‘I know that I see red’.”30 Bergmann’s submission is that such a distinction can be reconstructed with the aid of a hierarchy of pragmatic metalanguages. In other words, a reconstruction of propositional attitudes within the schema provided by a hierarchy of pragmatic metalanguages is capable of preserving the fundamental distinction between something being green and the knowing that something is green. Another similarity with the later views is the idea that this reconstruction can be accomplished without resorting to the notion of a self, which is rejected for Humean reasons. Maybe this solution is somehow favored by the verbal analogy between the different levels adding up to a hierarchy of languages and the idea that acts and their intentions belong to different levels. However, the details of such a reconstruction are not provided by Bergmann, who would abandon within few years the whole frame of reference based on the hierarchical machinery.31 One of the reasons for the abandonment was the increasing role of ordinary language in Bergmann’s view, but all that belongs to another story. The specific means by which the fundamental distinction is secured is in a sense a minor point, provided that the distinction is indeed secured. The Method The whole discussion concerning a proper conception of semantics can also be seen as a first implicit attempt at formulating Bergmann’s views on philosophical method. First of all, there is the idea of a reconstruction of traditional philosophical positions, to be attained by means of a formal language. Of course the solution was not completely new, since the basic idea of avoiding the absurdities involved in the traditional formulations of philosophical questions had been developed in detail by Carnap. Yet 30 31

Bergmann (1945), p. 201. Cf. Bergmann (1947b), p. 189, footnote.

48 Bergmann insisted that what is obtained in virtue of the switch to the formal mode of speech is not a dissolution of the traditional questions, as Carnap had claimed, but rather a reconstruction, which means that the reconstructed questions have a sense after all. The reasons of their having a sense are not clearly stated by Bergmann in these papers, but the major point is that ontological questions cannot be simply brushed aside. The unavoidability of metaphysics, though conceived in a peculiarly positivistic manner, is affirmed more than once: “it is…the system, the pattern of their [the positivists’] clarifications, linguistic, scientific, or otherwise, which ‘shows’ the pattern of the positivists’ metaphysics.”32 Such a metaphysics cannot ultimately be “said,” because of Wittgenstein’s famous prohibition, but that does not mean that it does not exist: “the gist of the positivists’ metaphysics is indeed, strictly speaking, ineffable;”33 that explains why people “have come to believe that they [scil. the positivists] have eliminated what is so tantalizingly elusive.”34 That means, in a sense, that “both parties are right, the critics who contend that positivism is just another implicit metaphysics and the positivists who protest that none of their work has any other than a merely historical reference to traditional metaphysics.”35 Therefore that is a question of relative emphasis, and

Bergmann (1945), p. 196. Cf. the much later passage in Bergmann (1960) in which Bergmann likens the method of ideal language with that of cubist painting: “At the beginning of the century a new style in painting emerged. Its name, cubism, alludes only to a part of the novelty. The other part is this. We are told that we are looking at a picture of a guitar, or of a woman’s face. At first glance, with an untutored eye, we see nothing of the kind. Looking more closely, we discern “elements”; different perspectives of the object in question in what seems an arbitrary scatter; or perspectives of the several different parts of the object; or, perhaps, discover that the apparent scatter is a very deliberate pattern which in its own way justifies the claim that we are looking at a picture of either a guitar or a woman’s face. Yet there is nowhere a representational image. The reconstructionist’s several cores are like the painter’s “elements.” His deliberate pattern is produced by the method. As in the painter’s case, elements and patterns are equally essential. Finally, as there is nowhere on the canvas a representational image, so the reconstructionist never speaks philosophically” (p. 18). 33 Bergmann (1945), p. 195. 34 Ibid. 35 Bergmann (1945), p. 196. 32

49 Bergmann chose “to emphasize the continuity between positivistic and non-positivistic thought.”36 That is the reason that Bergmann decided to speak of a positivistic metaphysics of consciousness. In fact Bergmann charged Carnap with holding a realistic metaphysical position. From Bergmann’s point of view the charge consisted in the position’s being realistic, whereas what probably most irritated Carnap was the allegation of holding a metaphysical position. But that is only an ad hominem argument, since for Bergmann everyone is compelled to hold a metaphysical position. The defect of Carnap’s lies in its being uncritical. The fact that Carnap did not recognize the unavoidability of metaphysics (though “reconstructed” along positivistic lines) only made it easier for him to slip into the eminently crude form of realism objected to by Bergmann. Looking at things from Bergmann’s perspective, his own view was at the same time more “metaphysical” than Carnap’s, since the legitimacy of metaphysics is explicitly acknowledged, and more “positivistic,”37 since it holds fast to a formal mode of speech (this seems to be the main purpose of the baroque hierarchy of languages devised by the early Bergmann). As has been shown, a similar “paradox” arises with respect to the “realism” issue: thanks to his faithfulness to the positivistic spirit, Bergmann considered himself to be less “realist” than Carnap, but at the same time the emphasis on the role of propositional attitudes paved the way to his own later brand of realism.

REFERENCES Bergmann, Gustav. 1944. “Pure Semantics, Sentences, and Propositions,” Mind, N.S. LIII: 238-257. Bergmann, Gustav. 1945. “A Positivistic Metaphysics of Consciousness,” Mind, N.S. LIV: 193-226. Bergmann, Gustav. 1947a. “Undefined Descriptive Predicates,” Philosophy and Bergmann (1945), p. 194. On the persistence of “positivism” in Bergmann’s views, even in later years, cf. Fred Wilson’s contribution to this volume, “Placing Bergmann”.

36 37

50 Phenomenological Research, VIII: 55-82. Bergmann, Gustav. 1947b. “Sense Data, Linguistic Conventions, and Existence,” Philosophy of Science, XIV: 152-163; then in Gustav Bergmann (1954). The Metaphysics of Logical Positivism, Madison: University of Wisconsin Press: 176-196. Bergmann, Gustav. 1954. “Bodies, Minds, and Acts,” The Metaphysics of Logical Positivism. Madison: University of Wisconsin Press: 132-152. Bergmann, Gustav. 1955. “Intentionality,” Archivio di Filosofia: n. 3: 177-216; then in Gustav Bergmann (1959). Meaning and Existence, Madison: University of Wisconsin Press: 3-38. Bergmann, Gustav. 1960. “Acts,” Rivista di Filosofia: 3-51. (Italian translation with the title “Dell’atto”); then in Gustav Bergmann (1964). Logic and Reality, Madison: University of Wisconsin Press: 3-44. Bergmann, Gustav. 1967. “Preface to the second edition,” The Metaphysics of Logical Positivism. Madison: University of Wisconsin Press: IX-X. Black, Max. 1945. “Review of Rudolf Carnap,” Introduction to Semantics, and Formalization of Logic,” Mind, N.S. LIV: 171-176. Carnap, Rudolf. 1937. The Logical Syntax of Language. London: Kegan Paul Trench, Trubner & Co., 1st ed. Logische Syntax der Sprache, Wien: Julius Springer, (1934). Carnap, Rudolf. 1942. Introduction to Semantics. Cambridge: Harvard University Press. Carnap, Rudolf. 1945. “Hall and Bergmann on Semantics,” Mind, N.S. LIV: 148155. Hall, Everett W. 1944. “The Extra-Linguistic Reference of Language. II: Designation of the Object Language,” Mind, N.S. LIII: 25-47. Hochberg, Herbert. 1994. “From Carnap’s Vienna to Meinong’s Graz: Gustav Bergmann’s Ontological Odyssey,” Grazer Philosophischen Studien, XLVIII: 1-49. Hochberg, Herbert. 2001. The Positivist and the Ontologist: Bergmann, Carnap, and Logical Realism. Amsterdam: Rodopi. Pinsky, L.O. 1954. “Positivism and Realism,” Mind, N.S. LXIII: 495-503. Sellars, Roy Wood. 1947. “Positivism and Materialism,” Philosophy and Phenomenological Research, VII: 12-41.

Bergmann as Historian ROBERT BAKER Union College

G

ustav Bergmann was born in Vienna on May 4, 1906. In his eightyone years of life Gustav Bergmann became a mathematician, a philosopher, a lawyer, a refugee, a psychologist, a professor of philosophy, President of the American Philosophical Association, and founder of what some have called the “Iowa School.” Bergmann’s contemporaries recognized Bergmann’s brilliance early on. His thesis adviser, Walter Mayer, recommended him to Albert Einstein as an assistant. Moritz Schlick, physicist, philosopher of science and founder of Der Wiener Kreis, or the Vienna Circle (1922-1938), recruited him as its youngest member. In 1929, the Circle published its manifesto Wissenschaftliche Weltauffassung: Der Wiener Kreis; in English, “A Scientific Conception of The World: The Vienna Circle.” In a profound sense, the doctrines enunciated in this document shaped the contours of Bergmann’s intellectual life. The manifesto announced a new empiricist and positivistic movement that would be different from earlier positivist movements by virtue of its commitment to a new methodology, logical analysis. The new logically analytic positivism, which was soon dubbed “logical positivism,” took as its primary missions the unification and defense of science, and the clarification and analysis of scientific and philosophical thought. One of its principal analytic tools was a meaning criterion that sorted statements into those whose truth was ascertainable as matter of language or logic—the truths of mathematics—and those testable in experience. Statements whose truth was not ascertainable by either of these means were considered meaningless nonsense. Since the truth of most theological and metaphysical statements was not testable by these means, most theological and metaphysical statements were dismissed as meaningless or nonsense. Acknowledging that not every adherent “will be a fighter” against nonsense, the manifesto nonetheless championed the extension of science into “per-

52 sonal and public life…education… architecture, and the shaping of economic and social life.”1 It also declared intellectual war on metaphysics, theology, and other forms of epistemic nonsense, which it condemned as sustainable only through reactionary political and socio-economic forces. The manifesto was intended to be provocative and was received as such by philosophers and the public. One student found its pronouncements so disturbing that he assassinated Schlick (on June 22, 1936). The anti-Semitic press, mistakenly believing Schlick to be Jewish, defended the assassin on the grounds that logical positivism was antithetical to the way “we Christians live in a Christian German state” (NA 2006). After the Anschluss, the annexation of Austria by Nazi Germany on March 9, 1938, Schlick’s assassin was released from prison. The Vienna Circle dispersed shortly after Schlick’s assassination. Bergmann, who was Jewish, fled Austria for America bringing with him the intellectual heritage of the Vienna Circle and the problem set that engaged its members. In 1939, Bergmann took a job at the University of Iowa, where he taught for three and one-half decades, and from this vantage point exerted a powerful intellectual influence at Iowa and on American philosophy in general. Bergmann as Historian I was both a second-generation positivist and a colleague of Bergmann’s. As an undergraduate, I had intended to become a historian but, perplexed by certain conceptual problems in the construction of historical accounts, I enrolled as a graduate student at the Minnesota Center for the Philosophy of Science. There I studied with another former member of the Vienna Circle, Herbert Feigl, and with one of Bergmann’s students, May Brodbeck. My years at Minnesota, 1959-1965, coincided with Bergmann’s peak productivity as an historian of philosophy. His essays in this area naturally piqued my interest and were to assert an indelible influence on my own scholarly work.

1

Sakar (1996), pp. 339-340.

53 Bergmann published his first historical essay, “Logical Positivism,” in 1950. Although originally published as a chapter in a book entitled A History of Philosophical Systems,2 Bergmann was acutely uncomfortable with the idea of himself as historian. He characterized his essay was “a systematic and critical account, preceded by a minimum of historical remarks.”3 Bergmann continued to deny any historical ambitions in his 1953 essay “Logical Positivism, Language and the Reconstruction of Metaphysics,” but his assertions became more ambivalent. On the one hand he stated, “this is not a historical paper,” yet he remarked four sentences later that he was “in a certain structural sense [an] historian.”4 By the end of the essay, Bergmann had become less restive about denominating himself a historian, remarking, “incidentally this essay represents a structural analysis, and in a sense, even a historical one, of the several branches of the positivist movement.”5 Within a year Bergmann became much less circumspect about his role as historian, opening a patently historical paper, “Some Remarks on the Ontology of Ockham,”6 with the observation that “Neither philosophy nor history have benefited from the cautious silence which…contemporary philosophers maintain about all but a few of their remote predecessors.”7 This essay was followed in quick succession by historical essays on Gottfried Leibniz,8 Nicolas Malbranche,9 and Gottlob Frege.10 Bergmann was clearly writing a type of essay about historically significant philosophers in a recognizably significant new style, as the editors of the leading philosophical journals of the period were quick to appreciate. Bergmann called his new approach “structural history,” in order to alert readers to both its weaknesses and its strengths. The weaknesses were patent. Bergmann’s histories ignore socio-economic factors. Thus his history of posiBergmann (1950). Bergmann (1954), p. 1. 4 Bergmann (1954), p. 32. 5 Bergmann (1954), p. 76. 6 Bergmann (1954). 7 Bergmann (1960). 8 Bergmann (1956). 9 Bergmann (1956). 10 Bergmann (1958). 2 3

54 tivism ignores the social context of the impact of Schlick’s assassination, of the Anschluss, and of anti-Semitism on the Vienna Circle, even though these factors caused the positivist diaspora. His structural history focuses instead on internal conceptual and logical conceptions and constraints that drove and limited the thought of the positivists. From his first paper on the history of positivism and in the papers that followed, Bergmann formulates his histories in terms of a set of pivotal propositions accepted by a philosopher or a philosophical movement. He then explores the range of possible and actual interactions and responses to these propositions, articulating connections and interconnections that might easily be overlooked by historians focusing on textual analysis or on external socio-economic factors. Bergmann had, in effect, translated the positivist ideal of logical analysis into a new method of approaching the history of philosophy (albeit a method that had been adumbrated, as so much in logical positivism was, by Bertrand Russell in his book on Leibniz.11) Bergmann’s 1954 essay on logical positivism displays the hallmarks of his approach to structural history. The essay opens with a set of propositions accepted by all who can be classified as logical positivists: (a) Humean views on causality and induction; (b)…the tautological nature of induction and mathematical truth…(c) [a conception] of philosophy as logical analysis…and (d) [the belief] that analysis leads to the rejection of metaphysics….12

The essay unfolds as a series of commentaries on this set of propositions laying out what Bergmann called the “linguistic turn” taken by positivism and explaining the structural reasons behind positivism’s devolution into ordinary language on the one hand and ideal-language philosophy on the other. The essay was a tour de force offering a perceptive analysis that revealed what none of Bergmann’s contemporaries had seen, at least not at that time: that Oxford ordinary-language philosophy, Ludwig Wittgenstein’s therapeutic turn, and W.V.O Quine’s naturalized epistemology were all legitimate offshoots of Schlick’s logical positivism. The style and perceptiveness of this essay was replicated in Bergmann’s later forays into structural history. I was profoundly impressed by 11 12

Russell (1900), Bergmann (1956). Bergmann (1954), p. 2.

55 Bergmann’s structural histories, and years later would draw on his approach to develop a structural analysis of moral stability and change in medical ethics. Concluding Remarks and Reflections I was Bergmann’s colleague at the University of Iowa from 1965 to 1969. As May Brodbeck’s student, I had venerated and idolized Bergmann. I initially owed my allegiance to the reconstructivist wing of positivism and my first publication in a philosophical journal was a defense of Bergmann against an unwarranted critique by Wilfred Sellars.13 Yet once I was Bergmann’s colleague our relationship did not turn out as either of us anticipated. As Bergmann might have put it, the tensions in our relationship were a product of external cultural, generational, and political factors, but more importantly, of a deeper structural tension. The external factors were threefold. Bergmann represented an older generation of European and Viennese culture. I had come to Iowa from the Bronx by way of the University of Minnesota, representing a young counter-culture popularized by our poet laureate and my Minnesota neighbor, Bob Dylan. Exacerbating these tensions was a defining external fact: on August 7, 1964, the United States Congress enacted the Gulf of Tonkin Resolution opening the way for American troops to officially engage in the Vietnam war. As the situation in Vietnam worsened opposition to the war increased and on September 28, 1967, an anti-war organization known as Resist issued “A Call to Resist Illegitimate Authority.” Among the approximately 200 initial signers were Noam Chomsky, Benjamin Spock, and the Rev. William Sloane Coffin Jr. They urged university faculty to assert their “legal right and moral duty to exert every effort to end this war, to avoid collusion with it, and to encourage others to do the same” and they “called upon universities to fulfill their mission of enlightenment” by organizing anti-war teach-ins. A copy of the Call to Resist found its way into the Philosophy seminar room and Edwin Allaire discussed it with the department’s three junior faculty members—John Burdick, William Robinson, and me. I can still see Allaire standing at the head of the table 13

Baker (1967).

56 as the four of us agreed to set up a local branch of Resist and to organize anti-war teach-ins at the university. We were remarkably effective anti-war educators, engaging the university, the community and the entire state of Iowa in a debate about the war. In the end, we paid a price for our activism. All of us had to leave the University of Iowa within a few years. Burdick left academia and now operates a major gallery dedicated to Inuit art. Allaire went to the University of Texas, Austin. Robinson went to Iowa State University, where last year he was named University’s Distinguished Arts and Humanities Scholar. I became a founding member of the applied ethics and bioethics movement14 and serve as William D. Williams Professor of Philosophy at Union College (Schenectady, New York). Beyond these generational, cultural, and political differences, however, was something that Bergmann the historian might appreciate: a structural tension. In my last years at Minnesota, I had begun to explore the ordinary-language and the Wittgensteinian/therapeutic branch of positivism. Allaire, who was my role model during my years at Iowa, had recruited me to represent that wing of the positivist movement at Iowa. As Bergmann fully appreciated in his essays on logical positivism, the ordinary-language and therapeutic branches of the movement remained profoundly antimetaphysical. My teacher Herbert Feigl, remaining true to the positivist imperative, was fond of remarking that “philosophy was the disease for which it should be the cure.” He believed that insofar as one’s views implied metaphysical commitments, one should “make the least of it.” Bergmann, however, was making the most of it, formally declaring himself a “realist.”15 and championing a florid form of realist metaphysics. In my last year at Iowa, I challenged Bergmann’s robust realism, offering a Wittgensteinian anti-metaphysical analysis.16 Although Bergmann never replied to my analysis, his position has been vindicated by history. Positivism’s anti-metaphysical stance, its taste for an ontology of desert landscapes and its forays into anti-philosophical therapeutics are passé today. Bergmann’s realist proclivities, in contrast, are quite modish. Realism—metaphysical realism, moral realism, scientific realism— realism Baker (2002). Bergmann (1964), p. 302. 16 Baker (1971). 14 15

57 flourishes in forms even more florid and exotic than Bergmann’s, including realism about possible worlds. Philosophy went in precisely the direction that Bergmann, the great structural historian, anticipated that it would. I suspect that had my younger positivist personae been Bergmann’s colleague in Vienna, or had my older personae been his colleague at Iowa, our relationship would have been different. As Bergmann might have remarked, however, relations and time have a reality that cannot be ignored. Nonetheless, time passes and perceptions change. This particular moment seems an appropriate one for me to acknowledge my deep intellectual debt to Bergmann. It was a privilege to have once been his colleague.

REFERENCES Baker, Robert. 1967. "Particulars: Bare, Naked, and Nude." Noûs. 1: (2). Baker, Robert. 1971. "Alice, Bergmann and the Mad Hatter." The Review of Metaphysics, 24: (4). Baker, Robert. 2002. “From Metaethicist to Bioethicist.” Cambridge Quarterly of Healthcare Ethics, 11 (4): 369-378. Bergmann, Gustav. 1950. “Logical Positivism.” In V. Ferm, ed., A History of Philosophical Systems. New York: Philosophical Library. Bergmann, Gustav. 1954. “Logical Positivism.” In Gustav Bergmann, ed., The Metaphysics of Logical Positivism: 1-16. New York: Longmans, Green and Co. Bergmann, Gustav. 1954. “Some Remarks on the Ontology of Ockham.” Philosophical Review, 53: 560-71. Bergmann, Gustav. 1956. “Some Remarks on the Philosophy of Malebranche.” The Review of Metaphysics, 10: 207-26. Bergmann, Gustav. 1956. “Russell’s Examination of Leibniz Examined.” Philosophy of Science, 23: 175-203. Bergmann, Gustav. 1958. “Frege’s Hidden Nominalism.” Philosophical Review, 67: 437-59. Bergmann, Gustav. 1964. Logic and Reality. Madison: The University of Wisconsin Press.

58 (No author). 2006. About Logical Positivism http://mindphiles.com/floor/philes/moritz_schlick.htm Russell. Bertrand. 1900. A Critical Exposition of the Philosophy of Leibniz. Cambridge: Cambridge University Press. Sarkar, Sahotra (ed.). 1996. The Emergence of Logical Empiricism: From 1900 to the Vienna Circle. New York: Garland Publishing.

Bergmann on the Synthetic A Priori Truth: Nothing Can Have Two Colors All Over at Once DONALD SIEVERT University of Missouri at Columbia

Every individual exemplifies at least one character; there is no character which is not at least once exemplified. That is the Principle of Exemplifi1 cation. Gustav Bergmann, “Synthetic A Priori” Every individual exemplifying a property of a domain exemplifies only that one property of that domain. That is the Mutual Exclusion of Properties Law. Emulating Gustav Bergmann

I

n the mid 1960’s, I was a graduate student in the Philosophy Department at the University of Iowa. Some of the finest teaching I have seen occurred here, with Gustav Bergmann among those finest of teachers. One could not take a course with Bergmann and forget it. He put much of his self into his teaching and in each week’s seminar he lived up to his own assertion that people are at their most interesting when they engage you about what they are then and there most concerned about. One semester I was asked to report on his “Synthetic A Priori” paper. Such reports were rarely given fully. They were largely occasions for Bergmann to expound his current thinking on matters at hand. Decades later, in the 2004-5 academic year, I was doing a project on Wittgenstein’s work with a singularly fine student. So splendid were our meetings that I started pointing out that faculty, too, have capstone experiences. Meeting twice a week, we almost made it through much of the Tractatus LogicoPhilosophicus in the spring semester. The more we worked on the 6’s of 1

Gustav Bergmann “Synthetic A Priori,” in Logic and Reality. Madison: University of Wisconsin Press, (1964), pp. 277-278.

60 the Tractatus, the more I started thinking back to the “Synthetic A Priori” paper. I had not thought of it in such detail in several decades. My recollection was that Bergmann developed a means of putting synthetic a priori truths within the general outlines of the Tractatus. I went back to reread it in order to check my recollection. Upon doing so, I thought: what a striking accomplishment it is. What I mean by this is that it shows a number of Bergmann’s intellectual virtues. Not least among these were the virtues of trying to accomplish much, to connect “his world” with those of philosophical ancestors, to answer critics, to show the empirical import of his ontological claims, to show how different components of “his world” fit together. Bergmann was interested in connecting with others while simultaneously doing his own thing. In other words, he tried to find a way in the philosophical world with the eternal dilemma of fusion and autonomy. For all the idiosyncrasy of referring to himself near the end of his teaching career as the living mind, when I knew him he viewed himself as a living mind in the philosophical community—a mere footnote to philosophy in the 20th century, as he said then from time to time. In “Synthetic A Priori” Bergmann sets himself the task of isolating a set of so-called synthetic a priori claims. He attempts this within the confines, broadly construed, of Wittgenstein’s Tractatus.2 Or, in other words, he attempts to show that there is a subcategory of truths that straddle the dichotomy of factual and logical truth. He tries to show that because this 2

Bergmann thus asked himself whether he could find a way to incorporate into the Tractatus some truths, such as that something cannot be two colors all over at one time, that Wittgenstein himself could not so incorporate into that work. Bergmann was well aware of Wittgenstein’s struggles here. Wittgenstein could not see his way to a Bergmannian solution, I think, because (1) he rejected synthetic a priori truth as not within the Tractarian framework and (2) he could not imagine looking at empirical facts to solve such a philosophical problem. Bergmann’s insistence, even demand, that his own work accord with some empirical constraints is what permits him to look to experience in the ingenious way that he does in the paper under discussion. What Wittgenstein would have made of Bergmann’s efforts can only be guessed. I would make the modest claim that Wittgenstein, while obviously turning his back on a number of aspects of the Tractatus, makes it difficult to determine what aspects of the Tractatus he would be willing to retain. He certainly continued to look at ways to accord some kind of privilege to some synthetic a priori claims, e.g., that nothing could have simultaneously two or more colors all over.

61 subcategory has affinity to both factual and logical truth it may be situated between them. This subcategory involves neither a distinctive new ontological kind of truth nor a mere reliance upon the working dichotomy between truths with which he seems to have been raised. Rather, it involves articulating how within that dichotomy one may nonetheless distinguish varieties of truth. So, making distinctions will be vital. Many a time Bergmann said, “Philosophy is the art of making distinctions.” For a long time, all I heard when he said that was that philosophy involves making distinctions. Much later, I was drawn to the artistry that is part of the conception.3 Wittgenstein was thinking similarly, I believe, when he considered as a motto for his Philosophical Investigations a line from King Lear, “I’ll teach you differences.” In both cases, it is not just that distinctions will be drawn and pointed out, but that this will be done so that one will notice, pay heed to, and learn from them.4 In other words, distinctions will be made and drawn artfully and heuristically.5 Bergmann could not be more aware that others, e.g., Kant, set themselves the same task of distinguishing such a class of truths. But Kant’s and Bergmann’s means were very different. One questioner, listening to the idea of the essay in lecture form, asks Bergmann whether his synthetic a priori is a “real,” i.e., Kantian, synthetic a priori. As Bergmann reports the exchange: In the discussion a distinguished colleague asked me whether "my" a priori was a "real" a priori such as, say, Kant's. I readily assured him that 6 it was nothing of the sort. 3

Philosophy was largely serious for Bergmann. Sometimes he could display a lightness and touch of humor. I think that is why, forty years later, I recall his saying in a philosophy of science course that one does not become scientific just because one describes someone’s walking along a street as follows: a man, m, walks down a street, s, at velocity v and at time t. I took this to be a way of saying that thinking and speaking scientifically involves entering a region of thought, representation and information, namely, science, and not just using scientists’ terminology as everyday jargon. 4 Wittgenstein and Bergmann share the desire to influence others to think for themselves. Naturally, as human beings, they are inclined to give priority to the results of their own, rather than others’, thinking. 5 Wittgenstein proposes publishing and reading the Tractatus and the Philosophical Investigations together, claiming that, so studied, each will be best understood. What an example of “teaching us differences!” 6 “Synthetic A Priori,” pp. 294-5.

62 So Bergmann’s a priori is not at all a “real,” i.e., Kantian, a priori.7 Bergmann is such a philosophical realist that he finds locating the source and/or nature of the a priori in the mind or intellect to be risking ontologically reprehensible psychologizing.8 So, instead of looking inward to the cognizing subject and cognition for the nature of the synthetic a priori, or even a contribution to it, he looks outward to the world for that nature and to language for its representation. I think it is fair to say that the distinctive variety (rather than new, ontological kind!) of a priori truth Bergmann points out centers on generality and simplicity.9 Think of a world book, a book with a complete description of all that is “in the world.” In this book there would be a subcategory of truths that are neither as general as logical and categorial truths, nor as specific as particular truths. Specifiable properties, and connections among them, i.e., relations among properties including the rela7

I think that this is because the means each uses and the results each offers are different. Still it might be the case that Kant could/would allow Bergmann’s solution as far as it goes but insist that it does not go far enough because it does not involve “true” universality and necessity. One can hear Bergmann saying that, indeed, his solution involves no such thing! Or, alternatively and correlatively, he might say that his view involves the only ontologically “true” and “factual” universality and necessity worthy of the name! 8 Bergmann is careful and respectful in his remarks about Kant. As a student, I heard primarily his dismissiveness of Kant, not the careful respectfulness. I now think that Bergmann thought well of Kant in a number of ways but thought that Kant went down some philosophical paths that Bergmann could not bring himself to go on. 9 In this essay I am trying to put Bergmann’s views in the best possible light. The gist of my effort is to indicate a taxonomy among sentences and, what for Bergmann coordinates with them, the facts that make them true. That Bergmann has both general and particular facts may be what puts some at ease: the generality of color incompatibility, some would argue, is not exhausted by the particular cases of such. That these general facts are empirical facts will leave those who think that the synthetic a priori truths are necessary truths unhappy: they will maintain that there is a necessity that needs addressing and which Bergmann ignores, rejects, or overlooks. Those who are unhappy with general facts will be unhappy with general facts either of empirical or of necessary variety. Bergmann believes that the facts in question, depending as they do on the contents of the world, are empirical facts. Even what might seem like logical and therefore necessary truths are in a sense contingent, as Bergmann’s dictum that “there is nothing logical about logic” suggests. Bergmann is sensitive to the claim/charge that he has created a “jungle.” I think that he tells himself and those who will listen that it is an empirical and needed (“dialectically necessary”) jungle. He presses on.

63 tion of properties to individuals, are the ontological basis of this subcategory. That is, if one specifies the appropriate properties and relations, and then one specifies their instantiations,10 one will be able thereby to distinguish within the class of all truths a subclass of synthetic a priori truths. As Bergmann himself saw and said, the ability to isolate this subcategory is but the last, and not very large, step in a series of steps that accompany providing an adequate ontology.11 I shall illustrate Bergmann’s solution, his isolation of synthetic a priori truth, with the claim that nothing can be two colors all over at once. Both this particular fact itself and the kind of fact that it is are important to appreciating Bergmann’s views. He encourages recognition of the nature of the particular fact as well as the nature of the kind of fact that it is. Putting it briefly, I will assert that Bergmann makes clear that color exclusion is an instance of the Mutual Exclusion of Properties Law. Bergmann’s solution is, in part, to point to the fact that nothing does, in fact, have two 10

For Bergmann, of course, there are no unexemplified properties or relations. The properties and relations are inseparable, categorially, from instantiation. Still this general categorial co-dependence is not identical to the actual, particular exemplifications in our world. Both are co-dependencies essential to Bergmann’s synthetic a priori truth. 11 There are several kinds of world book or world descriptions that may be employed here. One is an ontological inventory book, viz., a book specifying the ontological kinds that comprise the world and the kinds of categorial connections among these kinds. For example, Descartes’ dualism of finite substance, a correlative dualism of property, is a frequently used illustration of this idea. Another, partial inventory— along with categorial connection—is expressed by Bergmann’s Principle of Exemplification. (The matter of whether the Principle of Exemplification is a fact, even if a “logical fact” [!] or “formal fact”[!], like the issue of whether logical truths express such “facts,” is interesting and unavoidable when a “complete accounting of the world” is given. Indeed, Bergmann opens his “Synthetic A Priori” paper by both mentioning and quickly dismissing this nest of issues. These are matters far beyond what is appropriate to this discussion.) Still another kind of world book is an ontologist’s typology of the extant facts of the world. This book of kinds of fact, so to speak, is built upon some ontological inventory. But it goes beyond an ontological inventory by giving an inventory of kinds of facts instantiated in the world, i.e., kinds of actual, rather than merely possible, combinations of the ontological inventory to be found in the world. Mind-body unions in a Cartesian world would be a stock example of such a kind of fact. A third kind of world book is a historian’s record book, a narrative of all facts that constitute the complete history of the universe. Such a book could be written either in ordinary language or in a so-called ideal language if ontological perspicacity is one’s aim.

64 colors.12 This might strike one as curiously unremarkable. After all, aren’t synthetic a priori truths supposed to be truths, but truths that are necessary, or at least “privileged,” in some way?13 And therefore isn’t the facticity of such truths their least significant aspect? Enter Bergmann’s ontological vision and his insistence on making distinctions. A key ingredient in Bergmann’s approach is to distinguish the fact that nothing does have two colors from the allegedly “special” or “modal” nature of such a fact. Nothing’s having two colors is a matter of fact all right.14 He believes many parties to such a discussion grant this. Bergmann asks: what kind of fact is such a fact? Is there a way of distinguishing that kind of fact from other 12

By making the (mind-independent) fact be a basic part of the solution, Bergmann avoids Kant’s view about empirical connections’ being a consequence of the mind and its activities. His doing so perhaps is part of why, as he says, the problem of synthetic a priori truth was never of great interest to him: if such truth is a matter of fact, then synthetic a priori truths express facts among facts in a world of facts. The excitement associated with a new, hybrid kind of truth that is part of Kant’s worldview is lacking in this Wittgensteinian-Bergmannian view of the world as the totality of facts. 13 Cf., Wittgenstein’s mention of some sort of contradiction in everyday life in his 1929 paper, “Some Remarks on Logical Form.” This paper, (RLF), originally appeared in Aristotelian Society Supplementary Volume 9, Knowledge, Experience and Realism, (1929), pp. 162-71. It is reprinted in Essays on Wittgenstein’s Tractatus, edited by I. Copi and R. Beard. London: Routledge and Kegan Paul, (1966), pp.31-37. 14 Although Philonous, in Berkeley’s Three Dialogues Between Hylas and Philonous, believes that objects have but one of a given type of quality, he maintains that he cannot specify which property of a given range is the “real” property of something that has a property of that range. This is because, as he says, the property we attribute to a thing is a function of the conditions under which we perceive that thing and we perceive differing qualities of a given range depending upon the particular conditions of perception. He says we cannot determine the real quality of a given range for a given thing because we cannot give good reasons for selecting one set of conditions as “the” “ideal” or even simply preferred conditions. Russell, in an essay “The Ultimate Constituents of Matter,” capitalizes on this and takes the conversation in a new direction. He urges us to rethink what a thing is, to allow that things are complexes of their appearances from different perspectives, and hence that things do in fact, routinely, have more than one property of a given range of properties. Thus a given visible ordinary thing consists of more than one property of given ranges of properties, say size, color, and shape, precisely because it consists of, roughly speaking, all the differently sized, colored and shaped appearances arrayed in space-time. Thus understood, an ordinary object would as a matter of fact have more than one of a range of qualities. Whether one should maintain of the appearances that the Mutual Exclusion of Properties applies to them, even if not to ordinary objects, is a further question.

65 kinds of facts? He answers that it is an extremely general and extremely simple matter of fact, a more simple and more general matter of fact than other simple and general matters of fact such as that dogs are carnivorous or the fictitious fact that all triangles are yellow triangles. Color exclusion’s generality is indicated by its being a truth about all colors and all individuals that have color, i.e., about the domain of color. The generality also pervades our world: it is true worldwide, so to speak, that nothing has two colors and it pertains to all colors and all individuals exemplifying colors.15 Everywhere and anywhere, so far as we know, nothing has two colors at the same time. The simplicity has to do with nature of the fact at issue: it pertains to all and only individuals exemplifying colors, i.e., individuals and a single range of properties.16 The extension of the property In the discussion of this paper in Iowa City, one audience member suggested that synthetic a priori truths are “empirical principles.” Some use ‘principles’ in such a way that it might be misleading to describe them as “empirical:” if they are foundational in some way for experience in general, then they themselves are likely to be more general than synthetic a priori truths. They are not likely to be empirical in the way that empirical generalities are usually thought to be empirical, i.e., indicative of particular things/events/happenings following other things, or patterns among particular things/events/happenings. At the same time, one should grant that an empirical generality/generalization might function both to summarize what one has discovered to be true and as a “principle” for organizing inquiry into empirical facts. 16 Here Bergmann’s ontological enthusiasm might be pertinent. I think he believes that the exclusion of qualities in question is a kind of fact in its own right. Bergmann is not likely to take the reductionist route of analyzing the fact into all the individual cases of property exemplification. Rather, he is inclined, I believe, to see the latter as consequences of the former, i.e., he embraces the existence of general facts as well as particular facts. This shows both how he departs from some varieties of logical atomism, Wittgenstein’s, for example, and why Wittgenstein would not be readily able to accept Bergmann’s view of synthetic a priori truth. Without the commitment to general facts, however, and without the commitment to its being a contradiction to say that two colors are had by one thing, Wittgenstein and his Tractatus could include Bergmann’s solution. Wittgenstein in the 1929 “Some Remarks on Logical Form,” introduces a quasilogical connective, exclusion, in contrast to the exclusion involved in contradiction. This occurs as part of his renewed effort to account for color incompatibility. It is a striking maneuver on his part given his earlier insistence that logical signs could be, should be, and even had to be introduced entirely independently of the contents of the universe. The very idea that connectives might be introduced as a function of empirical contents of the universe is a departure from his Tractatus views. There he was notoriously evasive when pressed to give examples of logical atoms or the content of 15

66 color will be all colored things; all of those things will have but one color. Such a truth is like, but only like, a logical truth in that it is universally true; but it is both too empirical and too contingent a truth to be a logical truth.17 There are other domains of which it is also true that nothing can have two qualities from that domain: e.g., no note can have two pitches at the same time, no area can have two shapes, etc. The color case is specific thought. He thought it not his job to do this just because providing examples would be to engage more in empirical questions than he thought he should. He tinkers with this proposal for several years and then abandons it, moving toward a view indicated at the end of the 1929 paper, viz., that it is in linguistic rules we should look for the linguistic counterpart of incompatibility. But this presupposes “analysis of phenomena” which, by some lights, he also abandons. In the end, we are left with facts on the side of the world—some extremely general ones so general that we barely notice them, he tells us in the Philosophical Investigations—and linguistic rules when it comes to such special truths as the property/quality incompatibility ones. Note that some rules, created to reflect general facts of the world, might come to have both conventionalist and factindicating aspects. The very fact that Wittgenstein first attempts to solve the incompatibility problem with a new logical sign shows how much he sees it as a problem for him qua logician. It also shows a kind of empathy for Bergmann’s approach in that the exclusion sign is more empirical than Wittgenstein’s previously introduced connectives were. So color incompatibility is “closer” to logical truth, and “deeper” than other empirical truths. That he abandons this kind of approach and then focuses on ordinary language suggests a rejection of the Tractatus’s logical model for dealing with privileged truths—and much else. A circumstantial case may be made for claiming that his ontology is like that of Hume. If this is so, then his rejection of the logician’s conception of privileged truth might leave him unable and/or unwilling to contemplate a solution like Bergmann’s. It is also true that Bergmann’s solution is more empirical-ontological in character than was at all appealing to either earlier or later Wittgenstein. The early Wittgenstein did not wish to look at particular facts and preferred to look at the world “purely” logically. The later Wittgenstein wished to look vastly more at particular facts and cared almost not at all for looking at the world logically. (Hence Bergmann can write of the early and later Wittgenstein under the rubric “The Glory and the Misery of Ludwig Wittgenstein.”) Bergmann’s combined interests in ontology, some form of empiricism, logic, and issues of language and representation is a combination of interests that seems to exceed those of Wittgenstein at any given time. 17 Another useful example is that everything that is colored is extended and vice versa. One can see Berkeley’s use of this claim as like a categorical claim when he argues that because all colors are in minds it follows, given the principle, that all extension is in minds too. This fact about color and extension is used to locate colors and shapes with the category mind as part of his effort to undo the new science.

67 relative to the kind of fact that embraces all these other similar cases, i.e., it is but one instance of a kind of truth. I mention this so that we can see that there is a form of fact or truth that spans different domains: there is a kind of truth about mutual exclusion of qualities and it characterizes, or is true of, several specifiable quality domains. Color exclusion, pitch exclusion, shape exclusion, etc., instantiate the same kind of truth about quality exclusion. The fact that nothing has two colors is an instance of the still more general fact that there are numerous domains in which individuals possess but one property of the domain. The kind of truth in question may be expressed thus: every individual exemplifying a property of a domain such as color exemplifies only that one property of that domain. For brevity’s sake, I’ll refer to this as the Mutual Exclusion of Properties Law. Bergmann thus shows us how to see that the claim that nothing has two colors is an instance of a still simpler, general empirical truth: for certain kinds/domains of property, nothing has two qualities of those kinds/domains. This latter truth is also, as he puts it, “deep” enough— meaning, roughly, so simple and general—a truth among empirical truths that it approaches being a logical truth, i.e., it approaches being as “deep” as logical truths are. Not for one nanosecond is this to say that the variety of claim in question is a logical truth. Rather, by virtue of its generality and simplicity it approaches closer to being one than some other empirical truths do. One way to see this is to contrast the Mutual Exclusion of Properties Law with the Principle of Exemplification: Bergmann writes in “Synthetic A Priori”: “Every individual exemplifies at least one character; there is no character which is not at least once exemplified. That is the Principle of Exemplification.” The Principle of Exemplification mentions all individuals and all properties. It mentions a kind of relation, what Bergmann sometimes calls a logical relation in contrast to descriptive relations, between categories. So it is called a categorial truth. The Mutual Exclusion of Properties Law mentions a range or domain of properties and says that individuals having properties of that range or domain have at most one prop-

68 erty from that range or domain. So, the Principle of Exemplification is more general than the Mutual Exclusion of Properties Law.18 The Mutual Exclusion of Properties Law, being a generality about individuals and properties, is in that very respect like a logical/categorial claim such as the Principle of Exemplification. The Principle of Exemplification deals with ontological categories as such. It is more general than the Mutual Exclusion of Properties Law because the latter deals with instances of those categories. The Mutual Exclusion of Properties Law is a quasicategorial claim involving empirically and ontologically determined categories rather than simply ontological categories: we are dealing with empirically constituted domains of properties, individuals, and empirical facts thereof. That there is an empirical and hence factual caste to this fact is precisely part of Bergmann’s point. The generality and simplicity of the empirical caste is what both enables and requires us to “locate” the synthetic a priori truths “between” the logical and other more particular factual truths, so to speak.19 18

There is a kind of overlap between the two in that both rely on the claim that no character is unexemplified. To make this perspicuous, the Mutual Exclusion of Properties Law could be expressed thus: Every individual exemplifies at most one character of a domain; there is no character of the domain that is not at least once exemplified. This formulation shows the consistency with the Principle of Exemplification. The properties of the Mutual Exclusion Law are, however, but a subclass of the class of all properties. So they comply with the Principle of Exemplification without exhausting its scope. 19 Wittgenstein, in the The Blue and the Brown Books (pp. 55-56), seems to consider a view like Bergmann’s when he discusses color incompatibility. He says that such incompatibility is “somewhat analogous to” logical impossibility and he intimates that a universal truth is involved. His effort in the discussion is devoted primarily to “diagnosing” our ineffective attempts to provide a model for color incompatibility. We give ourselves a model of why two colors cannot “fit” in one place, so to speak, but we cannot apply the model. Still, he seems to be saying, nothing does in fact have two colors all over simultaneously. If this is correct, he gives up looking for a “logical connective” in which to “ground” color incompatibility and settles, instead, for the occurrence of two colors in one place as what is expressed by what he calls a universally false proposition. The passage in question is interesting in light of this one from Philosophical Remarks, edited by R. Rhees and translated by R. Hargreaves and R White. Oxford: Basil Blackwell, (1975), pp. 106-7: “‘Red and green won’t fit into the same place’ doesn’t mean that they are as a matter of fact never together, but that you can’t even say they are together, or, consequently, that they are never together.” Here Wittgen-

69 Bergmann shows us that a certain type of truth, the Mutual Exclusion of Properties Law, is a variety of truth different from other (varieties of) truth in our world, is a mind-independent truth and so is not a minddependent truth, etc. On this view, Bergmann distinguishes the truth in such a way that one may see its special character while also seeing that nothing ontologically new is needed for this special character. In other words, he shows how a synthetic a priori truth has a distinctive place in “his world”20 without recourse to something ontologically, or even cognitively, new.21 Bergmann’s view of synthetic a priori truth is very much in the spirit of the vision of the Tractatus.22 Wittgenstein tells us that if you have a sinstein shows his usual acuity by distinguishing between its being a matter of fact that two colors are not in the same place and its being “impossible” to say they are together “or, consequently, that they are never together.” He thus seems to contrast the “mere” matter of fact that two colors are not in one place with the impossibility of their being together and the consequent matter of fact that they are not in one place. In the (later) The Blue and the Brown Books passage, he seems to say: we cannot even make good on the impossibility to which he was so committed earlier, tied as it was to such analogies as fitting and colliding. That leaves open the possibility that he comes to the view that “they are as a matter of fact never together,” the possibility that Bergmann urges. 20 Bergmann does not even veer towards Kant’s manner of securing such truths. 21 I should make clear that Bergmann believes that there are general facts over and above their instances, so to speak. One way to express this is to say that he views laws as expressing connections among properties, for examples, and (so viewed) laws are not the kind of thing to be “reduced to” their instances. Synthetic a priori truths can involve general facts, but they are not themselves what induce Bergmann to commit to the existence of general facts. In other words, Bergmann already has general facts in his ontology when he writes “Synthetic A Priori” so he needs no additional kinds of facts to accomplish the aims of the essay. Russell thought similarly that there are general facts; Wittgenstein wished to avoid such facts in favor of particular facts only. In Russell’s idiom, their metaphysical zoos differed in content. 22 Wittgenstein himself makes no place in that work for such truth. This is largely because he insists that synthetic a priori truth be some kind of necessary truth. For Wittgenstein it is then incumbent upon him to come up with a way for such truth to involve logical necessity. Bergmann proceeds differently. Wittgenstein thought that logic required proceeding without looking at the world. Bergmann “looks at” the world, takes the facts, i.e., in this case the empirical regularities, associated with synthetic a priori truths to be primary. As he says sometimes, one takes the world as one finds it. For him the question becomes then: how may these facts be distinguished from other facts. And his answer is that they are a distinctive kind of fact, but not an ontologically or logically distinctive kind of fact.

70 gle proposition or fact then there is a sense in which you may see in it the whole logical world, i.e., the world understood logically. This is because every particular fact or proposition contains the logic of the world. Logic, Wittgenstein tells us, is the scaffolding of the world and pervades all of space including, obviously, logical space as well as factual space (so to speak). Every fact is a combination of “objects,” i.e., ontological simples. Factuality is a combinatorial matter. Objects either are or are not in combination. This is one way in which two-valued logic pervades every fact in a Wittgensteinian atomist’s world. Types of facts involve types of objects in types of combinations. Correlatively and, by implication, where there are certain types of combinations, there is a lack of certain other combinations of objects. Color exclusion is a case in which there are exclusively one-one combinations between individuals and properties in a certain property domain. That exclusive one-one combinatorial feature is a type of combinatorial possibility and hence determines a type of fact correlative to that domain. Color exclusivity among individuals is a combinatorial fact, which is determined, even if in part empirically, by the type or domain of property in question. Such property exclusivity is true of other domains as well. This shows that we are dealing with a still more general type of fact, property exclusivity for a domain of domains of properties. Bergmann has considerable affection, or at least respect, for the Tractatus. He takes into the empirical domain Wittgenstein’s general logical and ontological insight that facts are combinatorial matters and that there are types of facts. With the ontological basis of a synthetic a priori truth in hand, one may see and say the class of such truths. In other words, seeing the ontological-empirical type combination possibility that is the basis of the fact of color quality exclusion enables us to see the nature of all such facts. It enables us to see the Mutual Exclusion of Properties Law. Bergmann had a noteworthy ability to structure issues and their discussion. In this case, he had the ability to structure the discussion of synthetic a priori truth so that one may see how widely accepted facts, facts involving properties and relations, share a structure that is there for the noticing. If one will but look, think, and then see in light of the previous looking at and thinking about what one sees, one will come to see the distinctiveness of synthetic a priori truths in Bergmann’s world. One will see

71 the nature of a kind of fact and also that nothing ontologically new is required by the solution he proposes. Bergmann shows us, in other words, the general features and relations in the world, which with their instantiations quite literally enable us to pick out all and only synthetic a priori truths. Holding aside whether or not his full list of kinds of truths so generated is acceptable to all, the very idea of an ontological and realistic basis for such a list is what typifies Bergmann’s work. It is we who notice and discover that the world has such truths in it. We are reporters of those truths and generate their listing, but in no sense do we generate the facts that underpin those truths. That is the ontological realism that so drives Bergmann.23 What I wish to point out in addition here is that Bergmann shows us how the nature of the ontological enterprise applied to a traditional puzzle helps us to understand our world. We can, in this instance, see a class of truths, by seeing a kind of fact, in virtue of seeing the ontological basis of both. In another idiom, he helps us to understand the typology of facts and truths in our world. When it comes to the epistemology of such a synthetic a priori truth, Bergmann is not very helpful. I suspect that this is deliberately so. For one thing, his task is to try to classify truths he, and his audience, take for granted. Further, he sometimes insisted that epistemology is but the ontology of the knowing situation. That is one way to steer clear of a range of issues. Another, directly pertinent one is this set of remarks about knowledge of synthetic a priori truths: I also understand what one means when he says that we are more certain of these facts than we are of any other. One may reasonably doubt that what he says is true. That, though, is unimportant as long as psychologism is kept out by the two little words which are italicized.24

Clearly, there is a premium on avoiding psychologism. Bergmann, while not addressing how we know synthetic a priori truths, insinuates that they are not epistemologically privileged.25 23

Prichard thought that Kant confused knowing with making and went down the road of claiming that we know truths because we make them. 24 “Synthetic A Priori,” p. 296. 25 Here I should recall that Bergmann lists a variety of such truths. Some may be wholly presented, and so knowing them may seem unproblematic. It is not clear to me

72 Why is Bergmann so very adamant that his view of synthetic a priori truth is not Kantian? For Kant, synthetic a priori truth is conceived differently from the outset. What Kant called true universality and true necessity were hallmarks of “his” a priori. These are in contrast to what Kant himself calls merely assumed or comparative empirical universality and necessity. Kant thought that at one level Hume’s view of the phenomenal world was entirely correct as far as it went. There is the specified assumed and comparative, but no “genuine,” necessity or generality in that Humean, phenomenal world. In that world, it is, so to speak, just one damned particular thing after another. Hence, although one may speak of assumed, comparative and empirical universality and necessity in that world, there is no genuine universality or necessity there. One must look elsewhere for these. Consequently, Kant turned inward, toward the knowing subject, to “locate” “there” “true” universality and necessity.26 He believed that one may “locate” “true” universality and necessity among our cognitions. And surely, once this decisive move is made, we can expect a different solution because the very problem is different. Kant needs to establish how synthetic a priori truths have a distinctive cognitive origin and nature vis-à-vis other, e.g., phenomenal, truths. That is, he needs to indicate some features of the cognizing subject that enable one to distinguish such truths. Kant needs the art of distinction as much as anyone to make it all work and even just to show how it works for him.27 Sometimes Kant’s assimilation of ‘necessary,’ ‘universal,’ ‘a pri-

what the “correct” epistemological gambit for the Mutual Exclusion of Properties Law as such is, even if all of us are aware of numerous confirming, and no disconfirming, instances. Bergmann clearly focuses more on locating the law among other facts than on how we know the truth of such a law. In this way, again, he locates himself closer to matters of experience than Wittgenstein while also aligning himself with the Wittgensteinian task of focusing on kinds of facts rather than on (knowledge of) particular facts. 26 Speaking with some Iowans for a moment, I would say that to Hume’s desert ontology of the world, Kant added a jungle ontology of cognition. Kant, I believe, thought synthetic a priori truths could be generated only by an active mind qua agency manipulating concepts in certain ways. 27 That Kant and Bergmann share some eagerness to make both numerous distinctions and some similar distinctions, say between the formal/logical and the empirical, is also worth pointing out. But their manner of giving ontological implementation to these

73 ori,’ and ‘subjective’ are breathtakingly immediate and breathtakingly unsatisfactory. He introduces complexity—e.g., a concept of necessary connection, necessary synthesis of this concept with others in judgment, as necessary conditions for cognition of “objective,” “genuine necessity”— within subjectivity. This is how he achieves “true” or “objective” universality and generality for synthetic a priori truth.28 For all Kant’s considerable willingness, even eagerness, to introduce, to rely on and to insist upon distinctions, Kant needed still additional ones, it seems to me. His attempt to generate objectivity from subjectivity is a major effort. This is an ingredient of his “Copernican Revolution” in philosophy. A key idea in his effort, specifying the conditions for anything to be an object for a subject, shows that his notion of objectivity is tied to subjectivity. Whether he sought or achieved the kind of objectivity Bergmann seeks is at least arguable. Bergmann would, to use his own expression, reach for Ockham’s razor rather than embrace such a “real” and Kantian a priori.29 distinctions, e.g., to the general/particular distinction, is sometimes so different as to make comparison difficult at the very least. 28 One way to read Kant is to see him as saying that the mind supplies a connection among concepts such that what is a merely empirical regularity of phenomena becomes synthetic a priori truth/knowledge/judgment in light of the mind’s addition to a judgment. This risks being understood as some sort of idealism/psychologism by Bergmann and others. Prichard, for example and again, sees Kant as confusing knowing something with making it be so. By making judgments, and assimilating making judgments with making the facts those judgments are about, Kant (as Prichard sees him) achieves objectivity at the price of hyperactive subjectivity. Bergmann had a strong aversion to such idealism/psychologism. John Findlay once appalled some German Kant scholars. He indicated that he thought that there was some merit to Prichard’s reading and that the result is that on that reading the mind makes glass bottles, butterflies, and other such everyday objects. Findlay offended by saying that on this view some of Kant’s deductions are a “disgrace to human reason.” 29 Is Kant doing something wrong or doing something different? He is at least doing something different, if only because he, Kant, is very concerned with what might be called cognitions or judgments in the domain under discussion. This is what Bergmann tries (too?) hard to avoid by his appeal to the simplicity of thoughts, all thoughts, and his meaning nexus M—at least at the time of his essay “Synthetic A Priori.” It might be easier to see/say that Kant is doing something “wrong” by Bergmann’s lights if one focuses on views Kant sometimes articulates in terms of all combinations having their origin in the cognizing subject and in terms of empirical combinations/regularities being consequent to transcendental ones. Consider these remarks

74 As Bergmann’s views evolved, he moved toward what his critics would call his jungle ontology of the world, while he adamantly retained relatively extreme simplicity on the side of the knowing subject. I wonder if there are not troubling philosophical complexities—either “in the world” or “in the mind” or even “elsewhere” as Plato insisted—if we are to solve adequately philosophers’ problems in ontological terms. One’s most fundamental commitments, or, in another idiom, one’s opening gambits, no doubt help to determine where these complexities end up being located. Bergmann’s turn outward and Kant’s turn inward are such gambits. Human diversity being what it is, some wish to avoid ontological gambits altogether.30 by Kant: “The manifold of representations can be given in an intuition, which is purely sensible, that is, nothing but receptivity; and the form of this intuition can lie a priori in our faculty of representation, without being anything more than the mode in which the subject is affected. But the combination (conjunctio) of a manifold in general can never come to us through the senses, and cannot, therefore, be already contained in the pure form of sensible intuition. For it is an act of spontaneity of the faculty of representation; and since this faculty, to distinguish it from sensibility, must be entitled understanding, all combination—be we conscious of it or not, be it a combination of the manifold of intuition, empirical or non-empirical, or of various concepts—is an act of the understanding. To this act the general title 'synthesis' may be assigned, as indicating that we cannot represent to ourselves anything as combined in the object which we have not ourselves previously combined, and that of all representations combination is the only one which cannot be given through objects. Being an act of the self-activity of the subject, it cannot be executed save by the subject itself.” (Critique of Pure Reason, B 129-130) A quick reading of such passages certainly encourages attributing to Kant the view that no combination whatsoever occurs without “an act of the self-activity of the subject.” Such talk, suggestive of intensive and global idealism, would leave Bergmann the realist ready to reach for Ockham’s razor. Kant also asserts that it would be silly to maintain that empirical connections could be discovered any way but empirically: they are too independent of subjects for that to be true. 30 I am reminded of P. F. Strawson’s interesting and correct remarks about why philosophers, in contrast to scientists, for example, differ so much: “Other disciplines are defined by constitutive principles of selection among ascertainable truths. Agreement among experts in the special sciences and in exact scholarship may reasonably be hoped for and gradually attained. But philosophy, which takes human thought in general as its field, is not thus conveniently confined; and truth in philosophy, though not to be despaired of, is so complex and many-sided, so multi-faced, that any individual philosopher’s work, if it is to have any unity and coherence, must at best emphasize some aspects of the truth, to the neglect of others which may strike another philosopher with greater force. Hence the appearance of endemic disagreement in the subject

75 The current fashion, given a major impetus half a century ago by the later Wittgenstein, is to avoid complexity of cognition or complexity of ontology, indeed simply to avoid ontology, by focusing on language, human behavior, and community norms. Every now and then, as one listens and reads, one sees between the lines that the traditional issues are there to vex some of the discussants; so not all is lost. To think about distinctions in language alone and without regard even to the question what such distinctions might be getting at in the world is, as Bergmann would say, to “lose the world”—or, at least, to overlook or ignore or reject ontological questions about the world. What Bergmann’s synthetic a priori discussion, like much of his work, does is to illustrate a kind of attentiveness to issues and questions about our using language, even ordinary language, to represent the world—as surely we do all the time. He wishes to know what work we do with the variety of signs—for example, types as well as tokens, logical as well as descriptive signs, square as well as round brackets—that we use. He wishes to know whether what we “see” in the world is there to be seen or is in the cognizing subject. The mind, in Hume’s marvelous phrase, has a great propensity to spread itself onto the world. In the kind of case at hand, Bergmann wishes to direct our thought and attention away from the mind and to facts in the world that are, as he would say, the ontological ground of the class of synthetic a priori truths. Bergmann tries to be disclosing of both the views on which he relies and the reasons for adding to them. There is much of this in the paper. Thus, there is rigor and vigor to the paper that impresses me still, after several decades. I now appreciate better the task he set himself and the way he arrived at its fulfillment, and I understand better how the solution he proposes is a consequence of views he had been propounding previously as is something to be expected rather than deplored; and it is no matter for wonder that the individual philosopher’s views are more likely than those of the scientist or exact scholar to reflect in part his individual taste and temperament.” He says this in concluding his preface to Skepticism and Naturalism: Some Varieties, Columbia University Press, (1985). In keeping with the spirit of these remarks, Bergmann neither criticizes nor improves upon Kant’s efforts. Rather, he emphasizes how different their approaches are. Bergmann is interested in the facts that are part and parcel of synthetic a priori truths; Kant is as interested in synthetic a priori cognition as in the “mere” empirical facts of the matter.

76 well as views he developed in the paper itself. For example, long before anyone was saying, as we do nowadays, “Show me the money!” Bergmann was responding to those who demanded, “Show me the bare particular!” While it is almost a given that he could not satisfy everyone with his answers, he strove mightily to accept and to respond adequately to the challenge. It was but one of many cases where the ontologist, logician, empiricist, cautious-and-minimalist phenomenologist in Bergmann converged in trying to show others how he brought his disparate views into a coherent vision. What I am saying is that Bergmann’s paper shows him to be working at developing his views with an eye toward the issues and people that he valued most. While he is pleased with the solution he offers in the paper, he sees it as but a consequence of other views he developed and does not offer it as more than that. At the same time, the essay shows that ability he had to see structure, to see generality and particularity simultaneously, to comprehend a complexity of facts under an ontological simplicity—in this case a general form of fact. I mentioned earlier Bergmann’s propensity to keep the mind simple but to complexify the world.31 Anyone who knew him knew he worried about making too many distinctions, not making too few distinctions. Correspondingly, he also worried about introducing too many existents and/or subsistents. He makes this clear in his American Philosophical Association Presidential Address when he mentions that some of his best students had become ardent critics on this very issue. Still, he says there that he was fortunate to have such students. Let me add that I think his concern about new directions in which he was going, while his allies criticized him for so going, was something he was willing to put out for discussion, as that very address showed. Now for some concluding comments of a more personal nature. I made my first mistake with Bergmann when he asked why I chose Iowa for graduate study at a fall tea hosted by the Turnbulls. When I said that it was because of an assistantship, and did not say that it was so I could study with him, it was immediately obvious that I had said the “wrong” thing. Still, it was one’s philosophical efforts that counted most with him, not the 31

Kant had the opposite propensity; he was all too ready to complexify the mind.

77 particulars of social chitchat. I was able to overcome the quasi-faux pas with class work. During a wonderful course on Wittgenstein the next summer, one day I answered one of his questions directly, concisely and correctly. He was impressed. He declared to and about me “So, there are mental processes!” I knew then I was over the hump in our interactions, even if I did not especially like his manner of indicating as much. Bergmann could be forceful and aggressive, even bombastic and bullying. In short, he could be fierce. He applied these traits in his philosophical ruminations, both in developing his own views and in countering hidden and not so hidden nominalism, aversion to bareness, idealism, “undergraduates in search of culture,” and other ills he inveighed against. He could be unhappy even with his allies. When a visiting Reinhardt Grossmann asked about how many bare particulars there were in a given visible area, the response was anything but warm. He occasionally addressed this difficult side of himself. My favorite example was his saying, simply, “If you think I’m bad now, you should have seen me before my analysis.” One thing that struck me in his one-on-one interactions with me was his uncommonly strong ability to empathize with and see the structure of someone else’s view and to help the person to articulate that view even if he disagreed with it. Bergmann thus showed one how to distinguish one’s own views from others’ views and how to both articulate and defend each. I was a beneficiary of the rare and precious degree to which Bergmann possessed this talent. He not only read my work carefully and understood it sometimes better than I myself did, but he helped me to improve it significantly. Embedded in this kind of interaction is that idea that philosophy is the art of making distinctions. Bergmann helped me to make distinctions and to improve my distinction-making abilities.32 I still worry about making too few distinctions rather than making too many, both in philosophy and elsewhere. But I remain perennially in Bergmann’s, and other Iowa teachers’, debt for giving me this wonderful worry.

32

I have a personality that facilitates making distinctions with some grace and charm, even occasionally with wit. That helps with the artistry part.

78 Most significantly, Gustav Bergmann was a consummate professor. He was erudite, he had much to profess, and he did so superbly.33 He exemplified being passionate, forthcoming, unpretentious, extremely intelligent and profoundly thoughtful, ingenious, indefatigable, conscientious and candid. He was all of these in extenso, i.e., in the classroom, in his writing, in his office, in his home, in his life. His family went along with and facilitated his professorial way of life. Such was his dedication to his work that this required sacrifice by them. Many in the academic world continue to benefit from this remarkable person and his family’s support. I am fortunate to be one of them.34

33

Bergmann’s conscientiousness could manifest itself in reluctance. When I invited him to give a lecture, specifically funded to be a popular lecture in the philosophy of science in honor of A. C. Benjamin, he declined on the grounds that he had not kept up with developments in the field. He was, I believe, still teaching service courses, but not seminars, in these areas. 34 I am also the beneficiary of comments by Edwin Allaire, Phillip Cummins, Richard Fumerton and others about this essay. I have responded to their comments in this version. Michael Bostwick is the singularly fine student I mention at the beginning. His mathematician’s thoughts are what prompted me to distinguish more sharply between pure/logical/mathematical combinatorial possibility and kind-determined and/or empirically-constrained combinatorial possibility.

Gustav Bergmann’s Quest for the Ontology of Knowing: From Phenomenalism towards Realism GREG JESSON University of Iowa Ordinary objects exist independently of the minds which may or may not perceive them. If you don’t believe this truism you are mad. Believing it does not make you a realist. A realist is a philosopher who propounds an ontology which perspicuously reflects the truism…The recalcitrant case [of hallucination] presents a challenge to the realist. If he fails to meet it, he may be driven to idealism. If he cannot stomach idealism, he must 1 choose phenomenalism. Gustav Bergmann, “Realistic Postscript”

Introduction

F

ew philosophers in contemporary times have gone through as radical and public a philosophic transformation as Gustav Bergmann did in the last half of the twentieth century. As the youngest member of the Vienna Circle, he accepted what he called “a reluctant phenomenalism,” because he saw it as being slightly better than what he thought was the only alternative to materialism. However, materialism entails that there is no distinct realm of non-reducible mental states, and this was a conclusion that he could not support. Even in those early and formative years, Bergmann adhered to a robust empiricism, which not only relied on sense experience as the sole source of knowledge about the external world, but also consistently accepted introspective experience as a legitimate source of knowledge about the internal world. Bergmann came to hold that without both sources of experience grounding our knowledge, most philosophic problems were inherently unsolvable. Philosophic extravagances are born

1

Bergmann (1964), “Realistic Postscript,” p. 324.

80 out of intellectual despair; and the history of philosophy is strewn with philosophers who, unrestrained by the tethers of both kinds of experience, either wildly speculated, or prejudicially embraced, one source while ignoring the other. This principle of being open to both sources of experience would later form one methodological cornerstone for all of Bergmann’s thought, the Principle of Acquaintance (PA).2 We cannot know anything to exist unless we are “acquainted” either with 3 it or a with a part of it or, wholly or in part, with a thing or its sort. Everything we know is analyzable in terms of what we are acquainted with (otherwise we would not know what we are taking about). The parenthetical clause shows not only that as it stands the formula is epistemo4 logical but also its ontological impact. The only primary reason I have to believe of anything that it exists is that it is presented to me…This is what Descartes’ Cogito has long meant to 5 me.

Philosophy must begin somewhere, and Bergmann found it deeply problematic not to begin with what we are acquainted with in experience. He wrote: Each of us is acquainted with some things and facts (states of affairs). Synonymously, these things and facts are presented to us. What one is acquainted with he knows. Each of us also knows much else. These things and facts, however, he knows only by means of what is, or has 6 been, presented to him.

The Principle of Acquaintance is philosophically prior, in that it provides the foundational data of which ontology seeks to provide a coherent account. In this sense, the Principle of Acquaintance is ontologically infor2

Although Bergmann continued to toil steadfastly on these issues almost to the end of his life, my remarks will be limited to his middle, and most influential, period. This runs from The Metaphysics of Logical Positivism (1954), through Meaning and Existence (1959), to Logic and Reality (1964). 3 Bergmann (1964), “Physics and Ontology,” p. 110. 4 Bergmann (1964), “The Ontology of Edmund Husserl,” p. 217, emphasis in original. 5 Bergmann (1964), “Realistic Postscript,” p. 307, emphasis in original. 6 Bergmann (1964), “Ineffability, Ontology, and Method,” p. 45, emphasis in original.

81 mative, yet pre-epistemological. We do not need philosophy, especially epistemology, to inform us whether or not we know anything. Bergmann remarked: “Epistemology asks how we can know what we do know. Ontology accounts for what in fact we do know.”7 This first cornerstone prepared the way for Bergmann’s second methodological cornerstone, the philosophical priority of ontology (or metaphysics) over, and independent of, epistemology. Because epistemology can provide explanations only in virtue of what exists, ontology must be prior to any epistemological project. Epistemology is constructed from, and bounded by, ontology, whereas ontology should never be derived from, or in any way limited by, our epistemological endeavors. The heart of philosophy proper is ontology. For a while this insight was 8 lost. Now it is being recovered. What exists? What reason do we have to believe of anything that it exists? The first question leads into ontology; the second, into epistemology. Ontology is primary; epistemology is but the ontology of the knowing situation. That is the gist of my ontologism. Epistemologism re9 verses the order of precedence.

In Bergmann’s mind, the principle of acquaintance and the priority of ontology over epistemology are inseparable. When he said that “Ontology is phenomenological. The rest is science,” that is just another way of saying that analytic ontology must begin with and be rooted, not in the science of the day, but in those things with which we are acquainted.10 To be sure, ontology must be able to account for how we know the things we do know, both about our own experience and the world. Consequently, an adequate ontology must account for our epistemological position in the world, which can then elucidate how science is possible. Bergmann’s views here are very similar to the American New Realists who argued for “the emancipation of metaphysics from epistemology” in the early years of the twentieth century.11 In more recent times, 7

Bergmann (1964), “Realistic Postscript,” p. 313, emphasis in original. Bergmann (1964), “Physics and Ontology,” p. 108. 9 Bergmann (1964), “Realistic Postscript,” p. 304. 10 Bergmann (1964), “Realistic Postscript,” p. 339, emphasis in original. 11 Holt, et al., (1912), pp. 45-96. 8

82 Roderick Chisholm, in The Problem of the Criterion, argued for what he calls particularism over methodism. He claimed that in order to escape a vicious circle of epistemic justification we must establish what we know before we can establish how we know it. If nobody is smart enough to ascertain how we know, that does not in the least imply that we do not have knowledge. It is not a prerequisite of having knowledge that we be able to explain how we have that knowledge.12 G.E. Moore defended the view that knowledge does not require proof when he argued that if we did not know some things without proof, we could never know anything by means of proof. If all knowledge requires proof, then this would result in an infinite series of proofs, wherein the premises of each putative proof must be proven in order for that proof to be capable of providing any epistemic support. In such a case, nothing could ever be sufficiently proven and consequently, and paradoxically, nothing could be known.13 These principles, as Bergmann followed them through the years, allowed him to grasp and analyze the single most philosophically important fact of the world: the intentionality of acts of consciousness. Bergmann remarked that “the distinction between an ‘act’ and its ‘content’ is indeed the key to an adequate philosophy of mind...”14 The structure of the act and its capacity to refer to objects would drive him from his positivistic phenomenalism into what he called a “phenomenologically chastened realism.”15 Bergmann kept the best of positivism—its reliance on sense experience—but argued that this is not necessarily in conflict with a scientifically tempered metaphysics. This gradual revolution was not without some personal cost. Bergmannian folklore has it that when The Metaphysics of Logical Positivism first came out, Rudolf Carnap, Bergmann’s former colleague and friend, was so offended by that title that he claimed he would never speak to Bergmann again. Further, Herbert Feigl never quite got over his displeasure with Bergmann’s break from positivism. But Bergmann had found his foundational principles and, following the So12

Chisholm (1973). Moore (1959), “A Defense of Common Sense,” p. 44, “Proof of an External World,” pp. 148-150, “Four Forms of Skepticism,” pp. 210-236. 14 Bergmann (1964), “Acts,” pp. 5, 28-29. 15 Bergmann (1964), “Physics and Ontology,” p. 119. 13

83 cratic dictum, was committed to following the argument wherever it led. Such philosophical character—the courage to stand against the majority view and collegial pressure, and the determination to criticize one’s own views as rigorously as the views of others—is rare. Bergmann declared that: One who has struck out on his own, either ignoring or challenging the fashions of the day, will not, if he is sober, be certain that everything he 16 has gradually come to believe is true. [And,] I am very sober.

More than any other twentieth-century philosopher, Bergmann defended the legitimacy and the primacy of analytic ontological inquiry; and, although one may disagree with some of his conclusions, it is not possible to read Bergmann without being impressed with his philosophical seriousness and formidable rigor. After Bergmann, analytic metaphysics regained its proper role as first philosophy that had been lost in the short-lived linguistic and epistemologized vogue that had trivialized and stifled much philosophic inquiry. In this respect, he was similar to Plato, Aristotle, Arnauld, Reid, Frege, Husserl, and Moore. Many sought to refute logical positivism in toto, but, for the most part, neither the positivists nor their critics realized that positivism’s empirical foundations were logically independent of its rejection of metaphysics. The austere empiricism of logical positivism was understood to rule out metaphysics, but Bergmann argued that logical positivism had its own metaphysics. Even though he was to reject positivism’s negative stance towards metaphysical inquiry, Bergmann remarked that “…I remember with a very warm feeling the debt of gratitude which I owe to the logical positivists,”17 because they were committed to a scientific view of reality and correctly desired to rule out wild and unrestrained metaphysical speculations. Logical positivism had many good instincts; it just went too far. This is what Bergmann meant when he spoke of “the stultifying scientism of the logical positivists.”18 Rather, “science is common sense elaborated,” or, as Bergmann was fond of saying, science is just the “long arm of com16

Bergmann (1964), “Preface,” p. vii. Bergmann (1964), “Realistic Postscript,” p. 336. 18 Bergmann (1964), “Acts,” p. 21. 17

84 mon sense.”19 The task then for the philosopher is to give an adequate account of what is given without contradicting common sense. Once we leave the gravitational pull of commonsense there is no telling where we might end up. Idealism, materialism, phenomenalism, and positivism had all failed in this task and, therefore, each had to be rejected. Bergmann’s genius was his ability to salvage the deep and valuable insights from each and abandon whatever did not line up with common sense. For example, idealism, while correctly holding a “close and intimate” connection between an act and its object (its intention), went too far by collapsing the two into one thing.20 Metaphysical materialism, while rightly attempting to be scientific, went too far by embracing the “absurd but also very dull” claim that there are no conscious states.21 Phenomenalism “discovered the right ontological assay of ordinary objects,” in that objects are “temporal succession[s] of particulars each of which exemplifies certain nonmental properties and stands in certain nonmental relations.”22 It was, however, incorrect to claim that physical things are simply congeries of phenomenal things.23 Positivism, as Bergmann argued, correctly sought to be empirical and scientific, but incorrectly construed this as entailing that metaphysics was philosophically superfluous. Each of these “dead end” views is “paradoxical or absurd,” in that each denies, in all seriousness, what is clearly true. The deepest core of Bergmann’s comprehensive realism is that everything, both physical and phenomenal, is what it is “irrespective of whether or not either you or I or any one else perceives it remembers it, thinks of it, and so

19

Bergmann (1964), “Duration and the Specious Present,” p. 98; “Acts,” p. 20. Bergmann (1964), “Acts,” pp. 19, 20, 33. 21 Bergmann (1964), “Realistic Postscript,” p. 336, “Materialists1 [metaphysical materialists] claim that minds (conscious states) do not exist. Materialists2 [scientific materialists] merely claim that minds, though existents, since they are body-dependent, are not real.” Also, pp.337-338; “Acts,” p. 15, “Since then, though, [Russell] has steadily moved down the slope toward that dismal materialism (identity of mind and brain) which he now so loudly proclaims. It is deeply sad that this may be the last word of one who is nevertheless the dominant figure of the century.” 22 Bergmann, (1964), “Realistic Postscript,” p. 332, emphasis in original. 23 Bergmann, (1964), “Acts,” pp. 15, 18, 24; “Physics and Ontology,” pp. 118, 120; “Realistic Postscript,” p. 334. 20

85 on.”24 According to Bergmann, the purpose of philosophy is to describe accurately, and to provide a sufficient account of, the full range of what we have discovered in the world. Those who deny that there are minds or physical objects cannot account for “the world we actually live in,” and, therefore, will always leave us deeply unsatisfied. The only fish that always goes with the stream is the dead fish, but philosophy is compelling and intellectually valuable only if there are those who refuse to be carried along in the waters of inadequate, popular, and canonized opinion. Gustav Bergmann was such a fish. The Archimedean Point: The Act of Awareness Descartes’ dream argument has resulted in an epistemological nightmare that has lasted for nearly four centuries. During this time, much of philosophy had been dominated by attempts to solve the skeptical problem; but without an adequate account of the mind the task was impossible. Bergmann clearly realized that recurring variations of skepticism, representationalism, idealism, phenomenalism, thin empiricism, phenomenalism, and positivism all had the same fatal flaw: they lacked an adequate description of the act of consciousness. Because of this, their accounts of what we know about the world were inevitably inadequate. The great failure in all these responses to epistemological and skeptical problems is that they all presuppose an ontology of intentionality and of the epistemic act that makes knowledge of anything outside the mind impossible. Bergmann remarked: Just as clearly, everything depends on an adequate analysis of the act, in the case its intention is phenomenal, which as I now clearly see is all I have done so far, as well as in the case it is physical, which is the specific topic of this essay. To be adequate, the analysis…must, together with others, lead to the solution of all philosophical problems…[but not nec25 essarily] “simultaneously.”

To rectify this deficiency, Bergmann sought to give an adequate assay of the act of consciousness. How do acts differ from the historically rich ar24 25

Bergmann, (1964), “Acts,” p. 19. Bergmann (1964), “Acts,” p. 20.

86 ray of philosophers’ objects such as ideas and sense data? If we acknowledge the existence of conscious states, how can we avoid being confined within that private mental world? What reason do we have for believing that acts take us outside our private mental lives? Without answering these questions, how is it possible to build an ontological foundation upon which metaphysical and epistemological realism could be supported? In these efforts, Bergmann’s motivations, and many of his conclusions, were similar to those of G.E. Moore in his 1903 article “The Refutation of Idealism.” A curious difference is that Moore, who was legendary for working and reworking the same philosophical problem over and over, seems to have largely given up on finding an ontological account of consciousness, after he failed to figure out how a belief can be about something that does not exist. In his 1910-1911 lectures published as Some Main Problems of Philosophy, Moore asks: In order that a relation may hold between two things, both the two must certainly be; and how then is it possible for any one to believe in a thing 26 which simply has no being?

Then, a few pages later he announced: Possibly some positive analysis of a belief can be given, which would enable us to answer this question; but I know of none which seems to be perfectly clear and satisfactory. I propose, therefore, to give up the at27 tempt to analyse beliefs.

In contrast, Bergmann kept revisiting the issue of the intentionality of acts repeatedly for close to three decades in almost everything he wrote, convinced that it is at the center of the philosophic endeavor. The act was Bergmann’s Archimedean point, which unlike mental states as traditionally conceived (such as Lockean and Humean ideas), points beyond itself, thereby establishing the possibility that we are not sequestered within our private mental worlds. Bergmann was puzzled by those who could not admit that such intrinsically referring acts exist but, rather, embraced in-

26 27

Moore (1953), “Beliefs and Propositions,” p. 263. Moore (1953), “Beliefs and Propositions,” p. 266.

87 stead some form of idealism, skepticism, phenomenalism, or materialism. Bergmann remarked that: Awarenesses do exist…Like everyone else, philosophers are sometimes aware of their awarenesses. Many philosophers nevertheless deny that they exist. One very important one, Ludwig Wittgenstein, spent the second half of his life trying painfully to convince himself, not only that they do not exist, but even that there are none. Such persistent refusals to 28 admit the obvious are so strange that one must try to explain them.

In Logic and Reality Bergmann recounted how his journey from phenomenalism to realism had been gradual and was still in process. He declared that he was attempting to relinquish all the remaining phenomenalist dross that was blocking the full-bodied realism he sought. In describing himself, Bergmann claimed, “I, for one, am as intent as the most embattled classical realist on securing [metaphysical realism] dialectically”— and this defense was centered on the act.29 Acts of consciousness are unique in that, by their very nature, they are the kinds of things that are necessarily of or about something other than themselves. Our thoughts function as “natural signs”30 by pointing or referring to objects (or intentions) beyond those thoughts. (Although the words “object” and “intention” are often used interchangeably for what our acts of awareness are about, and even though Bergmann often used “intention,” in this study I will usually prefer “object” over “intention,” because the word “intention” is easily mistaken as referring to the act of awareness.) Usually acts of consciousness are not the objects of our awareness (although an act can be the object of another act); rather, they are mental states that direct one’s awareness beyond the mental states themselves. This was exactly what Moore’s analysis had concluded in “The Refutation of Idealism” when he proclaimed: There is, therefore, no question of how we are to “get outside the circle of our own ideas and sensations.” Merely to have a sensation is already 31 to be outside that circle. 28

Bergmann (1959), “Intentionality,” p. 4. Bergmann (1964), “Acts,” p. 19. 30 Addis (1989). 31 Moore (1948), “The Refutation of Idealism,” p. 27. 29

88 Although Moore, in “The Refutation of Idealism,” claimed not to be providing a proof of his thesis regarding consciousness, he did claim to show that the arguments usually given to support idealism fail. Nevertheless, Moore believed that his conclusion constituted a revolutionary turn in philosophy. At the beginning of the “Refutation of Idealism,” Moore claimed: The only importance I can claim for the subject I shall investigate is that it seems to me to be a matter upon which not Idealists only, but all philosophers and psychologists also, have been in error, and from their erroneous view of which they have inferred (validly or invalidly) their most striking and interesting conclusions….If it has this importance, it will indeed follow that all the most striking results of philosophy— Sensationalism [empiricism], Agnosticism [skepticism] and Idealism alike—have, for all that has hitherto been urged in their favour, no more 32 foundation than the supposition that a chimera lives in the moon.

According to Moore, the great problem with all these views is that they misdescribe the connection between our mind and the objects of knowledge: either they held that objects are intrinsically mental, or that the only objects we are directly aware of are mental, or that access to objects in the external world is impossible. Moore sought to show that the object of the act (Bergmann’s “intention”) cannot be a part or a property of the act. Intentionality is captured neither by the part/whole relation nor by the relation of exemplification. If X is a part of Z, then both X and Z must exist; or if X is a property of Z, then both X and Z must exist. Many things are parts of wholes, yet no intentionality is exhibited. Similarly, properties can be exemplified by particulars without intentionality being present. Moore showed that intentionality has a different “grasp” or reference to objects from these relations, but he did not go further in elucidating what the referential capacity of acts consists of, and he did not show how acts can direct our attention to physical objects and to other public realities such as mathematical and logical truths. To be sure, acts direct us beyond themselves, but this is not sufficient to show that acts direct us to external realities. Acts can direct us beyond themselves to other things within our own mind, or to public things, 32

Moore (1948), “The Refutation of Idealism,” p. 5.

89 or, it would appear, to nothing at all. How is this possible, and how can we distinguish when we are in each of these different states? At this point, in the discussion of the nature of the mind, Moore went no further—but Bergmann did. In the following pages, I shall briefly outline Bergmann’s account of intentionality, and then focus on two issues that are crucial in his quest for realism: first, how does Bergmann handle the case of perceptual error and non-veridical thought; and, second, how does his analysis of the act contribute to his quest for realism? As mentioned above, Bergmann extracted what was correct and often deeply insightful from thinkers who, in his opinion, had later gone off into extremes that nobody could believe. The most important mistaken view from which Bergmann learned was idealism. It is hard to be wrong on everything, and the profound insight of idealists is that in all conscious states there is a close and intimate connection between an act of awareness and what the act is about. This connection is so intimate that it is not possible to be wrong about what a conscious state, such as a perception or belief, is about. Berkeley’s idealism was forged on exploiting this closest of all connections. The connection between the perception that an Aspen tree is in my yard and being an Aspen tree in my yard is so close that Berkeley counted them as only one thing instead of two. He was extremely suspicious of epistemological and ontological gaps, and reasoned that unless necessary connections could be established between sensations and objects, irrevocable skepticism would crash in upon all human thought. The foundation of Berkeley’s views rested on his reasoned rejection of all such gaps wherein any possible skepticism might root. In the quotes below, Berkeley and Moore, each arguing from the same phenomenon, claim to be directly aware of objects, but their views of what objects are were incompatible. But, surely, it should give us pause that two philosophers could give such widely differing interpretations of what this close connection links together. Berkeley claimed: I am the farthest from Scepticism of any man. I know with an intuitive 33 knowledge the existence of other things as well as my own Soul. 33

Berkeley (1975), Philosophical Commentaries, #563, p. 373.

90 In contrast, Moore claimed: I am as directly aware of the existence of material things in space as of 34 my own sensations.

Berkeley and Moore were both correct on one point, and both were wrong on another. They were both correct in their description of consciousness as involving an intimate and, therefore, direct connection between an act of awareness and what the act is of, or about. My thought that saber-toothed tigers are extinct is about saber-toothed tigers being extinct; and my thought is about saber-toothed tigers being extinct, whether or not any saber-toothed tigers exist. We know with complete certainty what our thoughts are about, and any philosophic view that denies this can do so only by denying what is clearly right in front of our faces. To say that we are certain just means that there is no possible evidence that could convince us otherwise at the time.35 Any view that entails that we cannot know what our own thoughts are about seems to be self-refuting. Surely, to claim that we might be mistaken about what we are thinking of is to be in the grip of a theory that cannot account for even that claim. Facts should always drive theories as opposed to theories driving facts. However, both Berkeley and Moore were also wrong on another point. The entire thrust of Berkeley’s work was to establish that the close connection between a perception (such as seeing that an oar is bent) and what it is about (being a bent oar) entails that the perception should be identified with the object (a bent oar). Berkeley argued: For what are the forementioned objects [houses, mountains, rivers, and all sensible objects] but the things we perceive by sense, and what do we perceive besides our own ideas or sensations; and is it not plainly repugnant that any one of these or any combination of them should exist un36 perceived?

34

Moore, (1948), “The Refutation of Idealism,” p. 30. Addis (1989), pp. 51-56. 36 Berkeley (1975), Principles of Human Knowledge, paragraph 4, p. 90. 35

91 Further, Berkeley argued that insofar as our perceptions exist, and because objects are identical with our perceptions (ideas), the objects given in perception must exist. In Principles of Human Knowledge Berkeley argued: That what I see, hear and feel doth exist, that is to say, is perceived by 37 me, I no more doubt than I do of my own being. I can as well doubt of my own being, as of the being of those things which I actually perceive by sense: it being a manifest contradiction, that any sensible object should be immediately perceived by sight or touch, and at the same time have no existence in Nature, since the very exis38 tence of an unthinking being consists in being perceived. I shall only observe, that if at [the] table all who were present should see, and smell, and taste, and drink wine, and find the effects of it, with me 39 there could be no doubt of its reality.

Furthermore, in the Dialogues Berkeley has Philonous say: I see this cherry, I feel it, I taste it: and I am sure nothing cannot be seen, 40 or felt, or tasted: it is therefore real.

Berkeley believed that because we cannot be wrong about what our thoughts are about, and because we could be wrong about objects if they exist independently of our thoughts, it follows that what our thoughts are about are mental entities—ideas. Moore drew the opposite conclusion. He reasoned that because there is such a close and direct connection between an act and its object, and because it is not possible for objects to be a part or property of an act, we must be directly aware of objects that are not part of our acts and are, therefore, outside “the circle of ideas.” He concluded that we are aware of material things in the external world. It is curious that, in “The Refutation of Idealism,” Moore shifts from the object of an act being “blue” to drawing conclusions about material objects. From the fact that blue is not an idea, does it follow that blue exists in the world? This is a particularly complicated case because blue is a property. But 37

Berkeley (1975), Principles of Human Knowledge, paragraph 40, p. 103. Berkeley (1975), Principles of Human Knowledge, paragraph 88, p. 124 (emphasis in original). 39 Berkeley (1975), Principles of Human Knowledge, paragraph 84, p. 122. 40 Berkeley (1975), Three Dialogues Between Hylas and Philonous,249, p. 238. 38

92 from the fact that my perception is of a strawberry being red, and that a red strawberry is not an idea, does it follow that there exists a red strawberry in the world?41 It certainly does not seem to follow. Something being “outside the circle of ideas” does not seem to entail that it exists in the world. According to Bergmann, “One who wants to use the phenomenological material dialectically, must have the power and courage to ‘schematize’ it.”42 Both Berkeley and Moore failed to account for both our conscious and epistemological positions in the world, because each gave an inadequate account of the act of consciousness. Berkeley had no place for the external world in his account, and Moore had no place for the internal world in his account. From the fact that the connection between the act and what it is about is so close and intimate, it does not follow that what the act is about is mental; and, from the fact that the act reaches beyond itself, it does not follow that what the act is about exists in the external world. Further, and this possibility seemed to elude both Berkeley and Moore, from the fact that each act is of something, it does not follow that what a thought is about exists in any sense. (I think that Moore would agree, but he did not have a principled way to argue the point, once he claimed that merely having an idea is already to be outside the circle of ideas. This impasse was captured in his quote above concerning relations requiring two terms.) However, Bergmann’s early analysis of the act of consciousness revealed that there is a distinction between what a thought is 41

In “Realistic Postscript,” (1964), p. 318, Bergmann made this point by writing: “If you use ‘real’ as I do and accept my assay of P, then, if what Moore asserts were true, the table I perceived would be real (mind-independent) on the sole ground that I perceived it. Clearly, this is a short-cut to realism. It is true, although not analytic, that what we perceive is nonmental. But the (actual) nonmental does not, we shall see, coincide with the real. The realistic urge behind the short-cut is only too transparent” (emphasis in original). 42 Bergmann (1964), “Acts,” p. 28. A clear example of Bergmann’s use of the phenomenological method in the assay of an act is: “An act is either an acquaintance, or a doubting, or a remembering, and so on. What makes it just that? An awareness is a doubting because it exemplifies a simple character which accordingly I call ‘doubting.’ If I am asked how I know all that, I answer as Locke might have: Go and look into yourself! If asked in the same vein, how I know that these characters are simple, I answer: Go and look once more!” “Acts,” p. 31. This sounds very much like Husserl’s philosophic motto: “To the things themselves!” (Du den Sachen.) I think Bergmann would very much agree.

93 about and what actually exists. Identifying consciousness of an object with the existence of that object is the idealist’s error. The idealist can collapse the object into the act only by dropping intentionality out of the act. Ideas are not of objects; they are objects. It was part of Bergmann’s realism that he did not hold that we can determine what exists merely by an examination of our mental states. Just because one is conscious of something, either in thought or perception, it does not follow that the thing exists. Bergmann’s Account of the Act While some philosophers attempted to elucidate intentionality as primarily a linguistic or social phenomenon, Bergmann was determined to discover the underlying ontological structure of intentionality that makes such linguistic and social events, and even awareness itself, possible. The following points express Bergmann’s principal claims concerning the nature of the act of consciousness. 1. Acts of awarenesses of different kinds such as doubting, believing, perceiving (the various “species”) exist, and we know this because we are directly acquainted with them. 2. The ontological account of our acquaintance with species can be explained only by these species being properties that are exemplified by acts. This accounts for the close and intimate connection between the species and acts, and for our acquaintance with them. 3. Every act of awareness is intrinsically of, or about, something else; therefore, there are no acts of consciousness that do not refer to an object (i.e., Bergmann’s “intention”). 4. There is also a very close a connection between an act of awareness and what it is about. This connection is so close that an act can be described only by mentioning what it is about. The connection is logical, analytic, and specific in the same way that the connection between premises and conclusion in a valid argument is logical, analytic and specific. The idealist goes wrong by collapsing act and object into one. If my perception that there is an Aspen tree in my yard is not identical to being an Aspen tree in my yard, then there are two things and not one. If X is about Y, then they cannot be identical.

94 5. The intimate connection between the act and its object (what it is about) cannot be discovered by scientific or empirical means, because these can discover only synthetic connections such as statistical correlations or causal connections. Such correlations and connections are not sufficiently intimate to account for our knowledge of the connection between an awareness and the subject matter of that awareness. 6. An act and what it is about cannot be linked by a “descriptive relation,” or an external relation, in which the relata are ontologically and logically prior to their exemplifying the relation. Insofar as external relations are contingent they are not intimate enough to describe accurately the relation between an act and its object. 7. An act points or refers to its object by exemplifying a property ( the “proposition”). If the act did not have a “phenomenal character” (its “intentional property”43), it would not be about anything and, consequently, would not be an awareness. A proposition (the act’s “propositional character”) is not the object of the act of which it is a property (although it could be the object of another act), but is the means by which the act selects the object. 8. Each act of awareness uniquely selects or intends a specific object such that there is a correlation between each act and each object. Bergmann’s formulation is: Bs and Ps and P M P: S has a belief (Bs) and this belief has the intentional property of being of or about P (P) which, by virtue of what it is, selects, intends, or means the state of affairs P (P). 9. There is a categorical distinction between the act of awareness and the object of awareness. The act is mental, whereas the object may be phenomenal, physical, or abstract; or it may not exist. 10. Whether or not the object of the act exists has no bearing on the intentionality of the act. The act can be of an object that does not exist, i.e., a state of affairs that does not obtain. (This, of course, is fraught with conceptual difficulties, which will be explored in the next section.) 11. In order to account for the intimacy of the connection between an act and what it is about, it is neither necessary nor sufficient to claim that there 43

Bergmann was not really happy with ‘proposition,’ given its tattered history (1964), p. 32. Husserl’s term “matter” is slightly better. Addis uses the more illuminating and correctly descriptive designation of ‘intentional property’ (1989), p. 43.

95 are intentional objects in the mind which “inexist.” The AristotelianThomistic view held that the mind takes on the form of the object, and Malebranche “taught that a mind knows only what is ‘in it’ and that for anything to be in the mind it must be a property of the mind.”44 What these views have in common is that they hold the “nonrelational conception of the act” wherein the seeming intentional capacity of acts is accounted for by means of acquaintance with a phenomenal object. However, one need not abandon realism to account for the close connection we have to objects in our awarenesses of them. The act exemplifies a “propositional character,” which has an intimate logical, or intentional, connection to the object. 12. The propositional character, or “content,”45 is a property exemplified by the act, but which transcends the act to grasp an object that is not a part or property of the act. This explains how many people can think of the same thing, or how one person can think of the same thing at different times (by having the same propositional character), and the directness with which we are aware of the objects in our awarenesses. 13. Acts, therefore, utilize two close connections, the act’s exemplification of the propositional character, and the logical, intentional connection between the propositional character and the object of the act, to direct one’s attention to something outside the act of awareness. 14. The propositional character selects its objects by means of the objects’ properties. 15. The intentionality of the act is not founded upon the act being similar to the object. Many things are similar to each other without being about each other; and things can be more or less similar, whereas intentionality does not admit of degrees. 16. Propositional content is simple in that it is not made up of nonpropositional constituents that come together to form a propositional whole, nor is it made up of propositional simples that each intend a simple object which is then (somehow) combined together into a state of affairs or fact that is complex.46 44

Bergmann (1964), “Acts,” p. 26. Bergmann (1964), “Acts,” p. 34. “Content” is another word that is easily misconstrued. 46 Addis (1989), pp. 75-80. 45

96 17. Although the content of each act is simple, the intention (object) of each act is complex. A content is always of a state of affairs (e.g., wherein a property is ascribed to a particular). 18. Acts are the means by which we have access to the public world and are, therefore, why it is possible that we are not confined to a private, mental world. Thinking of What is Not: Something’s Got to Give

The problem of thinking of something that does not exist is as easy to state as it has been difficult to solve. In Husserl’s words: A representation without an object represented is not conceivable: there are, therefore, no objectless representations. On the other hand, however, actually existing objects do not correspond to all representations: there 47 are, therefore, objectless representations .

How is it possible for a representation to signify an object when that object does not exist at all? Not only is the object not immanent in the mind, but it seems not to exist in any sense; and, if this is so, how can our consciousness be directed at it? Further, because we can think of objects that exist as well as objects that do not exist, we need an analysis that can account for both cases. That analysis must account for the similarity—each is a case of thinking. But, also, the analysis must account for the dissimilarity of the cases—one is of something that exists, and the other is of something that does not exist. Acts of consciousness are unique in that they intrinsically signify or represent, because every such act is of, or about, something. This intentional feature of acts makes it seem as if cognition is always of an existent object. As stated above, Bergmann rejected any account of intentionality as a descriptive relation (external relation). One reason for this position is that such a relation could never capture the necessary connection between an act and what it is about—its object. Further, if intentionality is a descriptive relation, then, in non-veridical acts such as false beliefs, we will 47

Husserl, (1994), p. 474.

97 be forced to come up with some kind of object to complete the relation. It is the non-veridical cases, such as perceptual error and false beliefs, which show the need to have propositional characters, or intentional properties, in the act. In the case of thinking of a non-existent thing such as a unicorn, if there are no intentional properties, and if the object does not exist, then there is nothing on the side of the mind, and nothing on the side of the world, that indicates what the thought is of. Therefore, some have reasoned, it is impossible to think of that which does not exist. Meinong’s “solution” was to devise a theory of objects wherein every possible object of awareness either exists, or subsists, or is ausserseiend (beyond being). Consequently, Meinong argued, “there are objects of which it is true that there are no such objects.”48 Where does Bergmann fit in to this? In his article “Acts,” Bergmann argued against intentionality as a “descriptive relation” and distinguished his view from the “nonrelational” conception of the act, which restricts our direct access to what is only “in the mind.” On these two views, he never wavered; but, he also argued that the intentional nexus was such that an act could signify something that does not exist in any sense. He wrote: For a (binary) relation to be exemplified, its terms must both exist. There are no centaurs. Yet I now think of one. If the intentional nexus is descriptively relational, what in this case corresponds to the existing chair which I perceive? There is no need to bring in centaurs. I mention them merely because, like Meinong’s golden mountains, they are spectacular…One way out which many took was to crowd the ontological inventory with such entities as “false facts” which merely “subsist.” The revulsion from such overcrowding was another cause for the decline of the 49 act in recent philosophy.

Here again, Bergmann was still, and correctly, influenced by his former positivist temperament of avoiding unnecessary metaphysical baggage. One does not need to concoct an ad hoc repertory of bizarre philosophers’ objects to account for facts such as our ability to perceive centaurs and golden mountains, and our ability to think of the present King of France, the round square, and an imaginary number such as √-1. Bergmann main48 49

Meinong (1960), “Theory of Objects,” p. 83. Bergmann (1964), “Acts,” p. 25.

98 tained that in cases of non-veridical perception, such as would occur when suffering from a hallucination, if one hallucinates seeing an elephant, “It follows that [the] conscious state while…hallucinating was one of perceiving the elephant.”50 Later on, the claim that one was “really” perceiving an elephant may be withdrawn, but taken in complete isolation from everything else, it is possible that a non-veridical perceptual state is indistinguishable from a veridical perceptual state. In “Acts,” Bergmann was content to claim that while hallucinating he was perceiving the elephant, and that there was no elephant that he was perceiving. There is something the act is about, because the intentional property intends and provides a mental determination, even if the object intended by the act does not exist. If acts can signify or intend objects that do not exist, then the intentional function of an act is not derived from the existence of the object. Bergmann held that in the case in which Peter is not blond, yet one perceives, believes, or imagines that he is blond, the object, or intention, is the non-obtaining state of affairs that Peter is blond. With quintessential Bergmannian directness, he stated: 1. The sentence ‘Peter is blond’ can be accounted for the PA. 2. It refers to a physical fact. 3. This fact does not exist. 4. The simple character ‘the proposition Peter is blond’, being exemplified by my awareness in the act or acts in question, does of course “exist”; or so at least I would have to say if I spoke philosophically. That is all. To grasp these four things accurately is to see that where some philosophers saw a problem, 51 there really is none.

However, four years after “Acts,” Bergmann backed away from this claim. In “Realistic Postscript,” he was to “crowd the ontological inventory” by adopting the kind of objects he once rejected in order to account for nonveridical acts. Bergmann became puzzled how there could be an intentional connection between the belief or perception that P and the state of affairs P, if that state of affairs did not obtain, or exist, in some way. He claimed that:

50 51

Bergmann (1964), “Acts,” p. 42. Bergmann (1964), “Acts,” p. 40.

99 No realistic ontology, we know is adequate unless it provides an ontological ground for the connection between a thought (P) and its intention P. In my world, the connection is the intentional tie (M) in P M P. Since M and P M P both exist, the former as a subsistent, the later as a formal fact, the connection is grounded. But I do not see how P M P 52 could exist unless P has some ontological status.

A few pages later, he remarked: “If anything is presented (to me), so is its existence.” If one doubts that this is a truth of the hard core, let him ask himself how he knows what ‘exist’ means. Why the classical realists failed to grasp this truth we understand already. There are the difficulties they could not overcome be53 cause they did not recognize that possibilities exist.

Possibilities, then, have some ontological status for Bergmann. If they did not, they would not be able to accomplish what he thought they were needed for. It is unclear how this could be an account that is compatible with thinking about impossibilities, such as the round square; and attempting to save the point by claiming that impossibilities are possible seems, at best, seriously wrong-headed. It is interesting to note that Bergmann observed several times in Logic and Reality how much his thinking had been influenced by his former student, Reinhardt Grossmann. In The Categorial Structure of the World and The Existence of the World, Grossmann analyzed intentionality as an “abnormal relation” which can “span the abyss between being and non-being”54 by being able to “connect what is there with what has no being at all.”55 Insofar as Grossmann held that intending an object requires that object as part of the furniture of the world, he maintained that in nonveridical acts one is connected to “objects which do not exist.”56 Thus, the vast furniture of the world includes two realms of objects: those that exist and those that do not exist. Such non-existent objects do not have any features, for “things that do not exist have no properties,”57 and differ from the 52

Bergmann (1964), “Realistic Postscript,” pp. 307-308. Bergmann (1964), “Realistic Postscript,” p. 312. 54 Grossmann (1983), p. 189. 55 Grossmann (1992), p. 94. 56 Grossmann (1992), p. 114. 57 Grossmann (1992), p. 112. 53

100 “entities” that veridical acts intend, in that entities have properties and “everything that has a property ipso facto exists.”58 The lengths to which Grossmann went for the purpose of maintaining that intentionality is always a relation to an object (either existing or non-existing) led him into the problematic ontology of non-existent objects. How does a non-existent object differ from nothing?59 Non-existent objects supposedly add something to the universe, but what? Clearly, there is a difference between thinking of a mermaid and thinking of a unicorn, but if such non-existent objects do not have properties, how can such an ontology account for the obvious difference in thinking of each of them? Positing non-existent objects will not help, for how does a non-existent mermaid that does not have properties differ from a non-existent unicorn that does not have properties? Because Grossmann conceived of intentionality as requiring an object that, in veridical acts is an existent “entity,” or in non-veridical acts a nonexistent “object,” it follows that acts, in and of themselves, are not capable of intending. Rather, at least a partial explanation, or necessary condition, for the intentional capacity of acts, resides in the objects or entities to which they are related. Bergmann’s view in “Realistic Postscript” is very close to Grossmann’s, in that both hold that intentionality is not possible without each act having an object. Of course, the assertion that intentionality is not possible without each act having an object is ambiguous between: (1) every act representing an object; and (2) every act requiring a corresponding object, which is intended. The crucial difference between Bergmann and Grossmann on this point is that Bergmann was convinced in his later works that each act required an intention that exists on some level, wherein Grossmann thought it was possible for an act to be related to a non-existent object.60 Bergmann said, “In 58

Grossmann (1992), p. 113. If the intentional nexus can hold between an existent act and a non-existent object, then what prevents one from holding that there are relations between two or more nonexistent objects? Are there non-existent relations between non-existent objects? It does not seem that this approach is in the least helpful or illuminating. 60 Grossmann has clearly placed himself within the tradition of Twardowski and Meinong, and stated that Twardowski’s position was the correct solution of thought about that which does not exist (Grossmann (1977), p. XIX, (1983), pp. 197-200). The heart 59

101 of these views rests on two theses: first, “every act has an intention”; and, second, “intentions have properties and stand in relations irrespective of their ontological status” (Grossmann (1977), p. XI). Even though the content of the act determines what object is presented to the mind, all three of these thinkers held that there must always be an object that is intended. Some of these intentional objects exist, and some do not. Grossmann expressed his reasoning this way: “The notion of a relation with just one term is indeed absurd; a relation with just one term would be an ontological absurdity. But when we assert that the intentional relation can connect an idea with a nonexistent object, we do not imply that it is such a one-term relation and, hence, an absurdity. The intentional nexus is a two-term relation. But, and this is the crux of the matter, the entities which stand in this two-term relation need not both exist” (Grossmann (1977), pp. XIX-XX). This account raises the question of whether or not non-existent objects are literally nothing. If non-existent objects are literally nothing, then in non-veridical perception Grossmann is again faced with the problem of a one-term relation, which Grossmann claimed is “an absurdity.” If, on the other hand, a relatum in a relation can be filled by literally nothing, then it would be impossible to have a one-term relation, because everything would be a two-term relation, because everything could then be related to literally nothing. Again, an ontology of non-existent objects seems to raise far more problems than it solves: If literally nothing can be a relatum in a relation, are there relations in which both relata are literally nothing? If not, why not? If so, how many? Are there different kinds of literal nothings? How can one literal nothing be different from another literal nothing? How can a literal nothing be an account and explanation of anything? And so on. Further, Grossmann claimed that non-existent objects are not literal nothings. He wrote: “It is obvious that the golden mountain cannot be individuated by being localized in space and time. Hence it must consist, not of instances of properties, but of properties…Thus, while a real mountain is a complex of instances, the golden mountain turns out to be a complex of properties.we may ask how the golden mountain differs from the complex property of being golden and a mountain, since both of these entities presumably consist of the same properties; they both are conceived of as wholes which have the same properties as parts…[But, one concerns] the relationship between an individual thing and its properties and…the other [concerns] the relationship between a complex property and the properties of which it consists” (Grossmann (1977), p. XXI). This, however, seems to contradict the phenomenological data. When one thinks of the golden mountain or of Santa Claus, is one really just thinking of properties somehow linked together? Isn’t there a phenomenological difference between thinking of a string of properties such as being a man, being jolly, wearing a funny red outfit, having a large grey beard, riding in a sleigh pulled by flying reindeer, etc., etc., on the one hand, and thinking of Santa Claus on the other? What is necessary to form a “complex of properties”? Can any properties be strung together to form an object? Is that all the unity there is to an object? Grossmann’s account does not seem to explain how thinking of Santa Claus and thinking of a series of relevant properties are different acts with very different objects. Further, if a non-existent object, such as the golden mountain, is a complex property, and if a complex property such as being golden and being a mountain exists, then why

102 my world the intention of, say, a false belief [such as mistaking a real toy cardboard coin for a real metal coin] is a possibility (p-fact) and as such has ontological status (exists), even though it is neither actual nor…real.”61 He thought that there must be “some ontological status to the intentions of all acts,” and that “the mode of possibility provides the way out” when we are confronted with mistaken judgments about existing objects.62 Things get more complicated when dealing with cases of hallucination or delusion that are not simply cases of mistaking an existing object for something it is not. After idealism and phenomenalism, the greatest ontological and epistemological mistake “which undid the classical realist,” such as Moore, was that the external (what is outside the circle of ideas) was identified with the real. Following the logic of requiring an intention of some ontological status, Bergmann’s decisive move was to claim that “existential” cases such as seeing a ghost, entail that “some external particulars are not real,” and that “these [critical] particulars are not mindindependent; on the other hand, they are external, i.e., they are “in” nonmental facts.”63 From this it follows that these “External particulars which are not real are mind-dependent.”64 All acts, both veridical and nonveridical, require objects of some ontological status, but unlike veridical acts, non-veridical acts are distinguished by their objects existing only if the non-veridical act exists. In this sense, the critical particulars of such acts are mind-dependent. Bergmann remarked, “If…a is the particular prewould Grossmann deny that the golden mountain exists? Finally, isn’t there a difference between falsely thinking that the existent Santa Claus is not jolly, and correctly thinking that the non-existent Santa Claus is not jolly? It is not clear that Grossmann can distinguish the difference between these two thoughts. As Husserl remarked concerning Twardowski’s positing of non-existent but intentional objects, “But the theory does wholly unnecessary violence to the facts. Through it nothing is gained and nothing is explained” (Husserl (1994), “Intentional Objects,” p. 347). It is interesting to note that Twardowski, Meinong, the later Bergmann (to a great extent), and Grossmann put most of their emphasis on the objects of the act; whereas, Husserl, the early Bergmann, and Addis put almost all their emphasis on the existence and nature of the act of consciousness. 61 Bergmann (1964), “Realistic Postscript,” p. 308. Bergmann’s slightly awkward phrase, “has some ontological status (exists)” is quoted just as Bergmann wrote it. 62 Bergmann (1964), “Realistic Postscript,” p. 308. 63 Bergmann (1964), “Realistic Postscript,” p. 321, emphasis in original. 64 Bergmann (1964), “Realistic Postscript,” p. 322, emphasis in original.

103 sented to me when I saw a ghost, then, if the act, i.e., my perceiving the ghost, did not exist, neither would a.”65 In “Acts,” Bergmann was content that the phenomenon of illusion was explained—without remainder—by the intentional property (“proposition”) of each act that intends, irrespective of whether or not the object exists. But, by “Realistic Postscript,” Bergmann was convinced that such an account was not enough, and that there had to be an object of some ontological status to account for the intentionality of acts. In “Realistic Postscript,” Bergmann claimed that, because “the realism-idealism issue is a tangle of ontological and epistemological questions,” the only way to establish a robust realism is by detailing an ontology of acts and objects that accounts for what we know.66 The ultimate test of any philosophy of mind is whether or not it can adequately account for what we know about ourselves and the world. Comparing philosophy to a game of chess, Bergmann said that: A philosopher’s ontology is like the preceding moves [in a game]; his epistemology (ontology of the knowing situation), like the winning move. If his last move does not win, you have indeed no reason to ac67 cept the claims he makes for his ontology.

Competing ontologies can be evaluated only by determining which is the best all-around explanation of everything we know taken together. An analysis that accounted for everything except how knowledge of the external world is possible would be glaringly inadequate. Act, Object, and Metaphysical Realism Any view that can properly be called “realism” requires that the connection between acts and objects is neither one wherein acts create their objects, nor one wherein objects causally or logically depend on the acts that intend them. Further, if one’s analysis of the knowing situation cannot account for how it is that knowledge of objective, external reality is possible, then 65

Bergmann (1964), “Realistic Postscript,” p. 322. Bergmann (1964), “Realistic Postscript,” pp. 319-324. 67 Bergmann (1964), “Realistic Postscript,” p. 305. 66

104 such an account will be inadequate. Bergmann claimed that the whole issue turns on how thought intends objects. He said: (1) There are two kinds of things; some are physical; some, mental. Realists2 hold (1); idealists hold that all entities are mental. (2) Minds know physical things; the primary source of such knowledge is perception. (1) without (2) is an empty husk. Realism2 requires an assay of the knowing situation in general, and of the perceiving situation in particular that bears out (2). Or, briefly, realism2 stands and falls with a realistic2 account of perception. Such an account must connect a mind with its in68 tentions.

Mistaken views of intentionality leave us sequestered within the private confines of our minds and make it impossible for us to grasp public, external realities. Realism thus “stands and falls” on the issue of whether or not it can account for our epistemological position in the world. Before we even begin philosophizing, we find ourselves in the world knowing many things about many different kinds of things, and any philosophical view that cannot adequately account for this fact should be rejected as hopeless. Bergmann’s quest for the ontology of knowledge attempted to circumvent idealism by analyzing intentionality not as a relation of production between act and object, nor as one of causation between act and object, nor as one of logical entailment wherein the act entails the object. The intentional connection between act and object is not reducible to any other relation: it is sui generis. But is Bergmann’s account from “Realistic Postscript,” in which every act must have an intention of “some ontological status,” really the best way to achieve the full-bodied realism he sought? In what follows I will argue that even though Bergmann’s later account is founded on the M relation of intentionality, it requires that the ontology of veridical acts and non-veridical acts be quite different, not only because the kinds of objects differ, but also because the dependence relations between acts and objects differ. Veridical acts intend existent (real) objects.69 Often the occurrence of an act causally depends upon the existence of the object along with other 68

Bergmann (1964), “Stenius on the Tractatus,” p. 264. The claim that veridical acts intend existent objects is restricted to affirmative acts. Truth-makers must be real objects, but the truth-maker of an act may not be its object.

69

105 conditions. For example, veridical perception requires that the object be a partial cause of the act. However, this is not always the case. We can have veridical acts of objects that are causally inefficacious, such as of logical, mathematical, and geometrical truths. We cannot be causally connected to such objects, and acts of consciousness do not affect the nature or existence of these objects. There are mathematical and logical truths that have never been thought of. With respect to veridical acts, the act does not produce, cause, or entail the object. Consequently, the object is ontologically independent of, and therefore logically prior to, any act that intends it. As Aristotle remarked: The object of knowledge would appear to exist before knowledge itself, for it is usually the case that we acquire knowledge of objects already existing ...Yet it is equally true that, if the knowledge of a certain object 70 does not exist, the object may nevertheless quite well exist.

Veridical cases of thought and perception reveal one side of the intentional phenomenon and entail a necessary requirement for a realist account of objects. With respect to non-veridical acts, such as seeing a ghost, Bergmann must hold that the object cannot be a causal condition of the occurrence of the act, because if the act did not exist neither would the object. The core of Bergmann’s realism is that neither veridical nor non-veridical acts produce, cause, or logically entail their objects. In an important passage Bergmann wrote: The only connection between the two facts A [the act of perceiving] and B [the act’s intention, or object] is the formal fact BMB [the thought that B means B]….Since ‘BMB’ is analytic, B does not causally depend on A, even though…the awareness is simultaneous with the particulars in B….Let B stand for ‘a is white’. If a is real, there is no good reason for calling the connection a dependence of B on A. If anything, the occurrence of A depends causally, among other things, on B. If, however, a is the particular presented to me when I saw the ghost, then, if the act, i.e., my perceiving the ghost did not exist, neither would a. This is a good reason for calling a as well as the fact of its being white mindFor example, in “Santa Claus does not exist,” the truth-maker is the total condition of the world, but the object of the act is Santa Claus. 70 Aristotle (1941), Categories, 7b, 15-32, pp. 20-21.

106 dependent…. External particulars which are not real are minddependent. Unless some acts existed, they would not exist. There being such particulars is part of the intentional nature of mind. To see that there is nothing paradoxical about this one merely has to recognize the dependence for what it is. Then one will not be tempted to say of any act either that it creates its intention or that the latter either causally or logi71 cally depends on it.

But, is this account consistent with the claim that every object of such acts necessarily “has some ontological status”? Bergmann seems to be caught in a dilemma by claiming that in non-veridical acts, if the act did not exist, neither would the object. In such cases, which is ontologically and logically prior, the act or the object? On the one hand, if the act is logically prior to the existence of the object (of any ontological status), then the motivation to claim that every act must have an object of some ontological status is contradicted. If an act can exist without an existing object, then there is no reason to claim that every act must have an object of some ontological status. On the other hand, if the object of non-veridical acts exists prior to the act, then how do such objects differ from real and actual objects? If the object exists only if the act exists, and if each act requires an object that exists of some ontological status, then neither can be ontologically or logically prior to the other. If the critical particular exists, then the act exists, but if the act exists, then the critical particular must exist to provide “an ontological ground” between the act and the object. If every act requires an object of some ontological status, how, in the case of nonveridical acts, is it possible for the thought that-P to intend P, if P does not already have some ontological status? But, then again, how can P have any ontological status, if the thought that-P does not exist? If an act must have an existing object to grasp or, in Colin McGinn’s term, “lasso,”72 how can a non-veridical act intend when there is not already something there for it to grasp or lasso?73 71

Bergmann (1964), “Realistic Postscript,” pp. 322-323. McGinn (1991), p. 37. 73 This point illustrates how initially appealing, but ultimately inadequate, the typical metaphors used to describe intentionality are. If intentionality is literally unique, then there cannot be figurative terms that adequately distinguish it from everything else. One should not simply assume a metaphysics of mind from metaphors such as “inten72

107 Why, though, is it necessary for act or object to be logically prior? Idealism seems committed to the logical priority of the act insofar as the object is a part or property of the act. Furthermore, realism seems to require that the object is logically prior to the act, insofar as metaphysical realism maintains that whatever exists does so, and has the properties and relations it does, independently of deriving its existence or nature from being thought of or experienced. (Bergmann calls this “realism2” which is contrasted with idealism, whereas, “realism1” refers to the claim that universals exist and is contrasted with nominalism.) In other words, a necessary condition of metaphysical realism is that everything is what it is independently of its being an object of thought, perception, or reference. According to metaphysical realism the existence and nature of even thoughts and experiences are not comprised, even in part, by being objects of other thoughts or experiences. As Richard Fumerton remarked, “Let us say that the fact that P has representational independence if it is not constituted by any intentional state that has P as its object.”74 Finally, if what exists is not constituted, or derived in any way by being the object of a representation, then the representation of that thing does not change that thing. The problematic point concerning Bergmann’s later view is that it seems to require that neither act nor object be logically prior; or, in a peculiar combination of both idealism and realism, that both act and object be logically prior. But why call a view that so strongly depends on existent but non-real objects realism? Bergmann argued that “the awareness b is simultaneous with the particulars in B [the intention],” yet the only connection between b and B is the intentional connection.75 In non-veridical acts when b exists, B exists, and when b ceases to exist, B ceases to exist—yet there is no causal tional arrows”, “ideas pointing”, “thoughts grasping”, and “consciousness lassoing things”, as a substitute for evidence and argument. 74 Fumerton (2002), pp. 6-7. Thus, metaphysical realism, defined as I have here, is consistent with Leibnizian idealism or possibly Hegelian idealism in which everything is non-physical or somehow mental in nature; whereas it would be inconsistent with Berkeley’s idealism which holds that with respect to one set of objects, those known by sense experience, their existence and nature are derived from being perceived. Further, it is consistent with realism to claim that, as a matter of fact, it is possible that everything is the object of thought or experience; but from this it does not follow that anything is derived from, or constituted by, being thought of or experienced. 75 Bergmann (1964), “Realistic Postscript,” p. 322.

108 connection between the two. If realism requires such an intentional occasionalism, doesn’t this just mean that realism fails in the final move of the game? Realism is undermined by claiming that there is a necessary relation between an act intending an object and the existence of the object, and, although Bergmann always eschewed idealism, he seems here to be blurring the distinction between act and object by requiring that there be always a connection between the act and an object which “has ontological status (exists).” At the end of the nineteenth century, Cook Wilson remarked: Theories of knowledge and reality, in the futile attempt to explain apprehension...[often are] an overstatement of the interconnexion of the being of the object and the being of the apprehending thought; the tendency to reduce the object apprehended to terms of the apprehension of it, and the tendency to reduce the apprehension to terms of the object. Now neither can be reduced to the other; neither expressed or explained in terms of 76 the other.

Metaphysical and epistemological realism hold that the reality known exists independently of being known, and that what is known is as it is apart from the knowing process. The difficulty here is that, in non-veridical acts, there is a blurring of the act with the existence, however marginal, of the object. Bergmann was correct to argue that the object is never ontologically dependent on the act. The classical idealist made the object a part or a property of the mind, and Bergmann never swerved from his rejection of that. He said, we must maintain “the distinction between ‘seeing’ [the act] and the ‘seen’ [the object]. One who blurs this distinction is, we know, on the way to idealism.”77 Even though Bergmann always maintained that the object is ontologically independent of the act, his account seems to entail that, if there is an act, then there necessarily is an object that exists. This does not lead to classical idealism, but it does reveal an avoidable ontological blur between acts and the objects intended. The underlying issue here is that the problem of non-veridical acts cannot be solved simply by introducing new kinds of objects such as possibilities, or critical particulars that have “ontological status (exists),” yet fall short of 76 77

Wilson (1969), p. 76. Bergmann (1964), “Realistic Postscript,” p. 311.

109 being real or actual. Again, doesn’t a view that places so much weight on objects that exist but are not real depart from realism? Again, is realism really served by having a view that requires different levels or kinds of existence? What could it possibly mean to say that there are things that have “ontological status (exists),” yet are neither real nor actual? Different levels of existence are introduced only to distinguish the objects of non-veridical acts from those of veridical acts, but in Bergmann’s later thought it is not possible to think of that which does not exist because the object of every act must have some degree of existence. In such an ontology, is it possible to form the thought that “the non-existent dragon poses no danger,” if one can only think of some object that has some ontological status? If one has a hallucination of a ghost, does one really perceive something that does not exist? The Bergmann of “Realistic Postscript” must say “no.” Rather, such a perception is of something that exists, but is not real. Suppose that in a hallucination I perceive a ghost, which is white and larger than a bus. Given the principle of acquaintance, must one then hold that the critical particular actually is white and larger than a bus? I do not see how Bergmann can avoid saying “yes,” yet I am equally convinced that this outcome does not accurately represent the spirit of his passionate realism. Rather, it seems to me that the Bergmann of “Realistic Postscript” was in the grip of a theory—every intention “has ontological status (exists)”—that was pushing him away from realism, and away from his views in “Acts,” which do not suffer these problems. In their attempts to connect thought with reality, Berkeley, Moore, and Bergmann in “Realistic Postscript” made very similar mistaken identity claims. Berkeley’s error was identifying what is presented in perception with what exists. Moore’s error was identifying an idea (which, he argued, is already to be outside the circle of experience) with what is real. Bergmann’s error in “Realistic Postscript” was identifying what is presented with that which has some ontological status. Because all three overstated the closeness of the connection between thought and existing reality, they had no way to account for thought or perception of what does not exist. What is puzzling is why Bergmann thought that his account in “Acts” was inadequate. I suspect that he and Grossmann kept thinking of

110 intentionality as a straight-forward relation. Grossmann acknowledged that “The notion of a relation with just one term is indeed absurd; a relation with just one term would be an ontological absurdity.”78 However, he was so compelled by this account that he stretched the meaning of “relation” almost to the breaking point by holding that intentionality is an abnormal relation which, in some cases, connects a mind with a non-existent object. But, perhaps, thinking of intentionality as a genuine relation that always requires two existing relata (or with one relatum of some lesser ontological status) is not the best way to conceive of it. Intentionality is notoriously difficult to describe, but this, as Laird Addis remarks, “lies in its profound difference or apparent difference from everything else we encounter or know about in the universe.”79 If there is nothing else like intentionality, then it should not be surprising that our stockpile of descriptive words that are relevant, such as “relation”, “connection”, “pointing”, “referring”, “of”, “about”, “object”, “grasps”, and “between” must be used carefully, and with qualifications. On the one hand, intentionality often appears to be relational, because it appears to require an existent object. But, on the other hand, intentionality sometimes appears not to be relational, because it does not always seem to require an existent object. From the fact that intentionality is an essential reference or “connection” to an object, it does not follow that it is an existential connection to an object. Intentionality can be relational without being a relation between existents. (However, veridical cognition such as knowledge does require an existential connection.) In other words, the “ofness” commitment of the act does not imply an existence commitment. This unique feature of intentionality can be expressed in two different, but consistent ways: One can define “relation” as a two- or more-place property, which can, in these special intentional cases, have only one place that is instantiated with something that exists. Or, one can define “relation” such that all relations always require two (or more) existing relata, from which it follows that intentionality is not a relation. Admitting that intentionality is a relation defined in this way would, in cases of non-veridical cognition, imply that there can be an instantiation of a two-term relation by only one 78 79

Grossmann (1977), p. XIX. Addis (1989), p. 3.

111 term. This would be like claiming that someone is a father although there is nobody who is, or ever was, his offspring. While avoiding a semantic dispute about how to use the word “relation,” it is important to insist on the uniqueness of intentionality. This task requires two things: an accurate description of what intentionality does, and an ontological account that does not deny what it does. It is the main burden of this study to argue that Bergmann’s views concerning intentionality in “Acts” were essentially correct, and constitute a more nearly accurate ontological analysis of the intentional phenomenon than his later views. There is a necessary (analytic) relation between an act and its object, but from this it does not follow that there is a necessary relation between an act and its object existing in the world. Bergmann’s great insight with respect to the ontology of intentionality in “Acts” is that acts of consciousness have a sui generis property, by virtue of which acts of consciousness point beyond themselves. Intentionality intends something without thereby also necessarily connecting one to that something existing in the world as a real object or an actual state of affairs. Perhaps Bergmann’s distinctions can be filled out as follows: A: the thought that there is an Aspen tree in my yard, (a particular act) B: the proposition, or intentional property of A: being an Aspen tree in my yard, (a universal) C: an Aspen tree being in my yard, (the object, or the intention) There is a close connection between A and B, because B is a property of A. B is not the object of the act, but the means by which the act refers to C. Thus, a thought can, by virtue of exemplifying B, represent the character of C and, consequently, be about C, without C being a fact in the world. A thought can be of C, without an Aspen tree existing in my yard, because B is about that fact, whether or not there is an Aspen tree in my yard. Further, there is another close connection between B and C, in that the thought A exemplifying B intends C. The propositional character, B (being an Aspen tree in my yard), is exemplified by A; it is a genuine universal and, consequently, can be present in many acts. It is a signifying universal, and signifies a situation whether or not that situation exists, and whether or not it subsists. As a signifying universal, B denotes the same situation in each

112 instance of A, although each A might differ with respect to the species (doubting, perceiving, fearing, hoping, believing, etc.). This accounts for how different persons might each have a different A, which all, nevertheless, have the same B and are, therefore, each of C. This also accounts for how the same person can have many instances of A at different times, and each be of C. Every A is a token of being about C in virtue of being of the same type, or exemplifying the same signifying universal, B. Being an Aspen tree in my yard, is signified by the universal B, and this universal signifies what it does because of its nature alone, irrespective of whether or not there is an Aspen tree in my yard. Because the existence of C does not affect the intentional capacity of A, it follows that the capacity of A to intend C is not determined by C. This account of intentionality reveals how realism1 and realism2 are logically connected. Realism2, as opposed to idealism, maintains that objects are not constituted by being cognized. In order to maintain ontological independence between the existence of the act and the existence of the intended object (realism2), it is necessary that intentional properties, or specifying universals, exist (realism1), in order to account for both veridical and non-veridical cases of cognition. Bergmann’s account of intentionality in “Acts” is very similar to Husserl’s. Bergmann remarked: “Husserl built everything on the act. So do I.”80 Of course, there are significant and controversial differences, but the similarities are striking. Both held that the most significant feature of the mind, and that which makes the mind what we all experience it to be, is that the mind reaches out, by means of its acts, beyond itself. It does not, however, follow from the fact that acts reach out beyond themselves, that they necessarily reach out to the existent world. In other words, just because an act reaches beyond itself does not mean that its object exists as 80

Bergmann (1964), p. 194. In “The Ontology of Edmund Husserl” (1964), p. 193, Bergmann wrote of his close philosophical connection with Husserl: “There is reading and reading...Recently I spent several months reading Husserl [the Logical Investigations]. Naturally, I had read in him before, though not very much and not during the last fifteen years or so, while I worked out my own views. Now, having really read him, I am profoundly impressed by the greatness of his achievement. I also see how much of what slowly and painfully I have discovered for myself I could have learned from him.”

113 part of the external world. Different acts might intend objects outside the mind such as trees, or objects inside the mind such as other mental states, or that which does not exist in either place, or in any sense. Whether or not an intended object exists must be established by something other than a mere awareness of the object. Every philosophy must have a way of deciding which acts, both past and present, it is rational to regard as veridical. This does not necessarily require a definitive decision procedure, or a method that results in unconditional certainty. What made Bergmann and Husserl (at least the early Husserl) realists was their insistence that awareness is not limited to mental entities. If we are necessarily cloistered or “truncated” within the mental realm, as the idealists hold, and if what is presented is identified with what is real, then serious difficulties arise. In the words of Husserl: We understand what “round square” means, and so we have a representation. In that we now deny the existence of the round square, we have not denied the content of the representation, for of it we have an immediate knowledge…If a round square is immanent in the representation…then there would be a round square in the representation. The truth, “No round squares exist,” would no longer be valid in rigorous universality…Thus, there are round squares as often as “they” are represented. Since the representations actually exist, the existence of any and every 81 absurdity would also have to be fully and completely conceded.

Both the early Bergmann and the early Husserl were able to avoid such absurdities and maintain the distinction between what is presented within experience and what exists by rejecting the view that the mind is incapable of transcending its private mental states. Husserl made this point when he wrote: Representation is not the mere being-there (Dasein) of a content in consciousness—that is, its presence in the real context of psychical processes (Erlebnisse). Rather, representation is an intentional process, a certain minding (Meinen), whereby an object appears (whether it appears “itself,” or appears in image) or whereby an object is even just designated, 82 or meant via significations. [Such as mathematical symbolizations.]

81 82

Husserl (1994), “Intentional Objects,” pp. 349, 352. Husserl (1994), “Draft of a Letter by Husserl to Marty,” pp. 476-477.

114 His point, and Bergmann’s also, was that merely locating an object in the mind is not sufficient to represent that which is not a part or property of the mind. Locating a private object within the mind can never be an adequate substitute for grasping a public object outside the mind. Such a strategy is just another way of placing a veil between our acts and the world. The great advantage of Bergmann’s analysis in “Acts” over his later account is that it could handle cases of acts in which their objects do not exist in any sense whatsoever. Because an act of consciousness cannot be described without mentioning the object it intends, it appears that each act requires an existing object. If, however, an existing object is not readily available, such as when one hallucinates, there is always the temptation to posit some existing object, such as an image or a sense datum, to fill in as the object experienced. On the other hand, because consciousness of a particular object does not seem to require that the particular object exists of which one is conscious, there is a counter-balance to that temptation. Because consciousness was thought to require an existing object, many philosophers have taken this to be sufficient justification for positing stand-in mental or abstract objects to take the place of what someone mistakenly thought he or she was perceiving or thinking of. Whether or not the object of the act exists makes no difference whatsoever to the intentional capacity of the act. Non-veridical acts intend just as effectively as veridical acts. Edmund Husserl addressed the phenomenon of the existential independence of the object from the act when he wrote: I have an idea of the god Jupiter: this means that I have a certain presentative experience, the presentation-of-the-god-Jupiter is realized in my consciousness. This intentional experience may be dismembered as one chooses in descriptive analysis, but the god Jupiter naturally will not be found in it…But it also does not exist extramentally, it does not exist at all…If, however, the intended object exists, nothing becomes phenomenologically different. It makes no essential difference to an object presented and given to consciousness whether it exists, or is fictitious, or is perhaps completely absurd. I think of Jupiter as I think of Bismarck, of the tower of Babel as I think of Cologne Cathedral, of a regular thou83 sand-sided polygon as of a regular thousand-faced solid.

83

Husserl (1970), vol. II, pp. 558-559.

115 Whether the awareness in question is a perception makes no difference whatsoever. Suppose a patient complains to his doctor about hearing ringing in his ears. It would reveal a complete confusion on the part of the doctor if, after listening with her stethoscope on the patient’s ears, she should authoritatively proclaim that, because she can hear no ringing in his ears, he cannot be hearing any. Ringing “in” one’s ears can occur even if one no longer has ears, just as phantom pains can be felt “in” one’s foot even if one no longer has a foot. From the fact that one sees a tree, even in a clear and distinct manner, it does not follow that there exists a tree, which is seen. Because the existence of the object is independent of the intentionality of the act, it follows that the epistemic status of acts cannot be discerned simply by consulting these acts. In order to discern whether or not a particular act is veridical, one must be able to discern whether or not its object exists, and exists as intended. Further, because veridical acts intend as assuredly as non-veridical acts, justification, if possible, must be accounted for by means of something other than mental states considered as mere mental states. One cannot discern whether an act is veridical or nonveridical purely by means of another act, because the same problem will reappear with respect to the new act: is it veridical or non-veridical? If we cannot discern whether or not a particular act is veridical merely by examining that act, and if we have access only to acts considered simply as mental states, then we can never be in a position to know whether or not any of our acts are veridical. Consider the following argument: (1) If nonveridical acts are phenomenologically indistinguishable from veridical acts, and (2) if our epistemological access is limited to acts considered merely as mental states, then (3) “the epistemological problem” is irresolvable. If this argument is sound, then escape from the circle of ideas is impossible. On the other hand, if the epistemological problem is solvable, if we can have justified beliefs concerning the external world, then at least one of the two premises above must be false. The second premise seems the more vulnerable. If there are indirect awarenesses, then there must be something of which one is directly aware. The real question is whether or not direct awareness is limited to mental states. Justification can be attained by examining either the nature of the objects presented in acts, or how the acts cohere together, or by some combination of the two. Berg-

116 mann appealed to both of these approaches, but the foundational point here is that Bergmann gave an ontological account of the act that allowed direct access to the external world. Because of this, Bergmann, like Moore, Husserl, and Frege, found skepticism about the external world to be a complete non-starter. Such skepticism fails to describe accurately what we know about our access to the world. Bergmann often said that the skeptic is like a person who locks himself in jail, throws the key out the window, and then wonders why he cannot get out. Finally, Bergmann’s account in “Acts” avoids the problem of introducing philosophers’ objects to account for non-veridical perception. Introducing objects such as images, sense data, negative facts, possibilities, and propositions to supply an existing object for non-veridical acts simply reintroduces the problem of an open-ended act at another level. Anyone who holds that intentionality always requires an existing object should be troubled by non-veridical acts for they seem to be cases of acts without existing objects. If intentionality always requires an existing object, then in non-veridical acts, because the existing object is not what it seems to be (such as a ghost or a pink elephant), there must be some existing object or obtaining state of affairs that one is intending—but what? The putative objects for non-veridical acts that often are suggested are sense data, ideas and impressions for perception; images for memory and imagination; and propositions, negative facts, non-obtaining states of affairs, and possibilities for belief and judgment. If this account is correct, it should strike us as extraordinary that in non-veridical acts we are supposedly acquainted with objects such as sense data, ideas, images, negative facts, and the like that are nothing like the things we see in hallucinations such as unicorns, ghosts, or pink elephants. Adolf Reinach wondered how the object of an awareness could be recast into a different object solely on the basis of the existential status of the object intended. He wrote: Can the distinction between essence and existence be so far misunderstood that the denial of existence is confused with a modification of essence, of the essential characteristics? Concretely expressed: does a gigantic house of five floors, which I suppose myself to be perceiving, by

117 any chance become an experience when this perceiving turns out to be an 84 hallucination?

How can one confuse thinking of a gigantic house of five floors with thinking of a perception? If it turns out that the house one is thinking of (or perceiving) does not exist, it does not follow that what one was thinking of (or perceiving) was a mental fact. Such non-veridical acts must either intend non-existent objects, or there must be something within the intentional complex that remains “open-ended.” Something’s got to give. In other words, in non-veridical acts either the act intends what does not exist (in any sense), or the act must intend a non-veridical image, a non-veridical proposition, a nonveridical sense datum, or the like. But, once again, an open-ended act is forced upon us, for we would have a non-veridical image which is of nothing, or a non-veridical proposition which is about nothing, or a sense datum which is not of anything, or a possibility which is true of nothing. If an act requires an object to be intentional, then why doesn’t an image require something that it is an image of to be an image? Why doesn’t a proposition require an object of which it is about to be the proposition it is? It makes just as good sense to speak of perceptions, memories, or dreams that are not of existent objects, or of obtaining states of affairs, as it does to speak of viewing an image which is an image of nothing, or having a belief which is about nothing, or having a non-obtaining state of affairs which is about nothing. Images, propositions, possibilities, sense data, states of affairs, and the like are necessarily intentional, for they must be of or about something. On the other hand, if it is not necessary to concoct such obviously ad hoc objects to account for the intentional capacity of images, propositions and sense data, then there is no reason to require that every act have an existing object. If the universe is not overflowing with non-real objects and false facts to account for non-veridical acts, then one, if not the primary, reason for holding that there are such things as images, propositions and sense data disappears. Further, the introduction of such objects creates an unnecessary layer between the act and the world, and thus makes direct apprehension of the world somewhat problematic. 84

Reinach (1969), p. 198.

118 Attempts to avoid skepticism by including the object within the confines of the act, such as idealism or phenomenalism, ironically serve only to reduce most objects to what they obviously are not. The same mistake is made from the opposing viewpoint by requiring an existent object for each act. If this is so, then in non-veridical acts our beliefs and perceptions are of objects very different from what we normally take them to be about. The problem of non-veridical thought or perception about a common object, such as a physical object, cannot be solved by treating it as a veridical thought or perception of an uncommon philosophical object, such as a sense datum. Such confusions about what one’s experiences are about seem at odds with a thoroughgoing realism. Instead, by denying that intentionality requires an existent object, the ontological status of the objects of perception is logically independent of the ontological status of the perception of the object. Thus, no ontological conclusions can be drawn about the nature of the world merely from the nature of experience. Consequently, ontology is liberated from epistemology, for the object is not sequestered within the intentional act as a part or property. With this freeing of the object from the act, the philosophical ground is prepared to discover and explore the ontological status and various properties of the diverse objects of experience independently of the specific acts under which they are intended. Then it becomes possible to undertake a fruitful inquiry into how act and independent object come together in knowledge. Bergmannian Postscript Bergmann’s greatest achievement was his detailed analysis of the ontology of awareness and knowledge in defense of a robust realism. In this, he was the American counterpart to the early Husserl. Up to his time, no other philosopher in the English-speaking world had so carefully worked out the ontological foundations of metaphysical and epistemological realism. His wide-reaching work influenced a whole generation of ontologists, philosophers of mind, epistemologists, and philosophers of science. His writings formed a groundbreaking foundation of ontological analysis, which all subsequent serious students of intentionality and realism must take into account. On the last page of Logic and Reality, Bergmann wrote:

119 Philosophy…takes the world as it finds it. The most likely present threat to first philosophy is an unholy alliance among those who now propound these two opposite kinds of error [the Lebenswelt of certain interpretations of the later Husserl and the later Wittgenstein joined with scientism and materialism]. But, then, first philosophy, the fundamental dialectic, never attracted any but the happy few. Thus its position has always been 85 precarious. There is no more cause for alarm than usual.

But, perhaps, everyone would agree that he was mistaken on this point. Because of the life of Gustav Bergmann, there is less cause for alarm than usual.

REFERENCES Addis, Laird. 1989. Natural Signs: A Theory of Intentionality. Philadelphia: Temple University Press. Aristotle. 1941. Categories, in The Basic Works of Aristotle, ed. by Richard McKeon. New York: Random House. Bergmann, Gustav. 1954. The Metaphysics of Logical Positivism. Madison: The University of Wisconsin Press. Bergmann, Gustav. 1959. Meaning and Existence. Madison: The University of Wisconsin Press. Bergmann, Gustav. 1964. Logic and Reality. Madison: The University of Wisconsin Press. Bergmann, Gustav. 1967. Realism: A Critique of Brentano and Meinong. Madison: The University of Wisconsin Press. Chisholm, Roderick. 1973. The Problem of the Criterion. Milwaukee: Marquette University Press.

85

Bergmann (1964), “Realistic Postscript,” p. 340.

120 Fumerton, Richard. 2002. Realism and the Correspondence Theory of Truth. Lanham: Rowman & Littlefield Publishers, Inc. Grossmann, Reinhardt. 1965. The Structure of the Mind. Madison: The University of Wisconsin Press. Grossmann, Reinhardt. 1977. “Editor’s Introduction” in On the Content and Object of Presentations: A Psychological Investigation, by Kasimir Twardowski. The Hague: Martinus Nijhoff. Grossmann, Reinhardt. 1983. The Categorial Structure of the World. Bloomington: Indiana University Press. Grossmann, Reinhardt. 1984. “Nonexistent Objects Versus Definite Descriptions” in Australasian Journal of Philosophy, 62, p. 363-377. Grossmann, Reinhardt. 1992. The Existence of the World: An Introduction to Ontology. London: Routledge. Holt, Edwin; Walter Marvin, William Pepperrell Montague, Ralph Barton Perry, Walter Pitkin, Edward Gleason Spaulding. 1912. The New Realism: Cooperative Studies in Philosophy. New York: The Macmillan Company. Husserl, Edmund. 1970. Logical Investigations Vol. I & I. Translated by J.N. Findlay. London: Routledge & Kegan Paul. Husserl, Edmund. 1994. Early Writings in the Philosophy of Logic and Mathematics, Translated by Dallas Willard. Dordrecht: Kluwer Academic Publishers. McGinn, Colin. 1991. The Problem of Consciousness. Oxford: Blackwell Publishers. Meinong, Alexius. 1960. “Theory of Objects” in Realism and the Background of Phenomenology, ed. by Roderick Chisholm. New York: The Free Press. Moore, G.E. 1948. Philosophical Studies. London: Routledge & Kegan Paul Limited. Moore, G.E. 1953. Some Main Problems of Philosophy. London: George Allen and Unwin LTD. Moore, G.E. 1959. Philosophical Papers. London: George Allen and Unwin LTD. Reinach, Adolf. 1969. ‘Concerning Phenomenology’, translated by Dallas Willard, in The Personalist, Vol. 50, No. 2, Spring 1969.

121 Twardowski, Kasimir. 1977. On the Content and Object of Presentations: A Psychological Investigation. Translated by Reinhardt Grossmann. The Hague, Martinus Nijhoff. Wilson, John Cook. 1969. “Logic and Theories of Knowledge and Reality”, in Statement and Inference, Vol. I, Oxford: Oxford at the Clarendon Press.

Time for Bergmann’s Bare Particulars ERNÂNI MAGALHÃES West Virginia University

I

n many philosophical circles, a view is reduced to absurdity if it’s shown to imply that there are bare particulars. Quentin Smith’s view, which he defended in 2002,1 that past and future things lack the ordinary intrinsic properties they have when present, has been criticized for implying the existence of “nearly-bare” particulars.2 David Armstrong, without much of a look at the doctrine itself, endorses a principle of the rejection of bare particulars.3 And D. W. Mertz complains that the bare particular is the contemporary incarnation of Locke’s “I know not what” substratum, and that it lacks “content.”4 This, I would guess, is not among the philosophical circles in which ‘bare particular’ is a dirty word.5 Still, if Bergmann’s bare particulars are to do any constructive work, as I think they can, they must first be exonerated from the most common of the complaints against them. The reason bare particulars are in such ill-repute is that they are assumed to be bare. In Armstrong’s formulation, a particular is bare if it does not have any properties or perhaps even stand in any relations [loc cit.]. (I sometimes use ‘properties’ to include both non-relational and relational characteristics.) Because everything has some properties, or maybe because everything particular has some properties, this has been enough to impugn the bare particular. At one level, one cannot help but understand the objection. If one were only familiar with the phrase and not the idea it was used to express, it would be natural to infer that a bare thing is one that does not have the 1

Smith (2002). Zimmerman (1998), p. 212. 3 Armstrong (1978), p. 169. 4 Mertz (2001), p. 48. To be fair, Mertz, unlike the others, actually offers arguments for his rejection of bare particulars. 5 I refer to the group gathered for the Bergmann Centenary Conference. 2

124 relevant clothing. The relevant clothing in this case being properties, of course bare particulars would be the ones that do not have properties. But at another level those who impugn bare particulars on these grounds are guilty of wholly ignoring the actual idea that has been expressed by those who championed such things. It’s simply not true, as others have also noted,6 that bare particulars are bare in this sense. A bare particular is bare in the sense that it’s merely a particular and nothing else. It’s not, in other words, a property; nor is it something that involves properties as constituents. But it does involve properties in another highly relevant sense. So far from being propertyless, the bare particular is precisely that which has properties.7 It’s hard to fathom what would be the point of a particular that was bare in the critics’ sense. Why suppose that there was this particular entity out there that did not exemplify any properties? One could imagine an advocate of universals impressed by counter-examples to the identity of indiscernibles supposing a bare particular in each otherwise indiscernible object. So now two objects with all the same universals could be individuated in virtue of their brutely different bare particulars. But why stop there? Why not have the particulars that individuate those entities also be the exemplifers of their properties? Otherwise, one would have a miscegenetic marriage of the bundle theory and categorial dualism: objects as bundles of both properties and bare particulars. But even if bare particulars do not lack properties, isn’t there some other objection to them in this neighborhood? Isn’t it true that apart from its properties a bare particular has no properties? Isn’t that a problem? If “apart” means “considered apart,” so that the question is whether a bare particular considered apart from its properties lacks properties, then bare particulars are not propertyless apart from their properties. If I am considered apart from my money I am not thereby poor. If “apart” means “if it were not for,” so that the question is whether if it were not for its properties a bare particular would have any properties, the answer is somewhat more complicated but probably yes. Bergmann thought bare particulars could 6

The classic and still best critical discussion of bare particulars is in Loux (1979). This dramatic way of putting it is surely wrong, though, since properties have properties and enter into relations. In his elementarist moments, Bergmann does subscribe to the dramatic claim (1960), pp. 115ff. 7

125 not lack properties—even that it was unintelligible to suppose them to do so. If it’s just impossible and not outright nonsense that bare particulars lack properties, and counterfactuals with logically impossible antecedents are necessarily true, then so is the claim in question. If one holds that the antecedent is unintelligible, and thus lacks a truth-value, then the whole “statement” lacks a truth-value. Since I incline toward the view that it is impossible and not unintelligible for bare particulars to lack properties, the fact that bare particulars would lack properties if they had no properties, so far from being a reductio of accepting them, is rather an indication that they meet the minimal requirements of intelligibility. Every entity must be bare in the sense that if it had no properties it would have no properties. I believe this is the traditional picture of the bare particular, and it’s the conception of a bare particular I will consider. But I should mention that some of Bergmann’s apparently important claims about bare particulars do not obviously square with this conception. And they actually support the critics’ conception. For example, in “Ontological Alternatives,” Bergmann says that bare particulars are “only numerically different.”8 He also tells us, in Realism, that “the bare particular is a mere individuator,” and that structurally “[i]t does nothing else.”9 If bare particulars are only numerically different, then they are not qualitatively different. If bare particulars merely individuate, then they don’t also exemplify properties. The obvious reason they would not be qualitatively different is that bare particulars lack qualities. If bare particulars did have qualities, how could they fail to be qualitatively different? Maybe what Bergmann means is that the brute numerical difference between bare particulars is primary; that their qualitative difference is possible only in virtue of this brute difference. And Bergmann surely does hold that different particulars can differ qualitatively only because they are different in the first place. But this doesn’t strike me as a very natural interpretation of the ‘only’ in “bare particulars differ only numerically.” But recalcitrant though the textual evidence may be, I will assume that bare particulars do in fact have qualities and accordingly differ not only numerically but also qualitatively.10 8

Bergmann (1964), p. 133. Bergmann (1967), p. 25; see also (1964), p. 160. 10 Partly as a result of the comments about this paper at the Bergmann celebration, I’m less inclined to insist on calling these entities “bare particulars.” The expression is of9

126 So there may be problems with the concept of a bare particular, but obvious incoherence is not among them. Bare particulars may accordingly be considered as candidates for various philosophical jobs. The job I want them for is to be the individual content of time. I want bare particulars to be continuants and subjects of change. Change I will understand not in the Pickwickian sense of wholes made up of qualitatively diverse temporal parts located at different times, but in the sense that if something goes from being F to being not-F, the very thing that was F is the thing that is not-F. In terms of the contemporary literature, I want bare particulars to be the subjects of endurantist persistence and change.11 The alternative to endurance is perdurance—the sort of persistence that allows for what I derisively called Pickwickian change. Bergmann assumes persistence is endurance. He also thinks that endurance is possible only if there are absolute times— times construed as not reducible to the entities at them.12 I sympathize with the first claim; not so much with the second. Concerning the first, imagine something A is first F and then not-F. How can something be both F and not-F? The absolutist endurantist has an answer, Bergmann thinks: F is not, as it appears, a non-relational property of A but rather a relation between A and the absolute time, say, T.13 Not-F is a different relation also had by A, but there is no problem as long as A has F-at and not-Fat to different times. (It’s interesting that Bergmann sees this dialectical possibility long before its major public splash with D. H. Mellor’s 1980 Real Time.) Because Bergmann rejects absolute times, he rejects endurance and with it persistence. But could a relativist about times—the one who thinks times are reducible to located entities—think of changing properties as reten associated not only with what I take to be its central feature—its mere particularity—but with various of the other characteristics Bergmann associated with those entities: being momentary, lacking parts, and not belonging to kinds. 11 The problems of persistence and change are related. Something persists just in case it exists from one time to another. Something exists from one time to another only if it’s capable of changing. Endurantism, though often construed as a theory of persistence, may be best understood as an account of persistence and change. 12 Perhaps a more faithful interpretation is that for Bergmann, given a “subjectpredicate world”, change is only possible given absolute times. See “Some Reflections on Time,” (1960), pp. 230-231. 13 Bergmann (1960), pp. 230-231.

127 lations to times after all? I don’t think so. Suppose the relativist thinks that the fact that A is F at T is understood as the fact that A is F while B is G and C is H and D is J…. Given this, it’s hopeless to think of F as the relation is-F-while. Even if there is such a relational property, it’s plainly not fundamental. The spirit of this relativist line is to take A’s being F and B’s being G, etc., as the relevant primitives, connect them through simultaneity and so build up times. By the relativist’s lights, A’s having the relation being F-while to B’s being G is reducible to A’s being F and B’s being G, which are connected through simultaneity. The same sort of problem attends the idea that the fact that A is F at T is to be reduced to the fact that A’s being F belongs to the collection of A’s being F, B’s being G, etc. To think that there is yet the fundamental relation being-F-by-belonging-to (a certain collection) is to take what is plainly derivative for something primitive. What is really there on this conception is A’s being F, B’s being G, etc., which are then connected to each other by belonging to a certain collection. So it does seem hopeless to take changing properties of enduring objects to be relations to relativistic times. Fortunately for genuine persistence and change, the endurantist need not be a relationist—relationism being this view that changing properties are relations taking times as terms. Faced with the puzzle above that a changing A is both F and not-F, the endurantist may just insist that A is not both F and not-F. Either A is F and was not-F, or was F and (earlier) was not-F, or—you get the picture. The point is to insist that tense matters; that endurantist change is coherent precisely because tense matters. Bergmann would reject this alternative because he rejects tense. The sentences of his ideal language, in which the world is supposed to be most accurately pictured, are not sensitive to context in the way tensed sentences clearly are.14 As I see it though, a language without tense would not get the world quite as it is. One of the very interesting things about the world is that it’s changing all the time. A language that was not sensitive to this fact would be thereby a less than ideal language. Interestingly, the relationist—again, the person who thinks of changing properties as relations to times—who accepts universals can be shown to be committed to bare particulars. This may be seen by thinking about 14

Bergmann (1964), pp. 179-80.

128 just which properties the relationist should take to be relations to times. It’s not promising to take all and only changing properties to be relations to times. (A changing property is not one that changes but one with respect to which something changes.) Whether F is a changing property is a contingent fact. It’s a contingent fact, for example, that green is a changing property. It could have turned out that some things were green but that none of them changed with respect to being green. If a property’s being a relation to a time is determined by its being a changing property, and it’s a contingent fact whether something is a changing property, then it’s a contingent fact that properties are relations to times. So it’s a contingent fact whether a certain property is a relation or not. Since that isn’t a contingent fact, it can’t be that whether a property is a relation to a time is determined by whether it’s a changing property. Bergmann has the absolutist who believes in continuants holding that all properties of ordinary objects are relations to times.15 Although this choice raises the same problem I’m about to point out, perhaps the best route for the relationist endurantist is to hold that all changeable properties are relations to times, where a changeable property is one with respect to which something can change. Since it’s plausibly a necessary fact whether a given property is changeable, it will turn out to be a necessary fact about a property that it is (or is not) a relation to a time. All this is important because there is some question how this kind of endurantist will individuate his continuants. If our changing or changeable properties are relations to times, what accounts for our being different entities? I assume that individuation cannot be accounted for in terms of relations or relational properties. On my proposal, as long as some properties are not changeable, there might be some hope of non-relational qualitative individuation in these immutable properties. But the hope is small since whatever the immutable properties of continuants are, they’re surely such as to be capable of being shared by several continuants at the same time. And if this endurantist is a realist about universals, he won’t have the materials to individuate his continuants. Well, at least those materials won’t be found among properties. But this endurantist can appeal to non-qualitative bare particulars to individuate otherwise qualitatively identical entities. So even for the relation15

Bergmann (1960), pp. 230-231.

129 ist endurantist with certain commitments, it looks as if bare particulars are a helpful posit.16 But, as I said, I’m interested in bare particulars as the continuants of serious tensing endurantism, not the relationist endurantism just considered. I don’t see, admittedly, anything in the bare particular that especially recommends it for this job. On the other hand, it seems to me there are independent reasons to think that bare particulars are as good as it gets as an answer to the question what keeps the characteristics of individual objects together at a time. So my main concern is whether there might be some reason bare particulars can’t also be enduring continuants. One complaint against this move is that I am no longer even talking about bare particulars; as Bergmann describes the idea in Realism, “persisting bare particular” is an oxymoron.17 This raises the question what is central and what is peripheral in the concept of a bare particular. I take as central that bare particulars are only particular and that they exemplify properties. All else is peripheral. I assume it is coherent to take a bare particular to persist and to change. (Those who find it oxymoronic to talk of persisting and changing bare particulars may replace that expression with whatever they think is more appropriate.) I also assume that it is possible for a bare particular not to be simple. This is also a bit sacrilegious since, as far as I know, Bergmann consistently held that bare particulars are simple. Holding that they may be complex leaves open the possibility that the most obvious examples of persisting entities are genuine. One question that has daunted endurantists concerns the exact conditions in which something survives. Just what changes can something undergo? Can a human become a minotaur? Can a fetus become a human? Can a ship lose all its original boards and acquire all new ones? It may be thought that these questions are particularly awkward for the endurantist with bare particulars for continuants. Not having a nature, it might be argued, there is nothing in the “nature” of a bare particular to explain the kinds of changes it can or can’t survive. I’m not sure what to say in the final analysis, but there are several things I feel confident I can say. One is to suggest that, coming from some 16

Thanks to Adam Friedlander for helping me see this problem more clearly. Also, see K. Hawley (2001 and 1998). 17 Bergmann (1967), p. 117.

130 objectors at least, this complaint is hypocritical. It’s admittedly hard to state non-trivial conditions under which something does persist through time. But this is just as hard for perdurantists as it is for endurantists. I believe the problem is hard for any of them, but consider the worm theorist, just to fix our focus. The worm theorist says something persists by having diverse stages that have the appropriate relations of resemblance, causal connection, etc. Therein lies the rub: how much resemblance? Do some resemblances matter more than others? What kind of causal connection? How much? It’s notoriously difficult for causal theorists of all kinds to distinguish in a principled way the “right” from the “wrong” kind of causal connection. Even those who say persistence is a matter of convention face the analogous problem of discovering just what our conventions are about these matters. So as far as the problem of criteria is concerned the perdurantists and endurantists are in the same boat. But this is little solace if the boat we’re all in is sinking. My hunch is that there are no non-trivial (henceforth I drop this qualification) sufficient conditions for persistence.18 The most obvious reason for this is that persistence involves identity and there are no sufficient conditions for identity. I do think there are necessary conditions for the persistence of something. If B at T1 is A at T2, i.e., now, then whatever properties A has are properties B has, and vice versa.19 Thus persistence is governed by the indiscernibility of identicals. But since this is a serious tensing version of endurantism, it’s important to watch the tense. It’s not necessary in order for B at (the earlier) T1 to be A now that A have whatever properties B had. Nor is it necessary that B have every property A had. If these were necessary conditions for persistence, change would be impossible. Significantly, I don’t see any reason why bare particulars could not have the sort of connection B and A in the schema are supposed to. This necessary condition unfortunately yields little or no guidance in deciding whether the minotaur is the human, whether the adult is the fetus, 18

Compare Merricks (1998). But maybe this is not in the final analysis a non-trivial necessary condition for persistence. Notice that the condition itself asserts the identity of some entities, i.e., properties. Still, the condition is significant insofar as it shows the connection between the identity of A and B in terms of the identity of some items other than A and B. The paradigm of a trivial necessary condition would be one which required that in order for some entity to be F it must be F. 19

131 or whether the ship with the new boards is the ship with the old boards. It would be nice if one of the necessary conditions for persistence were something more substantive. Let me consider two such substantive conditions: spatio-temporal continuity and kind membership. Promisingly, spatio-temporal continuity would help rule out several possibilities and it fits nicely with bare particular continuants. Bare particulars are plausibly spatially located, and although in my opinion not strictly temporally located, yet the temporal continuity condition may be understood to require that a bare particular exist at continuous times. The main problem with spatiotemporal continuity is that it doesn’t strike me as a plausible necessary condition for persistence. Why couldn’t my mom pop out of existence for a while and then come back? Or why couldn’t she go from one place one moment to a place very distant the next? A kind restriction on persistence is similarly helpful with problem cases. The idea that once a K, always a K should help us figure out whether humans can become minotaurs and whether fetuses can become humans. If kind conditions do yield only necessary conditions for belonging to certain kinds then the help will be limited. Perhaps in order to be a K something must have DNA of a certain kind. But since at any given time lots of things will have that certain DNA, it will require further information to determine whether some K at T1 is still around at T2. More seriously, a kind condition seems to conflict with something apparently at the heart of Bergmann’s conception of bare particulars. Since they are supposed to lack natures, at least in the sense of essential properties, it’s hard to see how they can be fitted to belong to kinds. Are bare particulars still bare if, as I want them to, they exemplify essential properties? Actually, it’s not clear that kinds on this view need be essential properties in this sense. F is an essential property of A if A has F in every possible situation in which A exists. The kinds in question are not, or need not, be essential in this sense. A kind must be connected with the particular in such a way that if A is K at T, then A is K whenever it exists. I don’t see that having a property immutably entails having it essentially. So supposing they belong to kinds may not quite imply that bare particulars even have essential properties. And having immutable properties does seem compatible with what I take to be the heart of the concept of the bare particular—that it’s simply a particular. Unfortunately, and somewhat anticlimactically, I can’t accept

132 kind conditions because I can’t bring myself to believe that the persistence of things is restricted by their belonging to kinds.

REFERENCES Armstrong, D. M. 1978. Universals and Scientific Realism. Cambridge: Cambridge University Press. Bergmann, Gustav. 1960. Meaning and Existence. Madison: University of Wisconsin Press. Bergmann, Gustav. 1964. Logic and Reality. Madison: University of Wisconsin Press. Bergmann, Gustav. 1967. Realism: A Critique of Brentano and Meinong. Madison: University of Wisconsin Press. Hawley, Katherine. 2001. How Things Persist. Oxford: Oxford University Press. Hawley, Katherine. 1998. “Why Temporary Properties Are Not Relations between Objects and Times,” Proceedings of the Aristotelian Society 98: 211-216. Loux, Michael. 1978. Substance and Attribute. Dordrecht: D. Reidel. Merricks, Trenton. 1998. “There Are No Criteria of Identity over Time,” Noûs 32: 106-124. Mertz, D. W. 2001. “Individuation and Instance Ontology,” Australasian Journal of Philosophy 79: 45-61. Smith, Quinton. 2002. “Time and Degrees of Existence: A Theory of ‘Degree Presentism’,” in Time, Reality and Experience. Royal Institute of Philosophy Supplement: 50, ed. Craig Callender. Cambridge: Cambridge University Press. Zimmerman, Dean. 1998. “Temporary Intrinsics and Presentism,” in Metaphysics: The Big Questions, ed. D. Zimmerman and P. van Inwagen. Oxford: Blackwell Publishing: 206-220.

Bradley’s Regress: Meinong versus Bergmann FRANCESCO ORILIA Università di Macerata

Introduction

W

hat philosophers keep discussing under the heading “Bradley’s Regress” stems from this well-known passage in Bradley’s Appearance and Reality: We may take the familiar instance of a lump of sugar. This is a thing, and it has properties…It is, for example, white, and hard, and sweet; but what the is can really mean seems doubtful.…Sugar is obviously not mere whiteness, mere hardness, and mere sweetness;…for its reality lies somehow in its unity....The word to use, when we are pressed, should not be is, but only has.…The whole question is evidently as to the meaning of has; and, apart from metaphors not taken seriously, there appears really to be no answer... [A possible answer is:] ‘There is a[n independently real] relation C, in which A [a thing] and B [a property] stand; and it appears with both of them.’ But here again we have made no progress. The relation C has been admitted different from A and B, and no longer is predicated of them. Something, however, seems to be said of this relation C, and said, again, of A and B. And this something is not to be the ascription of one to the other. If so, it would appear to be another relation D, in which C, on one side, and on the other side, A and B, stand. But such a makeshift leads at once to the infinite process. The new relation D can be predicated in no way of C, or of A and B; and hence we must have recourse to a fresh relation, E, which comes between D and whatever we had before. But this must lead to another, F; and so on, indefinitely. Thus the problem is not solved by taking relations as inde1 pendently real.

No attempt will be made here to unravel what exactly Bradley intends to argue for in this and related passages (see on this, e.g., Candlish 2006), although I should mention that he is commonly taken to provide an 1

Bradley (1893), Ch. 2, pp. 16-18.

134 attempt at a sort of reductio ad absurdum of the thesis of the reality of relations, whatever this precisely means in his philosophical framework. I shall rather concentrate on a certain way of considering Bradley’s regress, namely as an argument or family of arguments constituting a challenge for those who accept states of affairs (or facts).2 In a paper of 1899, Meinong appears to confront a similar issue. Interestingly, however, the Austrian philosopher’s attitude toward it is rather different from Bradley’s: …a complexion is a relation together with its members. This must not be taken to mean that the complexion is merely the relation and its members. That would basically be the same as the...objective collective consisting of a, b and r. Rather, a and b stand in the relation r which can only mean that also a and b respectively, each stand in a relation to r, that a relation, that is, a relation r' or r" respectively which could also be the same. It is very clear that what we just said concerning a, b, and r can be repeated concerning a, r, and r' and also concerning b, r, and r". In this way, new relations can appear without end. As far as I can see, there are no theoretical difficulties hidden in the preceding statement, as there are also none in the thought that the division of a line leads to an infinite se3 ries of always smaller partial lines.

Just as for Bradley’s quotation, I do not intend to provide any historical exegesis of this text. I simply take it as an invitation to discuss states of affairs, without taking for granted that Bradley’s regress is vicious.4 This seems to conflict with Bergmann’s opinion on the matter, as is illustrated by the following passage, where Bradley’s regress is referred to as a paradox to be avoided: Take two spots; one red and round, the other green and square. Red and round are “tied” together. So are green and square. Red and square are I shall use “state of affairs” and “fact” interchangeably. Meinong (1899), p.147. 4 Thus, I shall not be deterred here by the consideration that it is not clear to what extent we can attribute states of affairs, as here understood, to Meinong. Obviously he acknowledges objectives, but these are rather like propositions, since they can be subsistent or non-subsistent just like propositions can be true or false (more on the distinction between states of affairs and propositions below). Note also that, whereas Meinong distinguishes the two relations r' and r", we shall be concerned with a single (exemplification) relation. In considering them Meinong might have in mind something like the relations corresponding to what linguists call thematic roles, such as agent, patient, beneficiary, and the like. I think that an account of them may be postponed for present purposes. For my view on the matter see Orilia (2000). 2 3

135 not. Nor are green and round. That is why there are two spots and not four and why these two are what they are. That shows two things. First. There is a sort of entities which are constituents of others and yet so “independent” that a complex is more than, as one says, the sum or class of them. Entities of this sort I here call things [categorematic]. This is of course a very special use of the word. Second. The something more, the “tie” which makes a complex out of the class, must have an ontological ground. Everything except sameness and diversity must have an ontological ground. One who does not understand that does not understand the task and nature of ontology. The entity or entities which are the grounds of the “tie” I call subsistents [syncathegorematic].… Categorematic and syncategorematic are two modes [of existence]…. Among subsistents the distinction simple-complex does not even make sense. Nor do syncathegorematic entities need further ones to connect them with the categorematic entities they connect (Thus Bradley’s paradox is 5 avoided).

Do the above passages really indicate a contrast between Meinong, as I take him here, and Bergmann?6 And if so, who is right, if either? In my attempt to answer these questions, it will emerge that we should distinguish two versions of Bradley’s regress and correspondingly two accounts of the unity of states of affairs. One is Bergmann’s (and Strawson’s) nexus internalism, as we may call it. The other is an approach closely related to what Meinong seems to have in mind in the quoted text and which I have dubbed fact infinitism. States of Affairs States of affairs are to be understood here as complexes standardly representable by formulas such as “P1a” or “Rna1 ... an” (to be taken here not as sentences as in first-order logic, but as complex names designating facts). They are complexes in that they have constituents or “parts” which occur (are present or contained) in them, although not in the mereological sense in which we say that a wheel is a part of a bicycle. A state of affairs repreBergmann (1964), pp. 128-129. Actually, in Realism, a masterwork dedicated in large part to a close scrutiny of Meinong’s views, Bergmann explicitly criticizes what Meinong claims in the above passage (1967, p. 348). He attributes the fact that, according to Meinong, the regress is harmless to his taking relations to be particulars. See Bonino (2006) for a discussion of this. Be this as it may, here we shall try to see whether we can make use of Meinong’s positive attitude towards a Bradley-type regress in a framework where relations are—as in Bergmann—universals.

5

6

136 sented by “P1a” is taken to involve as constituents a property, P1, and an entity, a, instantiating the property, e.g., a given lump of sugar’s being white (hard, sweet,...). A state of affairs represented by “Rna1 ... an,” in turn is taken to involve as constituents an n-adic relation, Rn, and n entities, a1, ..., an, standing to each other in the relation Rn, e.g., a given lump of sugar’s being on a certain table (I shall usually drop the superscript indicating the degree or “adicity” of the property or relation, when it is not particularly important to be explicit about it). The property or relation in question, P1 or Rn, is assumed to be a universal and the entities, a, a1, ..., an, are typically particulars, although they need not be. I neglect the possibility that they are not, unless it is specifically relevant for our concerns (see “Problems for the Hypotheses (A), (B) and (C)” below). We can call facts of kinds Rna1 ... an and Fa, relational (n-adic) and non-relational (monadic), respectively. In the following, I shall typically consider an arbitrary non-relational state of affairs, Fa, with the understanding that what I say is generalizable to relational ones. A fact Fa has in a sense mereological constituents, if a has mereological parts, but this is not relevant for present purposes; hence, by “constituent,” we mean here “non-mereological constituent.” Fa has at least F and a as such constituents, which we may call canonical, for we should not rule out at the outset, for reasons that we shall see, that there are other, non-canonical, constituents. At least in typical cases (wherein F is a contingent property such as a certain skin color for a person), Fa is a nonsupervenient complex (with respect to F and a); viz., it is possible that F and a exist, but Fa does not exist. To put it otherwise, we can imagine a world in which F and a exist (for there are, e.g., the facts Fb and Ga), without the existence of Fa. For instance, although there are the facts Bs and Wm (Stevie Wonder’s being black and Mick Jagger’s being white), there could instead be the facts Bm and Ws, were it the case that S.W. is white and M.J. is black. In such a circumstance, the properties B and W, and the individuals s and m would still exist but the facts Bs and Wm would not exist. Since they are non-supervenient complexes, states of affairs must be distinguished from propositions as they are understood nowadays at the end of a theoretical path springing from Bolzano’s propositions, Frege’s thoughts, Meinong’s objectives, Moore’s propositions and the propositions of Russell’s Principles of Mathematics (1903). Leaving aside immaterial details, we can say that, given any property P and an object x, there exists the proposition that predicates P of x, understood as a complex with P and

137 x as constituents. We could represent it as P(x), where parentheses are displayed to distinguish it from the state of affairs Px. This proposition may be untrue (it could be a non-subsisting objective in Meinong’s terminology), but nonetheless part of the ontological inventory, thereby granting a meaning to the sentence “x is P.” In contrast, the state of affairs Px might or might not exist (at least in the typical case in which P is a contingent property). For a correspondentist in truth theory, the existence or nonexistence of the state would make the proposition true or false, as the case may be. In sum, whereas propositions are supervenient on their constituents, states of affairs (typically) are not such. Now, once a state of affairs is admitted in the ontological inventory, its unity, that it exists as a complex in addition to its constituents, requires, at least prima facie, an ontological explanation. This is the problem of the unity of states of affairs.7 It is in confronting this issue that Bradley’s regress dangerously comes to the fore for the friend of states of affairs, for it seems to show that a certain way of addressing the problem leads to an infinite regress of the kind evoked in Bradley’s above quotation, a regress that is typically seen as vicious. Accordingly, some contemporary philosophers take Bradley’s regress to show that there are no states of affairs after all.8 In doing this they consider facts not as entities that can be observed, but as theoretical entities postulated in order to play certain roles in a philosophical framework, notably the role of truthmakers for true sentences or the propositions that true sentences express.9 It may sound strange that states of affairs are just theoretical entities. After all, to take a typical Bergmann-style example, when we see a red spot, we see the spot with a color, and thus, one might argue, the state of affairs consisting of the spot’s being red. But the foe of states of affairs can reply by noticing that we can give a different analysis of what we observe, for the red in question can be viewed as a particular, or “particularized” property, what is now usually called a trope (a perfect particular in the terminology of Bergmann 1967), an entity that cannot be “shared” in the sense in which universals are typically shared by different objects and that can therefore be “found” only in the spot in question.10 This dialectical exchange corresponds pretty much to Bergmann’s11 contrast between a thing ontology, where particulars Vallicella (2000, 2002, 2002a). Simons (1997), Dodd (1999). 9 Wetzel (2003). 10 Armstrong (1989). 11 Bergmann (1960), p. 84. 7

8

138 and universals are “things,” i.e., primitive simples out of which states of affairs are constructed, and a fact ontology where “facts” are taken as primitives. Bradley’s Regress Revisited Note that the regress arises in connection with a certain (hypothetical) fact, which can be taken to be its origin. In relation to the above passage from Appearance and Reality, the origin is the fact consisting, say, in a certain lump of sugar’s being white. In relation to Meinong’s passage quoted above, the origin is a’s being in the relation r to b. In the following formulation of the regress by Loux, the origin is the fact Fa: According to the realist, for a particular, a, to be F, it is required that both the particular, a, and the universal, F-ness, exist. But more is required; it is required, in addition, that a exemplify F-ness, ...a relational fact. But the realist insists that relations are universals and that a pair of objects can bear a relation to each other only if they exemplify it by entering into it. The consequence, then, is that, if we have to have the result that a is F, we need a new higher-level form of exemplification (call it exemplification2) whose function is to insure that a and F-ness enter into the exemplification relation. Unfortunately, exemplification2 is itself a further relation, so that we need a still higher level form of exemplification (exemplification3), whose role is to insure that a, F-ness and exemplification are related by exemplification2; and obviously there will be 12 no end to the ascending levels of exemplification that are required here.

As this quotation clearly shows, the Bradleyan argument invites us to postulate, step after step, newer and newer relations, which we shall call E1, E2, E3, etc., where the letter “E” reminds us that they are exemplification relations (where Loux deploys exemplification, exemplification2, exemplification3, I find it more convenient to use E1, E2, E3, etc.). Let us concentrate on the generic fact Fa, taken as origin of a regress. Following Loux’ reasoning, we can see that at the step where E1 is introduced, call it step p1, a fact E1Fa is hypothesized. It contains as constituent, beside F and a, E1 as well. Having started from Fa, the relation E1 which we introduce at step p1 is obviously a dyadic exemplification relation, which we can then also represent as “E2.” In general, we can imagine a sequence of steps such that, at every step pi a new relation Ei is introduced, i.e. (i + 1)12

Loux (1998), pp. 38-39. The “realist” in Loux’s quotation is the realist with respect to universals, for whom, traditionally, Bradley’s regress constitutes a problem.

139 adic exemplification, Ei+1. At step pi a fact Ei .... E2E1Fa, with constituents Ei, ...., E2,E1, F, a is also considered. Accordingly, the reasoning generates, so to speak, a sequence of (hypothetical) states of affairs, whose first member is the origin Fa and whose i+1th element is the fact Ei .... E2E1Fa introduced at step pi. This (potentially infinite) sequence may also be aptly called a (Bradley’s) regress (the context can tell us whether by this term we mean an argument or a sequence). It is convenient to assume that the origin Fa of such a sequence could also be represented by “E0a,” where E0 coincides with F. We leave it open for the time being whether the states of affairs postulated at each new step of a Bradleyan argument are ultimately all identical to the origin Fa. If this were so, Fa should be taken to have as constituents not only F and a, but also the non-canonical E1, E2, etc. F could be called a canonical constituent with a prima facie attributive role and a a constituent with argument role. Having admitted the possibility of the above non-canonical constituents, we must then also leave open the possibility that a prima facie attributive constituent is not the really attributive constituent of the state in question and that it is instead just another argument beside a. This possibility obtains when, e.g., Fa happens to be identical to a fact E1Fa wherein there is a non-canonical constituent E1 which is prima facie (and perhaps really) attributive. Call faithful a representation of a fact such as those we have seen, Fa, E1Fa, etc., inasmuch as there is in them a symbol for every constituent. For example, if the only constituents of Fa are F and a, then “Fa” is a faithful representation of Fa. On the other hand, if there is in Fa a non-canonical constituent, e.g. E1, so that Fa could more explicitly be represented as “E1Fa,” we shall say that “Fa” is not a faithful representation of Fa. Before we proceed, it is worth being clear on how we should understand the infinity and viciousness allegedly involved in Bradley’s regress. The argument, as we have seen, allows us to postulate an infinite sequence of (representations of) states of affairs, which, as I see it, should be considered as vicious insofar as admitting it leads to an unacceptable consequence, i.e., a self-contradictory or at least totally implausible thesis. Otherwise, the sequence had better be considered harmless, if the consequence is acceptable after all, albeit not in an obvious and/or particularly palatable way, or even trivially harmless, if the consequence is just an unsurprising trivial result. Correspondingly, Bradley’s regress, qua argument, can be considered as either (trivially) harmless, or as vicious. The viciousness in this case should be seen as “directed,” so to speak, to a certain assumption

140 that the argument presupposes, insofar as the argument is taken to show that it is precisely that assumption that leads to the postulation of the vicious sequence. In Bradley’s opinion, at least according to many interpreters, the argument is vicious with respect to the thesis that relations are real. Similarly, the foes of states of affairs take it to be vicious with respect to the very existence of state of affairs. Two Versions of Bradley’s Regress According to Loux, the first member of a Bradley’s regress (Fa, in his example) is a fact that is different from its second element (E1Fa); and more generally each En .... E1Fa is different from the successive element En+1En .... E1Fa. This is evident from the way Loux continues his presentation: ... it is reasonable to think that once the realists have told us that a is F because a and F-ness enter into the relation of exemplification they have completed their explanation of the fact that a is F. There is of course something new that the realist might want to go on and explain—the new [emphasis mine] fact that a and F-ness enter into the relation of exemplification; however, the failure to explain this new fact would seem to do nothing to jeopardize their explanation of the original fact that a is 13 F.

In this account, the difference between the nth element S of a regress with first member Fa and the subsequent one S' is due to the fact that S' contains an additional constituent En+1 not contained in S and thus external to it. Moreover, S is a state of affairs wherein En is the constituent with prima facie attributive role and En-1, ..., E1, F, a are the corresponding arguments. On the other hand, it is En+1 which has the prima facie attributive role in S'. Loux thus provides, at least as I interpret him, what we could call an externalist Bradley’s regress, i.e., an argument according to which the relation En introduced at step pn is external, in the sense explained, to the state of affairs postulated at the preceding step. From this point of view, the expressions “Fa,” “E1Fa,” “E2E1Fa ,”...are all faithful representations of different facts, whose prima facie attributive constituents coincide with their really attributive constituents. Typically, however, we find in the literature an internalist Bradley’s regress, i.e., an argument which, in inviting us to postulate an ever growing number of exemplification relations, E1, E2, E3, ..., considers them all as 13

Loux (1998), p. 34.

141 constituents of the origin Fa, in such a way that the sequence, Fa, E1Fa, E2E1Fa, ..., corresponding to these postulations, is such that its members are all mutually identical. In fact, we could say that at every step pi, for i > 0, it is postulated that Ei relegates Ei-1 to the argument role beside Ei-2, ..., E2, E1, E0 (= F), a, and proposes itself as the really attributive constituent of Fa. In general, this assertion implies that EiEi-1 ... E2E1Fa is identical to the preceding member of the sequence, Ei-1 ... E2E1Fa. We would then simply have a multitude of different unfaithful representations “Fa,” “E1Fa,” “E2E1Fa,”..., which correspond to just one fact. These representations purport to display more and more efficiently the presence in the state of affairs Fa of the non-canonical constituents E1, E2, etc. From the perspective of someone who analyzes Fa, these exemplification relations add themselves, with prima facie attributive role, to the canonical constituents F and a, in an attempt to keep them together; but each of them is immediately demoted from its hypothetical attributive role by the “next” exemplification relation and thereby definitely confined to an argument role. Here are some examples of this way of seeing the matter (In Hochberg’s quotation I have replaced his notation for exemplification with mine): There are many ways of construing the problem posed by Bradley. One way...is the following. We take an atomic sentence ‘Ga’ to state that a particular has or exemplifies a property. The existence of the indicated fact—the particular exemplifying the property—is the truth condition for the sentence. The fact is taken to consist in the particular, a, and the property G, in the exemplification relation, E2. But, supposedly, it cannot be so taken. For there must be a further constituent: a three-term exemplification connection that obtains of G, a, and E3.…But then there must be a further constituent connecting these four constituents, and so on. The supposed problem can be taken to be that to acknowledge E3 as a constituent of the fact that a is G is to prohibit allowing one to specify the factual truth condition for the sentence ‘Ga,’ since it is not the fact consisting of a and G in the relation E2. For that purported fact turns out to be [italics mine] a fact containing the three-term connection that, in turn, must be construed as a fact containing a four-term relation and so 14 on. ... [I]t is not enough simply to say that the state of affairs of a’s being F is a unity. This unity must be explained....One thing is for sure: the metaphysical cement cannot be provided by introducing another universal, the relation of instantiation, to hold between a and F for a vicious regress is threatened instantly....How are, a, F, and the instantiation universal uni14

Hochberg (1989), p. 75.

142 fied? If we introduce a further universal, a relation holding between pairs of particulars and universals and the relation of instantiation, the same question is begged. And so things will continue as we introduce further entities to try to glue the constituents together. Unity cannot be 15 provided by adding another item to the complex [italics mine].

Given that there are two versions of the regress, the use of the singular, “Bradley’s regress,” is not completely appropriate, but we can speak in the singular in saying something that holds generally for both versions, as I have done above.

The Request for an Ontological Explanation One could think that the exemplification relations introduced step by step in a Bradley’s regress with origin Fa should be understood as constituents not so much of facts E1Fa, E2E1Fa, etc., but of propositions E2(F, a), E3(E2, F, a), etc. After all, the fact Fa presumably corresponds to some true sentence “a is F,” which in turn expresses as meaning a proposition F(a). Moreover, one could plausibly hold that, in relation to any proposition G(x), logic postulates an infinity of logically equivalent propositions E2(G, x), E3(E2, G, x), etc., just as it postulates an infinity of propositions, it is true that P, it is true that it is true that P,..., logically equivalent to a given proposition, P. Hence, why shouldn’t we think that in a Bradleyan regress we postulate erroneously at every new step new facts E1Fa, E2E1Fa, etc., or an increasingly complexity in one fact, simply because we are conditioned by the existence of the propositions E2(F, a), E3(E2, F, a), etc.? If it were so, we could say that there is just one (relatively simple) fact Fa to which all these propositions correspond, just as only this fact corresponds to the propositions: it is true that F(a), it is true that it is true that F(a), etc. Bradley’s regress would be in this case trivially harmless in that it would simply be a reminder of the existence of an infinite series of innocuous tautologies. This proposal has been advanced by Armstrong.16

15 16

Dodd (1999), p. 150.

Armstrong, (1997), pp. 18-19. Something similar has been proposed by Hochberg (1989). Perhaps the same could be said of Russell (1903) § 99, Orilia (1991), and Gaskin (1995) were it not for the fact that these authors discuss the problem of the unity of propositions rather than of states of affairs. It seems to me that Hochberg discusses with some ambiguity both states of affairs and propositions.

143 But this does not work, for, as we noted, and as is evident from the above quotations, Bradley’s regress is triggered by the request for an explanation, which stems from the problem of unity. Given the fact Fa, the Bradleyan argument presupposes that we should explain why we are in a world that contains Fa over and above F and a, since F and a could have existed without Fa. Clearly, this explanation cannot rely on supervenience. If Fa supervened on F and a, the explanation could (perhaps) simply be based on supervenience: there is Fa, because there are F and a and Fa supervene on them. But since a typical state of affairs does not supervene on its constituents, this will not do. It must also be emphasized that the requested explanation is not of the causal brand. If it were, we could simply answer that Fa exists because another fact caused it. For example, if Fa is the fact consisting of Socrates’ eating, we could say that Fa exists, because it was caused by a fact such as Socrates’ being hungry. Obviously, the requested explanation is of another kind. We could call it ontological. The philosophical tradition has always acknowledged ontological (non-causal) explanations beside the causal ones. For instance, there are alternative explanations of the existence of concrete particulars (over and above that of universals or classes) along these lines:17 (i) there are universals in a compresence relation; (ii) there are substrates in which universals inhere; (iii) there are entities localized in specific portions of space. These explanations are not meant to be alternative, but rather complementary, to causal explanations such as those based on protons, electrons and the like. In general, given a (legitimate) request for an explanation of a proposition P, one should answer with another proposition Q, which is taken to be linked to the former by an explanatory relation, expressible with words such as “because” or “in virtue of .” If the answer is correct, Q explains, at least partially, P. Let us use the symbol “⇐” to represent this relation. Thus, “P ⇐ Q” means “P in virtue of Q” or equivalently “P because Q,” “P is an explicandum of Q,” “Q is an explicans of P.” A formal account of this notion can be found in Tatzel’s (2002) attempt to reconstruct Bolzano’s analysis of the concept expressed by “because.” Rather than “P ⇐ Q,” Tatzel writes (p. 16) “Q > P,” but he means essentially the same thing.18 He interprets this expression as follows: Q is a partial or complete 17

Loux (1989). Note however that Tatzel uses his symbol as expressing a relation that connects a collection of propositions, rather than propositions. Where he talks of a collection of propositions {P1, ... , Pn}, I would talk of a conjunctive proposition P1 & ... & Pn.

18

144 mediate ground of P. The disjunction “partial or complete” regards the fact that in asserting “P because Q” we do not usually think we have said all that must be said in an attempt to clarify P, although we do not rule out that this could be the case. The term “mediate” has to do with the fact that in asserting “P because Q” we do not rule out that there could be intermediate explanatory elements in between P and Q: P because Q, since there are propositions P1, ..., Pn such that P1 = Q, Pn = P and, for i = 1, 2,...n, Pi is an immediate (partial or complete) ground of (immediately explains)19 Pi+1. Clearly, the relation ⇐ is transitive just like entailment, but it must not be confused with it. For example, the former, but not the latter, can connect only true propositions and is irreflexive and asymmetric.20 The relation ⇐ is generic in that it could be either causal or ontological. An index “c” or “o” could be added, depending on whether the explanatory link is of the former or of the latter variety. Thus “P ⇐o Q” means that Q is an ontological explicans of P and “P ⇐c Q” that Q is an causal explicans of P. In the following, the index “o” may be considered as implicit, when it is clear that we are dealing with ontological explanation. Similarly, I shall typically say “explicans,” as elliptical for “ontological explicans.” Before closing with these technical details, let us define explanatory chain, for it will be useful below. By that we shall mean a set of propositions ordered by the relation ⇐. If C is an explanatory chain whose only members, P1,..., Pn, are such that P1 ⇐ P2, P2 ⇐ P3,..., Pn-1⇐ Pn, then C will be represented by “P1 ⇐ P2 ⇐ P3 ⇐ ... ⇐ Pn-1⇐ Pn.” At least as regards ontological explanation, it seems appropriate to say that Q is not an explicans of P if there is a possible world wherein P is true, but Q is not. For instance, if we explain the proposition that there are concrete particulars, by asserting that there are universals in a compresence relation, we cannot admit that there is a possible world wherein there exist no concrete particulars and yet there are universals in a compresence relation. And thus we assume: (N)

19 20

(P ⇐o Q) → ~◊(~P & Q).

See Tatzel (2002) for details on this notion. Tatzel (2002), p. 17.

145 Facts: Ontologically Brute Versus Explainable In sum, Bradley’s regress presupposes this assumption: (BBA) Bradleyan Basic Assumption. If Fa exists (over and above F and a), there must be a proposition P, not based on supervenience, such that: the proposition that Fa exists ⇐o P; i.e., there must be a P which provides an ontological explanation of the existence of Fa in addition to its “parts” F and a, without it being a proposition that merely asserts that, given F and a, Fa must also exist for it supervenes on F and a. Could not we reject (BBA), by arguing that states of affairs supervene on their constituents? In order to answer positively, we should admit that after all there are in the ontological inventory, albeit as non-subsistent, states of affairs which we prima facie would not include in it, e.g., Bm, Mick Jagger’s being black. Such states of affairs would correspond to false sentences in a correspondentist account of truth. Perhaps this theoretical move is present in Wittgenstein’s Tractatus. In this way, states of affairs become indeed supervenient on their constituents (given F and a, Fa must also be given, although perhaps as merely non-subsistent), but would be hardly distinguishable from Meinong’s objectives and thus from propositions.21 And at any rate this move would not be sufficient to reject (BBA), for it would lead us to accept states of affairs (or similia) of the type S’s subsistence, where S is a states of affairs. A state such as this cannot be considered as supervenient on its two constituents, subsistence and S. In fact, we must distinguish, e.g., worlds like ours wherein Wm but not Ws subsists (Mick Jagger, but not Stevie Wonder, is white) and worlds 21

As a matter of fact this view is explicitly proposed by Bergmann (though with a different terminology, for he uses “subsistent” in another way, as we have seen in the above quotation). According to him (1964), p. 136, there are in the ontological inventory merely possible facts “with the lowest ontological status of all,” which are what in the world corresponds to false sentences. Hence, Bergmann’s facts are more like propositions (as they are understood here, supervenient on their constituents) than like states of affairs (as they are understood here, non-supervenient on their constituents). I neglect this here and treat Bergmann’s facts as if they were non-supervenient on their constituents. This is motivated by the following observation. As we shall see in detail below, Bergmann, by introducing the nexus of exemplification, attempts to provide a solution to the problem of the unity of states of affairs which is intended to work independently of the thesis (attributable to him although stated here in my terminology) that states of affairs are supervenient on their constituents. Or at least Bergmann’s proposal can be so understood and has been so understood; Armstrong (1978), p. 109, Vallicella (2000), p. 241.

146 wherein Ws, but not Bs, subsists (S.W., but not M.J., is white). In the end we must admit any way states of affairs that do not supervene on their constituents and thus we simply postpone the problem of finding an ontological explanation of the existence of non-supervenient states of affairs. Moreover, the non-subsistent states of affairs are hardly needed for the correspondentist theory of truth, for false sentences can be simply taken to be sentences that fail to correspond to any fact. I shall thus assume that there are no non-subsistent states of affairs and that states of affairs, if they exist, are not (in typical cases) supervenient on their constituents. In spite of this, we cannot take for granted that (BBA) be accepted. As far as causal explanations are concerned, we readily admit that there may be “brute facts,” i.e., that some propositions can contribute to explain other propositions, without themselves be explainable in a causal sense. For instance, the propositions asserting the existence of certain fundamental forces may be taken to have this status in modern physics. Analogously, we should perhaps admit that there are “ontologically brute facts,” that is, that some true propositions cannot be explained in an ontological sense. For example, some nominalists, rather than acknowledging universals, view in this fashion the propositions that point to objective resemblances among particulars.22 Similarly, one could argue, we could hold that the existence of each state of affairs is “ontologically brute.” This means rejecting (BBA) and opposing to it the Brute Fact Approach: (BFA) There are facts, but, for a typical fact Fa, there is no proposition P (either based on supervenience, or of some other type) such that: Fa exists ⇐o P. We find this proposal in Van Inwagen.23 I shall not attempt to argue at this juncture in favor of either (BFA) or (BBA). I simply emphasize that both versions of the regress implicitly oppose the former and presuppose the latter, but differ in how they propose to provide the ontological explicans which (BBA) invites us to look for. A Geography of Possible Approaches If we set (BFA) aside and accept (BBA), there must be, given that Fa exists, a corresponding explicans (not based on supervenience). Presumably, 22 23

Varzi (2005), p. 63. Van Inwagen (1993), p. 37.

147 it asserts something of a certain entity, U, which we could call a unifier (or “connector,” in Vallicella’s terminology). To wit, it asserts that U brings about (in a non-causal sense) that F and a are unified in a complex characterizable as a state of affairs. Such an explicans can be represented as P(U, F, a). Alternatively, the explicans could assert the same thing not of a single unifier, but of a plurality of them, U1,..., Un, in which case the explanatory proposition is representable as P(U1,..., Un, F, a). For convenience, we shall speak in the singular and say “unifier,” even though it is not ruled out that there are many unifiers. There are various extant theoretical options regarding the nature of this explicans, which can be grouped, following Vallicella, on the basis of three possible hypotheses regarding the “location” of the unifier: (A)

Internalism. The unifier is a constituent of Fa.

(B)

Identity Hypothesis. The unifier is Fa itself.

Externalism. The unifier is not a constituent of Fa and it is not Fa.24 As regards (A), we should distinguish six ways of articulating it, namely (A1)-(A6), below:25 (C)

(A1) What makes Fa an entity that exists over and above F and a is the presence in Fa of dyadic exemplification, E2, which connects F and a, so that Fa = E2Fa. In sum, since by the above conventions E2Fa = E1Fa, Fa exists ⇐ Fa = E1Fa. (A2) What makes Fa an entity that exists over and above F and a is the presence in Fa of all the exemplification relations E2, E3, E4, etc., so that the explicans of Fa is something like this: for each n ≥ 2, En connects all the exemplification relations of degree less than n, as well as F and a, in such a way that Fa = ... E4E3E2Fa = ... E3E2E1Fa. In sum, Fa exists ⇐ Fa = ... E3E2E1Fa. (A3) The unifier of Fa is an exemplification nexus, N, which is not a relation. N is a non-canononical constituent of Fa that is neither an argument 24

Vallicella does not consider the possibility of a plurality of connectors, if not perhaps on the side when discussing McTaggart in (2002a), nor the possibility that a fact could have, e.g., an external and an internal unifier. I shall neglect this hypothesis, since as far as I know it has not been proposed and it does not seem plausible. 25 Vallicella does not distinguish between (A4) and (A5) and does not consider (A6). Otherwise, he presents the options offered by (A) analogously.

148 like a, since it succeeds in tying F and a, nor a prima facie attributive constituent, for otherwise Fa would be identical to NFa and consequently this position would be scarcely distinguishable from (A1) and the nexus N from the dyadic exemplification relation. In sum, Fa exists ⇐ N is a constituent of Fa in such a way that it ties F and a without being a relation that obtains of F and a. (A4) the unifier of Fa is the universal F itself, which succeeds in the unifying job because it is unsaturated. The idea is that universals, if “abstracted” from states of affairs, are “incomplete” or with “holes” and thus in themselves unsaturated. As such, they are not “objects” capable of being arguments in a fact. They can occur therein only with an attributive role, i.e., as saturated by an argument. In sum, Fa exists ⇐ F cannot exist if not as saturated by an object and a is the object saturating it. (A5) The unifier of Fa is F, which succeeds in the unification because universals are not only capable of occurring as arguments in a state of affairs, but have also a capacity, which only universals have, of occurring in a fact with an attributive role, a capacity that F exercises in Fa. We could also say, in the terminology of (A4), that in exercising this capacity F is saturated by a, provided it is clear that in this option one does not rule out that properties can occur in a state of affairs as saturating (as arguments) rather than as saturated (as attributes). (A6) The unifiers in Fa are both F and a, since both of them should be viewed as unsaturated, i.e., incapable of existing if not as constituents in a state of affairs. And thus, Fa ⇐ F and a are mutually saturated. Clearly, (A1) is implicit in the internalist version of Bradley’s regress. (A2) is attributed by Vallicella to McTaggart.26 As we shall see in detail, Bergmann defends a version of (A3) and the use of “nexus” in this context is his. The other major contemporary philosopher whose name is 26

Vallicella, (2002a), p. 209. However, according to Broad (1933), p. 85, as I understand him, McTaggart endorses a version of Armstrong’s idea, mentioned above, according to which the regress is trivially harmless, since it consists of a series of propositions tautologically implied by a proposition asserting the existence of a fact such as A’s being related to B by R: “McTaggart admits that there is this endless series in connexion with any relational fact, but he denies that it is vicious. His answer amounts to saying that the first term, i.e., that A has R to B, is a fact in its own right, and that the rest of the series consists merely of further consequences of this fact.” (But see also note 45, below.)

149 typically associated to this theoretical option is Strawson,27 but Armstrong could also be listed.28 Thesis (A4) can be attributed to an ontologist who accepts states of affairs and is influenced by Frege, who, as is well-known, speaks of concepts as unsatuarted. However, (A4), stricto sensu, is not held by Frege himself, since, in short, he does not acknowledge states of affairs and views concepts as functions, which yield truth and falsehood when saturated by “objects.” Moreover, it should be noted that, according to Frege, concepts, despite their unsaturatedness, can saturate other concepts, provided the latter are of a higher order. Some find (A4) in Wittgenstein’s Tractatus, by arguing that, according to it, relations can exist only as “relating,”29 i.e., in my terminology, as attributes in a fact. Thesis (A5) could be attributed to an ontologist who, in line with the early Russell of Principles of Mathematics, thinks that universals differ from particulars in that the former, but not the latter, can occur in a fact in two different ways. The first, as logical subject (as argument, in the terminology adopted here), is common to all entities; the latter, as concept or verb (in an attributive role in my terminology), belongs only to universals. According to (A5), a state of affairs is brought about, when a universal occurs in this way with respect to a certain argument (is saturated, in Fregean terminology; Russell speaks of an “indefinable way” of occurrence of the universal). Thesis (A6) could be attributed to a philosopher (perhaps Sellars 1962) who, inspired by § 2.03 of the Tractatus, extends the Fregean metaphor of the necessity of a saturation to all the constituents of a state of affairs. It is not obvious that (A6) is attributable to Wittgenstein, because it is not clear that his states of affairs have universals as constituents and because he seems to admit non-subsistent states of affairs (which we rule out here). Hypothesis (B) seems to be proposed by Armstrong, when he claims that “States of affairs hold their constituents together in a non-mereological form of composition.”30 Accordingly, (B) results in this proposal: (B1) The connector of Fa is Fa itself and the explicans for its existence is: Fa holds F and a together. In his above-mentioned writings, Vallicella has argued for (C) and in particular for the following thesis: 27

Strawson (1959), Ch. 5, § 2. See Armstrong (1978), p. 109, where the idea is traced back to Duns Scotus and Aristotle. 29 See Bonino (2001), and Read (2005). 30 Armstrong (1997), p. 118. 28

150 (C1) The connector of Fa is a transcendental subject (identifiable with God for a theist) capable of judgments (or similia) which concretize or realize, so to speak, states of affairs. The explicans then is: there is a transcendental subject who performs a judgment, call it *Fa*, which realizes Fa. But in relation to (C), we should also contemplate a thesis implicit in the externalist version of Bradley’s regress: (C2) What makes Fa an entity that exists over and above F and a is the state of affairs E2Fa, understood as different from Fa, in that E2 is taken to be the really attributive constituent of the former, whereas F is taken to be the really attributive constituent of the latter. In sum, since, by the above conventions, E2Fa = E1Fa, the proposal here is that Fa exists ⇐ E1Fa exists, where Fa ≠ E2Fa. Once we accept the existence of facts and (BBA), I do not see other theoretical options, beside (A1)-(A6), (B1) and (C1)-(C2). Problems for the Hypotheses (A), (B) and (C) In this section all the theoretical options that we have associated with (A), (B) and (C), except (A3) and (C2), will be analyzed and found to be highly problematic. Let us start with (A1). It is fairly clear that, in seeing Bradley’s regress as vicious, Bergmann has in mind the internalist version of it, understood as a tool that can be deployed to reject (A1), that version of internalism according to which the unifier is a relation of exemplification, E1, internal to Fa. Bergmann is right in thinking that the argument can be successfully used against (A1). We can see this as follows. With a trivial induction it can be shown that the following proposition is true: (IBR) Internalist Bradleyan Regress. If (A1) is true, then, for any number n > 1 and fact Fa, Fa = E1Fa = E2E1Fa =....= En ... E2E1Fa. Thus, (A1), via (IBR), forces us to postulate ad infinitum non-canonical constituents, E1, E2,..., in every state of affairs Fa and by the same token forces us to acknowledge that there is no finite number n that is the number of non-canonical constituents of Fa. For suppose there were such a number n. Then, Fa would be identical to En... E2E1Fa. But it would also be different from En+1En... E1Fa, which contradicts (IBR). Accordingly, it

151 seems wrong to say, as we should in accepting (A1), that there occurs in Fa one constituent, E1, which brings about the feat of “unifying” F and a so as to make Fa existent. For why shouldn’t we attribute this role to E2, E3, etc.? This question suggests (A2) as an alternative to (A1). Obviously, (A2) is in conflict, just like (A1), with the basic intuition that a simple fact like Fa must have a finite number of constituents. For (A2) asserts explicitly what (A1) implies, namely that the relations E1, E2, E3, etc. are all constituents of Fa. But could not we bite the bullet and admit that every fact has an infinite number of constituents? This means admitting an actual infinity, but, after Cantor, this need not be seen as problematic.31 However, there is a more serious problem: (A2) implies that no constituent of Fa is really attributive. Proof: if there is a really attributive constituent in Fa, this must be either F, or some exemplification relation. But it cannot be F, because this property is demoted to the argument role by E1. Suppose then that, for some n > 0, the exemplification relation En is the really attributive constituent of Fa. This is impossible, since En is demoted by En+1 to the argument role. Yet, in a fact we must distinguish between attribute and arguments. It is one thing is to say that dyadic exemplification is interesting and another thing that John exemplifies wisdom. Intuitively, the truthmaker of the former assertion is a fact with the property of being interesting as attribute and dyadic exemplification as argument, whereas that of the latter is a fact with dyadic exemplification as attribute and wisdom and John as arguments. But according to (A2), counterintuitively, in every state of affairs Fa, all the exemplification relations, and F with them, occur after all as arguments, just like, one would want to say, dyadic exemplification in the truthmaker of dyadic exemplification is interesting. Let us now move to (A4). It is hardly tenable that the simple presence of a universal in a state of affairs is sufficient to explain the existence of the state, for it must be admitted that a universal can occur in a fact not only as attribute (Kim Basinger is beautiful), but also as argument (beauty is desirable). Moreover, as urged by Armstrong 1997, one should acknowledge states of affairs working as laws of nature and consisting of a relation (as attribute) obtaining of two universals (as arguments). Let us then shift to (A5) (those who are not convinced by the above objection to (A4) should notice that the forthcoming objection to (A5) holds against (A4) as well). There are two ways of understanding the idea that a prop31

Vallicella (2002a), p. 209.

152 erty is saturated by an entity, i.e., that its hole is filled by the entity in question. Either we say that the hole is necessarily capable of “attracting” any entity (of a certain category, e.g., particulars) into it, or it is not. Let us examine the first horn of the dilemma. Granted that the fact Fa exists, and thus that a (a particular) has the property F, this property could be a contingent property, so that there is a possible world w without the fact Fa. Now, since F attracts necessarily any particular into its hole, even in w F attracts a into its hole. Hence, it is possible that F attracts a into its hole without the existence of Fa and thus by the principle (N) of “The Request for an Ontological Explanation” above, contrary to what (A5) asserts, the proposition that F is saturated by a is not an explicans of the existence of Fa. Let us then consider the second horn, according to which, at least for some contingent property F and some entity a, it is possible either that F is saturated by a, or that it is not. Analogously, we would say that it is possible either that Fa exists or that it does not, or that a exemplifies F or that it does not. In sum, “F is saturated by a” seems only a picturesque way of expressing either the proposition to be explained, that Fa exists, or other (allegedly) explanatory propositions already to be found in other theoretical options, such as (A1) and (C2), where one already appeals to exemplification. Hence, either (A5) is false, because it falsely says of a certain proposition that it is an explicans of Fa, or else is only verbally different from other options that we are already considering and thus it can be put aside. Mutatis mutandis, (A6) can be rejected in the same way. Hypothesis (B1) is based on an obscure metaphor. It uses the expression “holds together,” which taken literally should be used with respect to ordinary objects in a way that involves causality. For example, the different parts of a bicycle, its wheels, pedals, etc., are held together, in a causal sense, by various screws. But it does not make sense to say that an ordinary object, as a whole, holds together the parts that compose it, for it would be like saying that the whole precedes the parts, as if the parts could not exist if the whole did not already exist. This is impossible, as it implies the absurd thesis that wheels, pedals, etc. cannot exist unless a bicycle already exists. In other words, in order to understand the sentence “Fa holds F and a together” we must rely on a metaphoric interpretation of it, an interpretation in which we treat (i) “Fa” as if it stood for an ordinary object such as a certain bicycle; (ii) “F” and “a” as if they stood for, e.g., the wheels and the pedals of the bicycle in question and (iii) “holds together” as short for “causally holds together.” But this means that, either we do not understand “Fa holds F and a together” or we should take it to be false

153 just like “the bicycle holds its wheels and pedals together.” In sum, (B1) seems to be either meaningless or false, unless we take it to be a less explicit way of saying that the existence of a fact cannot be explained ontologically and that accordingly (BBA) should be dropped in favor of (BFA). Let us now turn to (C1). Prima facie, it is a very implausible thesis, for it appears like an invitation to kill an ant with a gun. But this is not a conclusive objection, if all the rival options have serious difficulties; and Vallicella has written interesting pages in an attempt to support it. Yet, I think that (C1) faces a more fundamental problem. If we accept (BBA) and we are looking for an ontological explanation of the existence of Fa, we cannot appeal to a transcendental subject who performs the judgment *Fa*, for this sounds pretty much like an illicit passage from an ontological dimension to a causal dimension wherein we say that a state of affairs exists because it is caused by another state of affairs. For example, if we view the transcendental subject as God, should not we say that (C2) amounts in the end, crudely put, to the thesis that the fact Fa exists, because there is the fact that God has decided to create a world that contains Fa? In conclusion, it seems to me at this point that all we can do before giving up the hope to provide an ontological explanation of the unity of facts and embrace (BFA) is to devote a serious consideration to (B3) and (C2). Nexus Internalism Thesis (A3) could aptly be called Nexus internalism. Let us see how Bergmann comes to embrace it. He sees very well that Bradley’s regress presupposes two fundamental tenets that he is most willing to make: the rejection of nominalism understood as the doctrine that denies that there are properties and relations and the acceptance, in his terminology, of a “thing ontology,” and thus of states of affairs qua complexes. Once these concessions are made, one is naturally driven to appeal to a relation of exemplification internal to a state of affairs to account for the problem of the unity of facts, that is, one is tempted by (A1), which involves (i) that exemplification32 is a relation and (ii) that it is a constituent of the state of af32

More precisely, as we have seen, we should speak of exemplifications of various degrees. On the assumption that we take a monadic fact such as Fa as our starting point, we may speak of “exemplification” in the singular, and refer to dyadic exemplification.

154 fairs in question. It is because of both (i) and (ii) that (A1) falls prey to the internalist version of the regress, which Bergmann correctly takes to be vicious. However, the viciousness for Bergmann is directed to (i) and thus in his hands Bradley’s argument becomes a sort of reductio ad absurdum of this assumption and exemplification a “nexus” rather than a relation. Bergmann makes this point very clearly in “Frege’s Hidden Nominalism”33 and shortly after in “Ineffability, Ontology and Method.”34 In Individuals, Strawson similarly argues that Bradley’s regress is to be countered by taking exemplification not as a relation, but as a “non-relational tie.”35 These two great philosophers appear to have arrived independently to what is essentially the same idea. It is interesting to see how Bergmann presents the matter. Consider first this passage, where it is proposed that in a fact consisting of a spot’s being green we have as constituents a particular (“individual,” in Bergmann’s terminology), a property (a “character,” in Bergmann’s terminology), and the nexus of exemplification: Consider a green spot. Call F the fact of its being green. It is as “simple” a fact that I can think of. Yet, being a fact, F, like any other fact, has constituents which are things. In this sense no fact is “simple.” Which thing or things are the constituents of F? Some answer that they are two, namely, an individual (the spot) and a character (the spot’s color). The answer seems obvious; to me it is obvious. But, then, it has not been and still is not obvious to many. Some...[might] insist that F ha only one constituent (which is a thing) and that this constituent is neither the spot nor its color, but rather, the “colored spot.” We have come upon the root of the realism-nominalism controversy, a disagreement so fundamental that one might expect it to be relevant to almost any question a philosopher is likely to raise. I shall take it for granted that F has two constituents which are things and which are presented to us whenever F is presented, namely the individual and the character. But I shall also argue that F has three further constituents which in some sense are presented to us whenever F is...The three additional constituents of F are (1) individu36 ality and (2) universality, and the “nexus” of (3) exemplification.

We can set aside for present purposes Bergmann’s insistence on individuality and universality as further constituents of the fact. Let us focus instead on how Bergmann relates the thesis that exemplification is a nexus to the regress. He goes on to argue that exemplification is not another “thing” 33

Bergmann (1958), p. 210. Bergmann (1960). 35 Strawson (1960), Ch. 5, § 2. 36 Bergmann (1960), pp. 20-21. 34

155 in the fact F, along with the spot, a, and its color, G, and thus that it had better be represented not by a further label, “E,” beside “a” and “G,” but by the mere juxtaposition of G and a. Thus, “Ga,” and not “EGa” is to represent the fact, Ga, which is the spot’s being green.37 Of course we can give a label, “E,” to exemplification, as we have just done. But labeling exemplification in the representation of a state of affairs, by writing “EGa” instead of “Ga,” Bergmann notes, “either leads to disaster or, at least, is futile.”38 It would be futile, if this were merely taken to be the proposal to designate faithfully (in the terminology I proposed in “Bradley’s Regress Revisited,” above), but less concisely, what “Ga” already quite efficiently designates. It would be disaster, if it were taken to signal that exemplification is a relation that links G and a (and thus, in my terminology of “Bradley’s Regress Revisited” above, that E is the really attributive constituent of Ga, and that G is simply a prima facie attributive constituent of it). The disaster of course is the internalist version of Bradley’s regress, or, as Bergmann puts it in this paper, “a famous argument proposed by Bradley.”39 Let us recall that, according to Bergmann, there are other nexuses besides exemplification and that he also calls them “subsistent” or “syncategorematic” entities as opposed to “things” or “categorematic” entities, i.e., particulars and universals. As far as we are concerned here, we can neglect nexuses other that exemplification, but I mention this point, since it shows up in the passage from Bergmann 1963 quoted in the introduction. I have submitted it to the reader’s attention, not only because it most explicitly relates to Bradley’s regress the thesis that exemplification is syncategorematic (i.e., a nexus), but also because it clarifies what this is supposed to mean, as well as an important point in the above quotation from Bergmann 1960, i.e., that in some sense exemplification is presented to us when we observe a fact. I take it that, in the 1963 passage, Bergmann is saying something like this. As we see a spot which is round and red, but not green and square, in spite of the fact that green and square are also somehow in the visual field (in another spot), we must account for an observed pre-theoretical “tie” between round and red (which we do not observe, e.g., between round and 37

Actually, Bergmann makes use of parentheses as visual aids and thus he would write “G(a);” I have avoided this, because I have stipulated above that parentheses be used to mark the distinction between states of affairs and propositions. 38 Bergmann (1960), p. 23. 39 Bergmann (1960), p. 24.

156 green). I would say we could make the same point with relations. We could observe a red spot on a green spot, which in turn is beside a white spot. Thus we observe a “being on” tied to red and green, but not to green and white, and a “being beside” tied to green and white but not to red and green. We have in other words a pre-theoretical datum that we can put as follows: (TOD) Tie Observation Datum. We can observe properties and/or relations that are mutually tied (such as the properties round and red when we see a red round spot) and properties that are not mutually tied (such as the properties round and green when we see a red round spot and a green square spot). Now, in an ontology based on tropes we could account for this datum, by postulating a “compresence,”40 which, in Bergmann’s example of the two spots, ties a red trope to a round trope. In contrast, in an ontology that appeals to states of affairs as constructed out of particulars and universals, we must account for it by admitting in a state of affairs an entity that mutually ties properties (round and red in the example) by tying them to a certain particular (a spot, in the example41): the nexus of exemplification. But on pain of Bradley’s (internalist) regress we must also grant that this entity is all that is needed for the observed tie to be there. The last point is crucial. Here is how Bergmann puts it in Realism: “Qualities, then, need nexus to connect or tie them into ordinary things. A nexus does not need a further entity to tie it to what it ties, otherwise we would have entered upon an infinite regress.”42 In contrast a relation needs this further entity and in this sense a nexus is not a relation (like being on or being beside). This might sound like a contradictio ex vi terminorum, given that in ordinary speech “to tie” and “nexus” might be taken as synonyms of “to relate” and “relation,” respectively. To answer this charge by saying that “nexus” and “to tie” are used as technical terms with non-ordinary meanings is not sufficient, for what these meanings are must be clarified. I think this clarification should go as follows. A relation is a thing that can be observed (at least in paradigmatic cases) as either tied or not to other things, as recorded 40

see, e.g., Armstrong (1989), p. 114. More precisely, according to Bergmann (1964), p. 126, the particular is a “bare particular” somehow present in the spot, as a further constituent together with the properties of the spot (round, red, etc.), and it is this bare particular that exemplifies the properties. But we can leave these subtleties aside for present purposes. 42 Bergmann (1967), p. 9. 41

157 in (TOD). A nexus is not something observed as tied, but it is something that we postulate to account for (TOD), as “ontological ground” for it. I guess Bergmann could concede that, for all we know independently of Bradley’s regress, we could think that this postulated entity is another relation, but (given Bergmann’s assumption that the nexus is internal to facts) Bradley’s regress intimates that a nexus is not a relation and this essentially means that, once we have postulated it, we cannot go on postulating further tying entities to account for the ties that we can observe. Nor should we, since there are no other observed ties that these hypothetical further entities should account for. Still one might insist that this approach does not work, because it leaves us in need of an ontological explanation of how the nexus of exemplification, F and a are unified in a state of affairs Fa. This problem has been raised by Vallicella, who puts it as follows: ... let it be granted that exemplification is a nonrelational tie or nexus, call it NEX, and that this nexus does not spawn a regress. There will remain the problem of accounting for the difference between a’s being F and the sum a + F-ness + NEX. Thus the problem is not primarily one of blocking a regress, but one of ensuring the unity of a fact’s constituents. If you try to do this with exemplification relations, you get for your trouble a vicious infinite regress. But if you try to do it with a non-relational tie, you avoid the regress, but are left with the unity problem.43

In other words, given that there is the fact Fa, what explains, ontologically, that we are in world wherein the nexus of exemplification ties F to a, rather than in a world wherein the nexus does not tie F to a? One could be inclined by this question to dismiss Bergmann’s position just as Vallicella does. But I think there is another option. One can reply that the unifying presence of the exemplification nexus in a fact Fa, which grants the very existence of it, must be taken as an ontological brute fact. If this is so, however, one might suspect that Bergmann’s nexus internalism is just a uselessly baroque version of the brute fact approach (BFA). For at the end of the day the latter says that it is a brute ontological fact that the exemplification nexus is a constituent of Fa, whereas the former directly proposes that it is a brute ontological fact that Fa exists. One might charge that this postulation of a nexus in Fa is too ad hoc to be illuminating and therefore that one should just accept (BFA), which, more economically, avoids the postulation of the nexus. 43

Vallicella (2000), p. 241.

158 But Bergmann could reply that the nexus is not idly postulated, because we need to account for (TOD). As we noted, we could account for it in a trope ontology by postulating a compresence. But similarly we must account for it in a thing ontology à la Bergmann. The brute fact approach neglects this and therefore it cannot be right. Nexus internalism complies with (TOD), by postulating the nexus, without taking it to be a relation, in order to avoid the internalist regress.44 Fact Infinitism Let us now consider the externalist thesis (C2). It can be viewed as an approach suggested by the idea that the viciousness of the internalist regress is directed against the assumption that exemplification is internal to a fact Fa rather than against exemplification’s being a relation (as is the case for Bergmann). On the other hand, once (C2) is accepted, an elementary induction immediately proves the externalist version of Bradley’s regress. This can be stated with some precision as follows: (EBR) Externalist Bradley Regress. If (C2) is true, then, for every number n > 1 and fact Fa, there are n facts E1Fa, E2E1Fa,...., En ... E2E1Fa mutually distinct and distinct from Fa, as well as an explanatory chain (of cardinality n+1), Fa exists ⇐ E1Fa exists ⇐ E2E1Fa exists ⇐ .... ⇐ En ... E2E1Fa exists. The thesis (C2) could be aptly called fact infinitism, since, via (EBR), it invites us to admit that, for each state of affairs Fa, there are infinite distinct facts E1Fa, E2E1Fa, etc. Perhaps fact infinitism or at least something in the same spirit may be attributed to Meinong on the basis of the quotation in the introduction.45 44

I had all too quickly dismissed nexus internalism in Orilia (2006a). Erwin Tegtmeier helped me to see its merits, by insisting in conversation on the importance for Bergmann of grounding ontology on phenomenological evidence, which led me to give to (TOD) the appropriate weight in evaluating this approach. 45 Of course, fact infinitism tout court cannot straightforwardly be attributed to Meinong, for it is not clear, as already mentioned above, that a category of states of affairs, as here understood, is acknowledged by Meinong. Something like fact infinitism appears to be entertained by Broad (1933), p. 86, and it may be suspected that he attributes it to McTaggart. But Broad dismisses it immediately, as he takes for granted the brute fact approach: “The fact that at every stage after the first the relating relations are purely formal and are merely repeated shows that we are now embarked on the selfevidently impossible task of explaining [italics mine], by means of particular relational

159 It should be noted that in this approach we account for (TOD), by postulating an additional state of affairs “behind” an “observed” fact. Thus, suppose we observe that green and square are tied, because we see a green square spot and we analyze the observation by appealing to universals and thus to two “observed” state of affairs, say Ga and Sa. The tie between G and S is accounted for by postulating the additional states EGa and ESa “behind” the “observed” facts Ga and Sa. These additional states of affairs ground the tie between G and a, and S and a, respectively, and thus in turn between G and S. One might suspect however that the externalist Bradley’s regress is vicious and in particular that it strikes against fact infinitism just as the internalist regress strikes against (A1). After all, for any fact Fa, there exists, according to this doctrine, an infinite explanatory chain, an infinite series of propositions linked by ⇐, without a last member, a “ground” supporting the alleged explanation of the existence of Fa: Fa exists ⇐ E1Fa exists ⇐ E2E1Fa exists ⇐ E3E2E1Fa exists ⇐ .... This is at odds with a principle often (implicitly) accepted, which we could call Absolute Foundationalism. It goes like this: (AF) All explanatory chains are finite. That is, for any explanatory chain C, there is a finite set S of propositions such that (i) C ⊇ S, and (ii) for any proposition P ∈ S, if there is an explicans Q of P, then Q ∈ S. Intuitively, (AF) is true if every explanatory chain coincides with, or is part of, a finite explanatory chain that terminates with what we could call a basic proposition, i.e., a true proposition B such that for no proposition Q it is the case that B ⇐ Q. Clearly, in the chain Fa exists ⇐ E1Fa exists ⇐ E2 E1Fa exists, ... there is no such basic proposition. Hence, if there is a fact Fa and (C2) is accepted as an ontological explanation of its existence, (EBR) leads us to deny (AF). Thus, to save fact infinitism, we must sacrifice (AF). Now, admittedly, to reject (AF) might seem strange. For intuitively it seems correct to say that we have an explanation for P, only insofar as there is, so to speak, an increase in our knowledge/understanding, when we contemplate P. But, one could argue, if in an attempt to explain P I begin an explanatory task wherein at every stage I must presuppose a succeeding stage, then there is no increase. For any such increase is an approximation judgements, that general relational form which is presupposed by all relational judgements whatever.”

160 to the final stage, and if there is no such stage, then there is no explanation. And thus there cannot be infinite explanatory chains. But in fact good motivations in favor of such chains can be offered.46 That at any given stage we can continue the explanatory task does not show that no knowledge or no understanding is provided at any stage. It merely shows that at no stage we know/understand everything that there is to know/understand about the explicandum that gives rise to the explanatory chain. And noting that the explicandum in question gives rise to such an infinite chain may be considered part of our understanding of it. And so I think that (AF) can be sacrificed and fact infinitism saved after all.47 Conclusions Bergmann is right in taking the internalist Bradley’s regress to be vicious. Nevertheless, Meinong’s positive attitude toward Bradley’s regress is justified, if we understand this attitude as directed to the exernalist version of it. From this perspective, there is no real disagreement between Bergmann and Meinong (as we have understood the latter here) on whether the regress is vicious or not, for there are two regresses at issue. However, there is disagreement on the problem of the unity of facts, for Bergmann tackles it by “deriving” his nexus internalism from the internalist regress, whereas a “Meinongian” fact infinitism can be “derived” from the externalist regress in order to address the problem in quite a different way. Both approaches seem to me better than the other available theoretical options. But which one among them should we prefer? The former postulates a category of nexuses, whereas the latter avoids this, by viewing exemplification as a relation. On the other hand, the latter postulates an infinity of states of affairs for any given fact, a commitment that the former eschews. It is thus hard to decide from the point of view of Ockam’s razor. Although I personally feel more inclined to endorse fact infinitism, perhaps we should resort to Castañeda’s dia-philosophy48 and contemplate both theories as viable tools in our effort to penetrate the nature of reality.49 46

See Klein (1998, 2003). For a more detailed defence of fact infinitism, see Orilia (2006). 48 Castañeda (1980), p. 104. 49 The material in this paper has been more succinctly presented at the Gustav Bergmann Centenary Reunion, Iowa City, 19-20 May 2006. I wish to thank Laird Addis for his kind invitation to join it and all the participants for the useful discussion. Many parts of this paper can already be found in Orilia (2006). Meinong Studies and its publisher Ontos Verlag have granted permission to reproduce them here. Finally, I would 47

161 REFERENCES Armstrong, David. 1978. Nominalism and Realism. Universals and Scientific Realism, Volume I. Cambridge: Cambridge University Press. Armstrong, David. 1989. Universals: An Opinionated Introduction. London: Westview. Armstrong, David. 1997. A World of States of Affairs. Cambridge: Cambridge University Press. Bergmann, Gustav. 1958. “Frege’s Hidden Nominalism,” The Philosophical Review, 67: 537-573, (reprinted in Bergmann 1959: 205-224; references are to this version). Bergmann, Gustav. 1959. Meaning and Existence. Madison: University of Wisconsin Press. Bergmann, Gustav. 1960. “Ineffability, Ontology and Method,” The Philosophical Review, 69: 18-41, (reprinted in Bergmann 1964b: 45-63). Bergmann, Gustav. 1964. “Ontological Alternatives,” Bergmann 1964b: 124-157. Bergmann, Gustav. 1964b. Logic and Reality. Madison: University of Wisconsin Press. Bergmann, Gustav. 1967. Realism: A Critique of Brentano and Meinong. Madison: University of Wisconsin Press. Bonino, Guido. 2001. Russell’s and Wittgenstein’s Tractatus, doctoral dissertation. Vercelli: Università del Piemonte Orientale. Bonino, Guido. 2006. “Why There Are No Facts in Meinong’s World (according to Gustav Bergmann),” Meinongian Issues in Contemporary Italian Philosophy (V. Raspa, ed.), Meinong Studies, 2: 213-266. Bradley, Francis. 1893. Appearance and Reality. London: Swan Sonnenschein and Co. Broad, C. D. 1933. An Examination’s of McTaggarts’s Philosophy, vol. I. Cambridge: Cambridge University Press. like to say that in writing this paper I have found very useful to consult the special issue of Rivista di estetica, ed. by G. Bonino and G. Torrengo, dedicated to Bergmann (Rivista di estetica, Il realismo ontologico di Gustav Bergmann, 44, 2004).

162 Candlish, Stewart. 2006. “Francis Herbert Bradley,” The Stanford Encyclopedia of Philosophy. Fall 2006 Edition, E. N. Zalta (ed.), URL = . Castañeda, Hector-Neri. 1980. On Philosophical Method. Bloomington: Noûs Publications. Dodd, Julian. 1999. “Farewell to States of Affairs,” Australasian Journal of Philosophy, 77: 146-160. Gaskin, Richard. 1995. “Bradley’s Regress, The Copula And The Unity of The Proposition,” The Philosophical Quarterly, 45: 161-180. Hochberg, Herbert. 1989. “Russell’s Paradox, Russellian Relations and The Problem of Predication and Impredicativity,” Rereading Russell, C. W. Savage and A. C. Anderson, eds. Minneapolis: University of Minnesota Press: 63-87. Klein, Peter. 1998. “Foundationalism and The Infinite Regress of Reasons,” Philosophy and Phenomenological Research, 58: 919-925. Klein, Peter. 2003. “When Infinite Regresses are Not Vicious,” Philosophy and Phenomenological Research, 61: 718-729. Loux, Michael. 1998. Metaphysics, A Contemporary Introduction. London: Routledge. Meinong, Alexius. 1899. Über Gegenstände höherer Ordnung und deren Verhältnis zur inneren Wahrnehmung, “Zeitschrift für Psychologie der Sinnesorgane,” 21: 182-272; reprinted in A. Meinong, Gesamtausgabe, edited by R. Haller and R. Kindinger, with the collaboration of R. M. Chisholm, vol. 7, Akademische Druck-u. Verlagsanstalt, Graz, 1969-78, vol. II: 377-471. English translation, “On Objects of Higher Orders,” in A. Meinong, On Objects of Higher Orders and Husserl’s Phenomenology, edited by M.-L. Schubert Kalsi, The Hague: Martinus Nijhoff. (Page references are to this translation). Orilia, Francesco. 1991. “Type-free Property Theory, Bradley’s Regress, and Meinong and Russell’s Reconstructed,” Grazer Philosophische Studien, 39: 103-125. Orilia, Francesco. 2000. “Argument Deletion, Thematic Roles, and Leibniz’s Logico-grammatical Analysis of Relations,” History and Philosophy of Logic, 21: 147-162.

163 Orilia, Francesco. 2006. “States of affairs. Bradley vs. Meinong,” Meinongian Issues in Contemporary Italian Philososophy, V. Raspa, ed., Meinong Studies, 2: 213238. Orilia, Francesco. 2006a. “Stati di cose, esemplificazione, e regresso di Bradley,” Rivista di Filosofia: 97, n. 3: 349-385. Read, Stephen. 2005. “The Unity of the Fact,” Philosophy, 80: 317-342. Russell, Bertrand. 1903. The Principles of Mathematics. Cambridge: Cambridge University Press. Sellars, Wilfrid. 1962. “Naming and Saying,” Philosophy of Science, 29: 7-26. Simons, Peter. 1997. “Un mondo senza stati di cose,” Discipline Filosofiche, 7, 97/2: 29-47. Tatzel, Armin. 2002. “Bolzano’s Theory of Ground and Consequence,” Notre Dame Journal of Formal Logic, 43: 1-24. Vallicella, William. 2000. “Three Conceptions of States of Affairs,” Noûs, 34: 237259. Vallicella, William. 2002. “Relations, Monism, and the Vindication of Bradley’s Regress,” Dialectica, 56: 3-35. Vallicella, William. 2002a. A Paradigm Theory of Existence, Onto-theology Vindicated. Dordrecht: Kluwer. Van Inwagen, Peter. 1993. Metaphysics. Boulder: Westview Press. Varzi, Achille. 2005. Ontologia. Bari: Laterza. Wetzel, Thomas. 2003. “States of affairs,” The Stanford Encyclopedia of Philosophy, Fall 2003 Edition, E. N. Zalta (ed.), URL =

Bergmann’s Thinkable Inexpressibles WILLIAM HEALD University of Iowa

O

n the first page of his posthumously published New Foundations of Ontology (henceforth NFO) Bergmann states that “Ontology accounts for everything there is in terms of simples.”1 In this book Bergmann articulates a small number of ways in which simple things (particulars and universals) go together to compose all of the complex existents that we can think (intend) and—as he once would have said and continued in NFO to say—express with sentences of the ideal language (I.L.). But he also offers arguments for the somewhat paradoxical claim that there are (there exist) some “objects” of thought that cannot be expressed in the I.L. His arguments for this claim and their implications are what I want to examine here. 1. The ways in which things go together to compose complex existents are expressed in NFO by means of formulae Bergmann calls ‘canons’: (i) If there is a particular, α, and a nonrelational universal of the first type, β, then, since the nexus of nonrelational exemplification exists, there also exists the fact α-exemplifying-β; (ii) if there is a complex, α, and another complex, β, then, since the connective disjunction exists, there also exists the fact α v β; (iii) if there is a determinate, α, and another determinate, β, then there exists the circumstance α-being-diverse-from-β2 The canons are, 1

Bergmann, Gustav, New Foundations of Ontology, Madison: University of Wisconsin Press. 1991, p. 43. 2 The categorical system Bergmann introduces in NFO involves three categories of simples (things, functions, and entities) and three categories of nonsimples (facts, circumstances, and classes). The category of determinates contains things, facts, circumstances and classes. A determinate is thus either a simple determinate or a nonsimple determinate “produced from” things in accordance with the canons.

166 in essence, ontological counterparts (nonlinguistic restatements) of the formation rules of an ideal language (I.L.). The philosophical function of the I.L. is to represent every thinkable complex with a sentential expression that perspicuously reflects the ontological content and structure of that complex. The canons are supposed to specify all of the most general ways that simple things can go together to comprise a thinkable complex existent. The general “features” that we attribute to the ordinary objects of thought and discourse are thus ontologically accounted for, grounded, in terms of constituents “in” them and the ways these constituents go together. The assumption that the features of ordinary objects are to be accounted for by existents “in” them and that I.L. sentential expressions display these constituents and the way in which they are combined was always Bergmann’s guiding methodological principle. But it is also one that he realizes (not without prodding by critics3) has inescapable limitations. These limits become apparent at the level of the simple things of which all complex existents are composed, and again, at the level of the complexes that—in accordance with the canons—they compose. (1) Simple things are not completely simple. Any given particular (for example), has two “aspects”: (a) it is a particular, sharing this “feature” with all other (and only other) particulars, and (b) it is also the very particular that it is, distinct from all other particulars. If we are to account for the features of all existents in terms of their constituents then does not the double aspect of all things require us to suppose that there are subsimples “in” them to ground these two aspects? But then does this not render the socalled simples complex? And wouldn’t the same problem arise again anyway with the subsimples? (2) There is more to complexes than the things in them being connected in the way that they are (canonically) connected. The way that the canons are 3

The critics of whom I speak are Panayot Butchvarov and John Peterson, both of whom Bergmann explicitly acknowledges in NFO as the source of his realization of a need for a type of composition radically different than that exhibited by sort of complexes represented by I.L. sentential expressions. See Peterson’s “Bergmann’s Hidden Essentialism,” Review of Metaphysics, 22, 1969, pp. 660-75. See Butchvarov’s “The Limits of Ontological Analysis,” in The Ontological Turn, Iowa City: University of Iowa Press, 1974.

167 phrased expresses Bergmann’s conviction that nonactual facts (Bergmann calls them potential facts) have ontological status. In short, they exist. We can thus distinguish between the existence and “identity” of a complex, on the one hand, and its “status” (Bergmann refers to this status—or its ontological ground—as a mode) as being either actual or potential, on the other. But if we are to account for the distinction between the two modes a fact may have in terms of further constituents, do we not then make a fact being actual (or potential) itself another fact, which would then be—like all facts—either actual or potential? Does each fact then have not only a mode but also serve as a pedestal for an infinitely ascending hierarchy of metamodes (indeed, an infinite number of distinct infinitely ascending hierarchies of metamodes)? Despite these ominous problems Bergmann does introduce a new category of existents to ground the ultimate sorts of things and the modes of complexes. He calls the members of this new category entities and the manner in which they combine hylomorphic composition (in honor of Aristotle’s composition of form and matter). Each thing (universal or particular) is a canonical simple, but is also a hylomorphic composite of two entities, an item and an ultimate sort, the latter grounding its logical type and the former grounding its distinctness from all other things of the same type. And every fact is a canonical complex, but is also pervaded by a mode. The pervasion of a fact by its mode is the second type of noncanonical (hylomorphic) composition. Things and the non-simple existents “produced” from things in accordance with the canons Bergmann calls determinates. The hylomorphic composites—items and ultimate sorts into things and the pervasion of complexes by their modes—Bergmann calls ‘Two-in-Ones.’ A major distinguishing characteristic of the entities is that there can be no expression in the I.L. that designates them. Equivalently, the Two-in-Ones cannot, as can all of the canonical composites, be represented by sentences of the I.L.4 Bergmann’s reasoning in support of the existence of the Two4

I am overstating the case a bit here. Of the canonical composites only those that are complexes are represented by sentences of the I.L. Complexes are either facts or circumstances. The paradigmatic type of circumstance is the diad, or diversity. The diad a-being-diverse-from-a-exemplifying-F, for example, is classified as a circumstance rather than a fact because being-diverse-from is neither a relation nor a function. Diads, like all circumstances, are thus internal relationships, whose ground is the re-

168 in-Ones and their “ineffability,” although elusively expressed, is fairly clear.5 [A] (1) In some rare instances a single thing can be the entire intention of an act of awareness. In such cases, and only in such cases, one is presented with the ultimate sort of that thing.6 (2) One cannot think—is never in any case presented with—a thing being of a different ultimate sort from the one that it is of. Nor can one think the negation of a thing being of its ultimate sort. (Bergmann refers to these phenomenological claims as, respectively, the “unthinkability of the different ultimate sort” and the “unthinkability of the negated ultimate sort”.) (3) Suppose that there were expressions in the I.L. designating items and ultimate sorts. If so then there would be I.L. sentences expressing the combining of items and ultimate sorts into things. (4) Every I.L. sentence can be negated (as well as conjoined with others, disjoined with others, and so on); the result of doing so is always another well-formed I.L. sentence. (5) Every I.L. sentence expresses something that can be thought (intended). (6) There can, therefore, be no signs in the I.L. representing items and ultimate sorts and thus no sentential expressions representing their combining into the things they compose. [B] (1) In some cases one may be presented with the mode of a complex of which one is aware.7 lata (which Bergmann calls terms) of the relationship. Only complexes have mode. Nothing that does not have a mode is represented by a sentence. Classes, though not simple, do not have modes and are therefore not represented by sentences. 5 The arguments, as formulated here, are somewhat difficult to see in the text of NFO itself, although I am convinced that they are there. See especially the last chapter of NFO, pp. 326-335. My interpretation of these arguments is largely shaped by a letter (unpublished) written by Bergmann in 1970 to Reinhardt Grossmann. 6 Bergmann claims that a single thing can be the entire (Bergmann’s term is “explicit”) intention of an act of awareness only if the act is of a special species he calls grasping. One can be “implicitly” aware of things by being aware—by means of acts of perceiving, remembering, and other species—of complexes of which the things are constituents. 7 When the mode of a complex does become apparent to someone who intends that complex, Bergmann describes the complex as having become transparent. A complex

169 (2) One cannot think—is never in any case presented with—a complex being pervaded by a mode being in turn pervaded by a mode. [Bergmann calls this phenomenological claim the “unthinkability of iterated modes.”] (3) Suppose that there were expressions in the I.L. designating the modes, actuality and potentiality. Then there would be I.L. sentences representing the pervasion of complexes by these modes. (4) If there were I.L. sentences representing the pervasion of a complex by a mode then these sentences would themselves represent complexes. Thus there would be sentences describing the pervasion of a complex by a mode being pervaded by a mode. (5) But every I.L. sentence expresses something that can be thought, so there can be no I.L. sentences representing the pervasion of a complex by its mode. The two arguments are, of course, exact analogues of each other. Each rests on two phenomenological claims, one of them to the effect that something is thinkable (and thus exists) and the other to the effect that something is not thinkable (and thus does not exist). To these phenomenological claims are then applied the “principles” that (i) every I.L. sentence expresses something one can think, something with which one can be presented, and (ii) every well-formed I.L. sentence can be intelligibly negated. The larger conclusion is that there are certain features of the world that one can literally think that are not expressible in the I.L., a linguistic schema in which, presumably, everything thinkable can be expressed. That is, there are features of the world that “straddle the line” between what can be thought and said on the one hand, and what cannot be said because it cannot be thought on the other, features that are, in other words, thinkable but not expressible.

may become transparent either immediately or recursively. Complexes that may become transparent either immediately or recursively Bergmann calls analytic. This term covers all those complexes that he earlier had classified as analytic and many more besides these. But none of those complexes that most philosophers would classify as contingent, even those of which one is directly aware, are now classified by Bergmann as analytic. Thus one is never presented with the mode of complexes—those involving one’s own present conscious states—of which one is directly aware. The phenomenalism Bergmann once espoused is, in NFO, finally dead.

170 2. In a penetrating examination of the problem posed for Bergmann by the ultimate sorts and the modes, Panayot Butchvarov sees Bergmann’s Two-inOnes as less of a solution of the problems they are invoked to solve than as mere tinkering that is evidence of a deeper and more general defect of Bergmann’s general ontological method, the assumption that the proper way of accounting for the features of things is in terms of what is in them, in terms of their constituents. This assumption construes ordinary things as complexes, as wholes consisting of parts. The allure of the whole-part analogy, Butchvarov argues, rests on a deeper analogy between the wholepart relation itself and the relation between an arbitrary collection and its members: Insofar as we think of an object as a whole consisting of parts, we recognize that it would not be what it is and would not exist at all if its parts were not what they are or did not exist at all. And insofar as we think of an arbitrary collection of entities as itself an entity, we recognize that it would not be what it is and would not exist if its members were different 8 or ceased to exist at all.

Construing ordinary objects as wholes consisting of parts is attractive, Butchvarov argues, because we are covertly, though mistakenly, supposing that a whole consisting of its parts is not significantly different from an arbitrary collection consisting of its members. And since it is indeed plausible to suppose that once we know what the members of an arbitrary collection are, we know everything there is to know about the collection, if it were the case that a whole is to its parts as the members of a collection are to the collection—and were also the case that an ordinary object is a whole consisting of its constituents—then it would follow that once we know what the constituents of an ordinary object are, we would know everything—or everything ontologically relevant—about that object. But the appealing luminosity of our secondary analogy breaks down in the cases that Butchvarov indicates: (1) accounting for the unity of complexes in terms of constituents of those complexes, and (2) accounting for 8

The Ontological Turn, p. 27. It is crucial to keep in mind here that Butchvarov is using the term ‘entity’ here as Bergmann himself did prior to NFO, as a general term for anything that exists or has ontological status.

171 the nature of the simples of which all complexes are supposedly ultimately constituted in terms of something “in” them.9 We cannot account for the unity of a complex in terms of a further constituent of the complex—as Bergmann originally tries to do with fundamental ties and then later, apparently, tries to do with the modes—any more than we can unify an arbitrary collection by adding another member to the collection. And we cannot account for the ultimate sorts of simple things in terms of something “in” them—as Bergmann seems to be trying to do with his notion of the hylomorphic composition of items and ultimate sorts—without running afoul of the assumption that these things are simple, and thus have no parts. The first point to make is that it would surely be implausible to accuse Bergmann of confusing the whole-part relation with that between an arbitrary collection and its members. Bergmann assays the relation between spatial wholes and their parts as an external relation, which would make it impossible to construe a spatial whole as reducible to its parts. And Bergmann was certainly sensitive to the difference between a complex and a collection, once asserting as a fundamental principle of general ontology that “a collection of existents is not itself an existent.”10 Nevertheless, 9

It should be emphasized that in NFO Bergmann no longer regards exemplification, the connectives or the quantifiers as constituents of the complexes they “tie together.” In his earlier work he classified these existents as subsistents, and did regard them as constituents of complexes. In NFO they are classified as functions and—although not constituents of complexes—they are “in” complexes in the lesser sense that they are among the existents that are members of the ultimate foundations of the complexes they “tie together.” The entities are “in” other existents in a still different and even lesser sense. The item and ultimate sort composing a thing are not “in” the foundation of a thing, since a thing is not a canonical composite and so has no foundation from which it is produced. A complex is pervaded by its mode: its mode is not a member of a complex’s ultimate foundation, the collection of things and functions from which the complex is produced by a sequence of “canonical steps.” Nevertheless, in this least strong sense of “in,” entities are in other existents and their being “in” them is needed to ontologically ground certain features of the existents they are “in.” The modes are needed to account for whether the complexes they pervade are actual or merely potential. The ultimate sorts are needed to ground the logical type of things, while the items are needed to ground the difference between things of the same ultimate sort. 10 See p. 9 in Realism, A Critique of Brentano and Meinong, University of Wisconsin Press, Madison, 1967. In fact, Bergmann phrased it as follows: “A collection of entities is not itself an entity.” I have modified it for the obvious reason that Bergmann no

172 Bergmann was fully aware of Butchvarov’s critique and (in NFO) explicitly accepts the accuracy of Butchvarov’s description of the Two-in-Ones as marking “the limits of ontological analysis.” This acceptance is expressed by Bergmann’s admission that the composition of items and ultimate sorts into things and the pervasion of facts by their modes cannot be regarded as canonical composites of particulars and universals and thus cannot be represented by sentences of the I.L. Beyond this, moreover, Bergmann acknowledges that his earlier notion of constituency is too unrestrictive. He replaces this liberal conception of constituency with a more conservative conception that allows only determinates—things, facts, circumstances, and classes—either to be or to have constituents (although neither things nor classes can have constituents). Thus existents that he earlier would have classified as constituents of thinkable complexes— including the functions, the ultimate sorts and the modes—no longer are so describable. Nevertheless, despite how nuanced matters have become in NFO, it is still fairly obvious that Bergmann sees what he is doing as accounting for the general features of objects by finding something “in” them to “ground” these features (recall the slogan about “accounting for everything there is in terms of simples”). Even though functions are no longer regarded as being “in” the facts that are the values of these functions in the (new and more restrictive) sense of being constituents of them, they are still “in” them in the somewhat looser sense of being members of the ultimate foundations from which facts (and other nonsimple determinates) are “produced.” And these foundations are indeed construed as collections in much the same sense that Butchvarov had in mind. Furthermore, even though the entities—the items, ultimate sorts, and modes—are not even regarded as members of the ultimate foundation from which a canonical nonsimple is “produced,” the very phrase “Two-in-One” connotes a species of whole-part connection (Two-“in”-One). Thus Butchvarov might be excused for concluding that, despite all of the distinctions introduced by Bergmann, he is unable to escape the gravitational pull of the whole-part analogy and, indeed, is locked by it into a methodology that tacitly, if not longer uses the term “entity” as a generic term for any existent, but rather for that distinctive subcategory of existents involved in the Two-in-Ones.

173 explicitly, regards ordinary objects as reducible to what is “in” them in much the same way that collections are reducible to their members. While Butchvarov’s suspicions are well founded, I would like to defend Bergmann’s reluctance to abandon the whole-part analogy by emphasizing what I take to be his underlying motive for embracing it. The source of Bergman’s allegiance to the analytic ideal is not so much the prospect of reducing complexes to their constituents without remainder, but rather his abhorrence of any form of idealism no matter how timid or tentative. Unless the ontologically relevant features of objects are accounted for in terms of something that is—in some reasonable sense—“in” these objects, then the only plausible alternative is to suppose that the ground of these features is “in” us, specifically, in our conceptual and/or linguistic activities. Bergmann accounts for the phenomenon of qualitative sameness in terms of universals that are constituents of the complexes that are qualitatively identical. It does not rest merely on the existence of the conventions governing the use of linguistic expressions that we use to describe things or concepts that we use to group things into classes. Bergmann accounts for the phenomenon of the numerical diversity of qualitatively identical complexes in terms of particulars that are constituents of the complexes. These particulars are not something “carved out” of enduring substances or the sensory manifold by our glances or perceivings. And Bergmann accounts for the togetherness of a particular and universal in a fact in terms of the exemplification nexus, a function of which the fact is a value, and which, though no longer (in NFO) considered a constituent of the fact is still “in” it in the sense of being a member of its ultimate foundation. None of these ontologically crucial aspects of objects is regarded by Bergmann as a “creation” or “reflection” of our intellectual or linguistic activity.11 Rather, each is something that belongs to the object, and would still belong to the 11

It will not have escaped notice that Bergmann speaks of nonsimple determinates being “produced from” simple determinates (in accordance with the canons). This provocatively anthropocentric metaphor is chosen precisely (Bergmann claims) so that it can be abandoned and replaced by a more technical notion that does not suggest that the canonical nonsimples are produced by someone. In any case, my use of Bergmann’s metaphor should not be taken as a covert endorsement of the view that he explicitly rejects, that the complex objects of thought and experience are created by the intellectual or linguistic activity of those who are aware of them.

174 object even if we were not to experience it, think of it, or talk about it. This conviction is the essence of his realism. It is Bergmann’s realistic instincts that underlie his notion of the Two-in-Ones. In conceding that the hylomorphic composites are not expressible by sentential expressions of the I.L., Bergmann takes himself to be conceding that the closest we can come to representing them (nonmetaphorically) is metalinguistic statements. But he does not move from this concession to the conclusion that such statements—at least if they are understood as statements about the ideal language—do not correspond in some fashion to extralinguistic reality. “Being-to-the-left-of is a binary relational universal” cannot be transcribed into the I.L. But when we understand what this pseudo-statement purports to say, we have before our minds something more than what is expressed by the metalinguistic statement “‘to the left of’ is a binary relational predicate.” This something more is a hylomorphic composite of an item and an ultimate sort. If there were no such extralinguistic ground “corresponding” to the pseudo-statements attributing ultimate sorts to things, then the most reasonable conclusion to draw would be that the ultimate sort of a thing—and thus its “capacity,” as specified by the canons, to combine with other things into complexes— depends upon the linguistic conventions governing the formation of sentences in the language we use to represent these complexes. Bergmann also admits that a statement such as “every complex that is a disjunction of a fact and its negation is actual” cannot be represented with an I.L. sentence. But when we understand what this pseudo-statement purports to say about facts having this form, we have before our minds more than what is expressed by a metalinguistic statement to the effect that sentences of a certain syntactical form have all ‘Ts’ under the main connective in their truth tables. This something more involves the pervasion of a fact by its mode, the second type of hylomorphic composite. If there were no such extralinguistic ground “corresponding” to the pseudostatements attributing modes to complexes of certain forms then the most reasonable conclusion to draw would be that the principles of logic expressed by, for example, the truth tables, depend upon arbitrary conventions.

175 Addendum: Potential Complexes In the preceding pages, I presented what I take to have been Bergmann’s arguments for regarding a-thing’s-being-of-its-ultimate-sort and acomplex’s-being-pervaded-by-its-mode as occupying the peculiar status of being thinkable (intendable) but not expressible (in the I.L.). But something more should be said about Bergmann’s insistence that all complexes, whether actual or potential, have ontological status. The arguments presented above might suggest that this is a phenomenological claim, that experiential acquaintance with the modes of complexes compels acceptance of the existence of non-actual complexes. While it is true that such acquaintance does play an important role in Bergmann’s expanded conception of analyticity (in NFO analytic complexes are those whose modes are “accessible”), Bergmann introduced the modes—and thus the complexes, actual and potential, that they pervade—before he was willing to concede that we encounter them in thought and experience. What compelled him to do so? What reasons might one give for embracing the existence of potential facts, especially if one is, as Bergmann surely was, rather hostile to the notion of possible worlds?12 In what follows I will sketch what I take to be the dialectical pressures leading to this rather startling metaphysical outcome. (A) Whenever I intend anything, under any species of awareness (perceiving, believing, doubting, remembering, etc.) I am presented not only with what I am presented with but also with its existence. Otherwise, Bergmann claims, “I, for one, would not know what it means to exist.”13 Bergmann is claiming here, of course, not that I am always presented with the mode of any fact that I intend, still less that I am presented with the ac12

Concerning this hostility see, for example, “The Philosophical Significance of Modal Logic,” in Mind 69, (1960), pp. 466-85. 13 In NFO Bergmann enunciates what he refers to as the “Principle of Presentation”: “Whatever exists can be presented (i.e., intended), and whatever can be presented exists.” The first part of the principle reaffirms the anti-speculative character of his ontology. Taken together with the first, the second part of the principle might be taken to provide a sort of definition of the ontologically relevant sense of ‘existence’: something exists if and only if it can be explicitly or implicitly intended. That is, something exists if it can be the entire intention of an act of awareness—of whatever species—or is “in” something that can be the entire intention of an act of awareness.

176 tuality of any fact that I intend. Bergmann’s point, presumably, is that there is no significant difference between the various species of awareness as regards the existence of that of which one is aware. Every species of awareness is an awareness of something. Yet since some awarenesses are, as one says, nonveridical, some of the intentions of some of one’s awarenesses must be nonactual (indeed, a veridical act of the species disbelieving must have a nonactual intention!). Perhaps the idea is that from the point of view of a conscious being there is no apparent feature of the intentions of different species of mental act distinguishing actual from nonactual intentions. All alike present their intention to the conscious being as simply being there, before the mind, so to speak. Other than the most firmly committed Meinongians few will find this entirely persuasive. (Bergmann himself would once not have found it persuasive. Of this more below.) Certainly the direct apprehensions one has of one’s present sensations and mental acts, and also, perhaps, perceptual awarenesses of ordinary material objects in one’s immediate environment, do present their intentions as really being there. But imaginings, doubtings, entertainings and disbelievings, while of course having a specific content, do not seem to present their intentions as existing. But even though perceptual awarenesses do present their intentions as existing (as “real”) and doubtings and imaginings do not, it can certainly turn out that what I perceptually intended was not really there while, on the other hand, what I doubted or imagined was really there after all.14 Perhaps Bergmann is simply emphasizing the Cartesian point that there is no mark in the intention of an act of awareness that indicates whether that act is veridical or not.15 And if one’s ontology is to be erected on a purely phenomenological basis then, since there is no phenomenological difference between the intentions of veridical and nonveridical acts, there is no basis 14

Concerning Bergmann’s analysis of the concept expressed by the word ‘real’ see “Realistic Postscript,” in Logic and Reality, Madison, University of Wisconsin Press. 15 In NFO Bergmann rejects the claim that there is nothing in the intention of any act that is a mark of the veridicality of the act. In some acts of belief—those whose intentions are analytic—the mode of the intention is accessible to the believer. Note that I am using ‘belief’ in a broad sense here to include acts of the species disbelief. What is noteworthy, however, is that Bergmann rejects the view, which he seemed to have tacitly embraced earlier, that one is presented with the actuality of one’s intentions when one is directly acquainted with one’s own present mental states and sensations.

177 for regarding the intentions of one species of act as any more or less of an existent that the intentions of any other species of act. (B) Even if one does insist on granting a kind of priority to direct (i.e., introspective) and perceptual awarenesses—claiming that only they present their intentions as existing—one must still concede that some perceptual awarenesses are nonveridical. Having begun as a phenomenalist, Bergmann had once been well-equipped to deal with this problem as many phenomenalists do, by claiming either that (1) what is before the mind during a nonveridical perception is a sense-datum, or (2) what is before the mind whenever one perceives is a sense datum.16 But Bergmann eventually rejects the phenomenalist assay of ordinary objects that makes this way of responding to perceptual error available (indeed, eventually—in NFO—he abandons completely the notion of a sense datum). Still retaining the structurally “Berkleyan” assay of an ordinary perceptual object as a spatiotemporal mosaic of momentary particulars exemplifying simple perceptual (formerly, simple phenomenal) properties and relations, Bergmann develops an account of the distinction between veridical and nonveridical perception in terms of the modes of the complexes that are the intentions of acts of perceiving. Nonveridical acts of perception intend complexes that are pervaded by the mode of potentiality and veridical acts of perception intend complexes that are pervaded by the mode of actuality. Not only do acts of perception present their intentions as existing, their intentions do— always—exist. But the mode of the intended complexes may differ. The modes thus serve as a way of securing a kind of direct realist, and rejecting a representationalist, conception of perception. (C) Bergmann’s pre-NFO assay of the connection between the propositional characters in mental acts and the intentions of the acts that ex16

See “Realistic Postscript” for the details of this change of position, pp. 302-340. Bergmann’s essential motive for making this move, I believe, was his discomfort with an assay of the ordinary perceptual objects that construed them as complexes of existents that have essentially the same status as toothaches: nonmental (i.e., nonintentional) things that are nevertheless mind-dependent (that is, which exist only when intended). The way that phenomenalists handle the discomfort caused by this, of course, is to construe statements about ordinary objects as actual and possible sense data. Bergmann rather explicitly rejects accepting subjunctive conditionals into the I.L., however, and so he could not have accepted transcription of statements about ordinary perceptual objects as partially consisting of subjunctive conditionals.

178 emplify these characters is centered around an existent he calls the intentional nexus, represented in his I.L. by the upper case ‘M’ (for “means”). Any given mental act consists of a particular (momentary, as all of Bergmann’s particulars are) exemplifying two nonrelational universals, one of them—referred to as the species—grounding the type of awareness it is (perceiving, remembering, doubting, etc.) and the other—called the propositional character (or sometimes the thought)—accounting for which complex the act intends. Thus a perceiving that b is blue is the act perc(a)& bl(b) (a), where ‘perc’ designates the species exemplified by a and ‘bl(b)’ designates the propositional character intending the fact bbeing-blue. The fact that the propositional character intends the fact it does is the fact: bl(b) M bl(b). Bergmann construes this fact as analytic, having essentially the same status as standard truth table tautologies, and he construes ‘M’ as designating an existent having the same ontological status as the connectives “in” molecular facts. Doing so is attractive since connectives can, presumably, connect existing facts (and things) to nonexisting facts.17 In NFO Bergmann expands his conception of analyticity to include a newly introduced category of complexes he calls circumstances. The primary subcategory of circumstances are diads, or diversities. The diad (a,b), for example, is a-being-diverse-from-b. The diad (f(a),g(b) v f(a)) is f(a)being-diverse-from-g(b)-or-f(a). (The existents whose diversity diads ground are called the terms of the diad.) The significant features of diads are these: (1) any two determinates, of whatever category or type (things, facts, circumstances or classes) are the terms of a diad, (2) the diversity of the two terms of a diad is not grounded by a distinct relation or connection of diversity. Diversity is, in other words, an internal relationship (not: an internal relation) holding between any two determinates. A major symptom of this latter feature of diads is that the I.L. sign by means of which diads are represented is not—unlike the signs designating functions (exemplification, the connectives and quantifiers)—a sign that designates something. The symbol ‘( , )’ does not designate an existent (the relation or connection 17

Bergmann clearly did not always believe that the intentions of nonveridical acts exist. That fact that he did not was a motive for classifying the intentional nexus as a connective-like subsistent tie.

179 of being-diverse-from) that ties together diverse determinates.18 Rather it is what Bergmann calls a diacritical sign. A diacritical sign is one that appears in an I.L. expression designating a nonsimple determinate that does not designate anything “in” (in whatever sense of “in”) the nonsimple determinate represented but which “draws attention” to the sort of nonsimple determinate that it is. There is nothing “in” a diad except its terms (and, of course, whatever—if anything—is “in” these terms). The mode of the diad is determined by nothing except “the being and nature” of the terms themselves. In NFO Bergmann decides that the intentional nexus between a thought and its intention is a circumstance rather than—as he had earlier claimed—an odd sort of molecular fact, and from this he draws the appropriate conclusions: (i) complexes such as bl(b) M bl(b) are, though still analytic, analytic in the new broad sense in which complexes such as (f(a),g(b) v f(a)) are analytic, and (ii) the symbol ‘M’ is, like the sign ‘( , )’, a diacritical sign. But if ‘M’ is a diacritical sign, then the nexus between a thought and its intention is an internal relationship whose mode is determined entirely by “the being and nature” of its relata.19 And if this is so then the relata had better exist. Even if one had no difficulty with a connection between what does exist (a thought) and what does not exist (its intention), once the existent that was supposed to ground this connection is no longer granted ontological status we have nothing left except the relata themselves. If they do not exist, then the connection between them either has no ontological ground at all or one that is contained in only one of the relata. (D) Bergmann conceives of the world as a collection of nonsimple existents each of which emerges ultimately, from a “bottom layer” of simple things by a sequence of canonical “steps.”20 Atomic facts are “produced from” particulars and universals (of the appropriate sorts) by the function 18

The diacritical sign ‘( , )’ is a component of larger sentential symbols, like ‘(a, c)’, which can be read that “a is different from c” or “a’s being diverse from c.” 19 The phrase “being and nature” is the phrase Bergmann himself uses in NFO. This in itself represents a rather startling turnaround, since he was always hostile to the notion of nature so employed. He does not really explain his use of the phrase. But he on occasion uses the notion of the identity of an existent in a roughly equivalent fashion. See his discussion of this latter notion in NFO on pp. 266-267. 20 See, in NFO, especially in the first chapter, “Simples and Canons”.

180 of exemplification. Elementary diads (diads whose terms are things) are produced from similar materials without any functions. Molecular complexes are then produced from atomic facts (or, for that matter, elementary diads) by the connectives and quantifiers (which are also functions). Further diads may have molecular facts as terms, and these diads may serve as the material used to create further facts. And so on. This way of conceiving of nonsimple existents makes the existence (and the identity) of nonsimple existents dependent upon the simpler “materials” from which they emerge. Unless the particulars and universals that are the constituents of atomic facts exist (and, indeed, the functions of which atomic facts are the values also exist), then these atomic facts would not exist. And unless the facts f(a) and g(b) exist, then their conjunction f(a) & g(b) would not exist. But consider now the negative fact -h(c). Suppose that an act intending this fact is veridical. According to Bergmann this fact is then actual. Others— though not all others—would concede that it exists. But what then of its immediate constituent, h(c)? If it does not exist then what is it that is negated to produce -h(c)? And if it is claimed that where -h(c) exists h(c) does not exist, then what is the difference, ontologically speaking, between -h(c) and any other negated fact? Suppose, for example, that both ‘-f(a)’ and ‘-h(c)’ are true. What are the constituents of the facts represented by these two statements? Do they have any constituents? Do they have different constituents? Or is what makes them different what some conscious being who asserts one of them is thinking? In other words, if I assert ‘-f(a)’ and do so on the basis of seeing that g(a)—while believing that being-g is incompatible with being-f—is the truth maker for my assertion different from what would be the truth maker for the “same” assertion made by someone else who did on the basis of seeing that, say, h(a), while believing that being-h is incompatible with being-f? Are two such people even making the same assertion? Or are they making one assertion that has two (or potentially an indefinite number of) distinct truth makers?21 Many will no doubt find the acceptance of negative facts every bit as problematic the acceptance of potential facts, if they even see any real difference between them. Such critics would probably claim that the reason21

One can ask essentially the same question in terms not of the identity of assertions but rather of the identity of acts. Are the thoughts in all veridical acts of awareness whose intentions are negative facts identical?

181 ing sketched in the previous paragraph is a reductio of Bergmann’s conception of facts as complexes consisting of, and distinguished by, their constituents. But unless one is very comfortable with the notion that any facts but atomic facts are created by thinker/speakers—something for which Bergmann feels a deep repugnance—one cannot help but feel the force of this reasoning.22 (E) Bergmann’s account of mental acts rests on the contention that the constituent in an act determining its intention—the thought, or propositional character—is a simple nonrelational property, ontologically on par with, say, a specific shade of red. Bergmann does not acknowledge the existence of propositions, nonmental abstract objects that serve as the immediate objects of intentional states. Those who do embrace propositions generally regard them as complex objects whose structure mirrors or reflects the formal structure of the facts that would make them true (if those facts were to obtain). It is, presumably, this isomorphism of structure that allows the propositions to stand in for the facts that are their truth-makers when we reason deductively about nonactual facts (at least for those realistically inclined philosophers who reject entirely the view that the grammatical structure of the sentences expressing our thoughts creates the logical structure of the facts that are the truth-makers for these thoughts). When we reason from thoughts of nonactual facts to other such thoughts, however, it is puzzling how our reasoning is to be guided if these thoughts are simple nonrelational properties, since all such properties are of exactly the same form. If propositional characters are assumed to always intend an existent complex, whether potential or actual, then there will always be something—some existent—that possesses the logical form that guides deductive reasoning.23 22

In an earlier version of the NFO chapter on “Thought and Language” Bergmann suggests that the traditional view that thinkers literally construct negative facts out of the positive facts that are “really there” ultimately ends in absolute idealism. These comments unfortunately did not make it into the published version of this chapter. 23 In “Ontological Alternatives” (contained in Logic and Reality) Bergmann mentions that thoughts—propositional characters—are “unique in that they and they alone among the world’s things reflect its form” (p. 157). I believe that it is significant that in this essay—which is devoted to a discussion of Frege, who did embrace a conception of propositions as complex Platonic existents—is, as far as I can determine, the earliest assertion Bergmann makes of his case for the existence of potential facts. He

182 (F) Bergmann expends a significant proportion of his energies in NFO trying to find a way to ground the logical forms of all complexes (and, more generally, of all determinates), and to do so, moreover, in a way that makes a complex (or determinate) being of (each of) its logical forms expressible in the I.L. In short, for each complex (each determinate) he wants to specify another complex that grounds the former being of (one of each of) its logical forms. A complex may be, for example, each of the following: (i) a complex, (ii) a fact, (iii) a disjunction, (iv) a disjunction one of whose disjuncts is a nonrelational atomic fact of the lowest type and the other of whose disjuncts is an elementary diad,—and so on. Each of these is a logical form of the complex. Focus on being-a-disjunction, one of the logical forms of a-exemplifying-f-or-b-exemplifying-g. Call this fact ‘γ’. The general pattern of Bergmann’s assay of the complex a-exemplifying-for-b-exemplifying-g-being-a-disjunction is as follows: (∃α)(∃β)(γ=v(α,β)) [where the “variables” ‘α’ and ‘β’ range over all complexes]. This complex will be actual if and only if the “place” held by γ is a disjunction and will be potential otherwise. But the mode of this complex (Bergmann calls these form complexes) is entirely independent of the mode of the complex “put in the place” occupied in it by γ.24 More generally, a complex may actually be of a certain form even if it is not itself actual. This seems to make more sense if complexes exist whether or not they are actual. (G) In NFO Bergmann defines an analytic complex as one whose mode is either immediately or recursively accessible to those who intend the complex. He admits that this makes analyticity an “anthropocentric” notion, since it defines analyticity in terms of a “relation” between complexes and thinking beings. But corresponding to this anthropocentric notion, Bergmann claims, is the non-anthropocentric notion of a tautology. A complex is a tautology if and only if every other complex is of a form such that every complex that is also of that form is actual.25 (And a complex is a contradiction if and only if it is of a form such that every complex that is also of that form is potential). But in spite of the close connection between says in a footnote: “…‘P ’ and ‘not-P’ both represent entities, one a fact, one a p-fact” (p. 137). 24 All of this linguistic phraseology—“in the place of,” “variables,” etc.—must of course must be understood ontologically. 25 NFO, p.145.

183 the notions of form and tautologousness Bergmann insists that it is important to keep them distinct.26 It is, after all, possible to create calculi with more than the two “truth values” of standard logic. If we were never presented with a complex of a given form being actual (or potential), what basis could we have for supposing that all complexes of a given form are actual (or potential)? And if we were not on some occasions presented with complexes being of one (but not both) of the two modes, what basis could we have for supposing that the logic of our world is not specified by one of the many-valued calculi it is “possible” to construct? It will not do to appeal to “principles of logic” such as ‘(x)(f)[f(x)v-f(x)]’ since this is—by Bergmann’s lights, anyway—simply another statement which, like all statements, represents a fact. In order to know whether the statement is true, we must know whether the fact it represents is actual and we could not know this unless we were presented with its mode. The techniques typically used to assess the “modal status” of statements (e.g., truth tables) clearly themselves presuppose the “principles” of logic and so cannot be the ontological (or epistemological) ground of logic.27 These lines of reasoning—(A) to (G)—are not easily separated from one another. Nor is any of them alone, or even all of them taken together, likely to be entirely persuasive to those who find the notion of potential facts as repugnant as Bergmann finds many of the implications of rejecting them. Most resisters would probably be tempted to claim that the problems Bergmann sees with the alternatives to potential facts can be easily handled by treating certain types of phenomena subjunctively. Acts with nonactual intentions (nonveridical perceptions prominently among them) do not present their “owner” with an existing complex pervaded by the mode of po26

Ibid. Since Bergmann makes the ground of our knowledge that the logic of the world is a two-valued logic a matter of “experience” in this fashion and yet also explicitly denies that the modes of contingent complexes are ever presented it would not be unfair to ask him how he can be so sure that there really are only two modes. Since the modes of contingent complexes are never presented then how do I know that it is not the case that there are really more than two modes, some of which pervade complexes whose modes are not accessible to me? And even if I did know there are only two how do I know that some of these complexes don’t have more than one of the two modes? He does consider the question (NFO p. 86), claiming simply—and not altogether satisfactorily—that it is a matter of “extrapolation.” 27

184 tentiality; rather, they present their owner with what would exist if the intentions of these acts were to exist (if the act were veridical). Statements asserting the diversity of two complexes both of which are nonactual do not have as a condition of their truth the obtaining of an internal relationship of diversity between one or two existing determinates one or both of which happens to be potential; rather, the condition of their truth is the diversity between two complexes that would obtain if the two complexes were to exist. The validity of deductive inferences between thoughts with nonactual intentions is not based upon the formal features of potential facts; rather, they are based upon the formal features these facts would have if they were to exist. And so on. It far from obvious, however, that there is any benefit to be derived, in terms of ontological economy, from such a reconstrual. After all, subjunctive assertions still have to be made true (or false) by something. If we are to translate “a-being-blue-is diverse-from-c-being-green” into “abeing-blue-would-be-diverse-from-c-being-green if both of these facts were to exist” (or, if you like, “—if a were blue and c were green”) in order to avoid conveying the impression that there are true statements about merely potential facts. Then what makes the latter statement true? Perhaps the translation should go like this: “‘a is blue’ is a different sentence from ‘c is green’.” Then the truth makers for our problematic statements concerning potential facts would be metalinguistic facts. While the assumption that ontologically relevant features of the world can be stated only by speaking informally about the syntax and interpretation of an ideal language was indeed once a presupposition of Bergmann’s philosophical methodology it is clear that he always believed that these features are not merely metalinguistic facts. It is reasonably certain that Bergmann’s commitment to the existence of potential facts—as are virtually all of his ontological commitments—is motivated ultimately by his fervent realism.

Placing Bergmann FRED WILSON University of Toronto

T

his study aims to characterize and bring out some of the philosophical achievements of Gustav Bergmann, achievements that are too often not recognized. There are those who hold about fame that it’s a crap shoot. Then there are those who hold about fame that genius will out. Neither of these seems to me to be true. But what else? Well, there are those who explain the absence of recognition as the simple result of neglect. That seems too simple. Certainly, Bergmann was not the only significant philosopher who was never invited to present a paper at Princeton. Another, for example, was Roy Wood Sellars.1 Perhaps the neglect of both was the result of an assumption often made in the Eastern Division of the American Philosophical Association that what goes on in the Western Division can safely be ignored. Or perhaps the neglect was willful, as a way of avoiding too sharp a critic—it was easier, down east, for example, to simply let anger be one’s response to such a book as Bergmann’s Philosophy of Science2 rather than attempt a reasoned response to the criticisms of one’s position that are to be found in that book.3 Or perhaps it is because Bergmann is too much of a positivist,4 as he 1

See W. Preston Warren, Roy Wood Sellars. Boston: Twayne Publishers, (1975). Warren mentions a paper, “The Double Knowledge Approach to the Mind-Body Problem,” that Sellars read to the Aristotelian Society in London in 1922. He goes on, “Sellars had read a similar paper in French before the French Philosophical Society a year earlier. It is interesting that Sellars did not get comparable invitations in the United States. Like the Lowells and the Cabots of New England tradition, the Ivy League east was quite self-contained” p. 25. 2 Gustav Bergmann, Philosophy of Science. Madison: University of Wisconsin Press (1957). 3 See, for example, Hilary Putnam’s review of Bergmann’s Philosophy of Science, in the Philosophical Review, 69, (1960), pp. 276-77. 4 For Bergmann’s early experience as a member of the Vienna Circle, see his “Memories of the Vienna Circle: Letter to Otto Neurath,” in F. Stadler, ed., Scientific Philosophy: Origins and Development, Dordrecht: Kluwer, (1993), pp. 193-208. Neurath had

186 clearly is in his Philosophy of Science; and we all know that positivism is as dead as any philosophical movement can be.5 Or perhaps it is because he is too much a metaphysician, as he clearly is already in his first collection of essays, The Metaphysics of Logical Positivism and even more clearly in his later books. Positivism may be dead but it has left as its heritage a sentiment that ontology is not the thing one tries to do. These remarks suggest that Bergmann is both a positivist and an ontologist. But many, perhaps most, would think such a combination not really possible. Carnap, for one, thought not, and read Bergmann belatedly out of the Vienna Circle for arguing that one could have a positivistic metaphysics.6 Even those who are sympathetic to Bergmann’s views are inclined to say that he could not be both. Well, I am inclined to say the opposite, that he can be both, and I propose so to argue in this essay, and, indeed, to argue that the one supports the other. But just consider two titles. The first helped Bergmann in 1938 in his flight from Vienna. He made possible Bergmann’s passage to New York and all he asked in return was that Bergmann write this memoir. Bergmann composed it while crossing the Atlantic, discussing it daily and in detail with a fellow refugee on the ship, the Austrian novelist Hermann Broch. The two shared an apartment briefly in New York, until, with the help of Herbert Feigl, another member of the Vienna Circle, who had earlier fled to America, and Albert Einstein, Bergmann obtained a position at the University of Iowa as an assistant to the psychologist, Kurt Lewin. Bergmann soon concluded that Lewin’s gestalt theories were not up to positivist standards. But, when Feigl, who was in the Philosophy Department at Iowa, left to go to Minnesota, Bergmann was able to secure a position teaching philosophy of science in the Philosophy Department. He later met the behaviorist psychologist Kenneth. W. Spence who was in the Psychology Department, and a fruitful collaboration resulted. Bergmann was to remain at Iowa for the rest of his career. Everett Hall was then chair of the Philosophy Department at Iowa, and for several years, Wilfrid Sellars was also there. Bergmann had fruitful interactions with both. 5 John Passmore puts it simply: “Logical positivism is dead, or as dead as a philosophical movement ever becomes.” However, he does add a “But…”: “it has left a legacy behind.” See Passmore’s article “Logical Positivism,” in P. Edwards, ed., The Encyclopaedia of Philosophy, 8 vols., New York: Macmillan, (1967), vol. 5, pp. 5257, at p. 57. 6 The story is told in Herbert Hochberg, The Positivist and the Ontologist: Bergmann, Carnap, and Logical Realism, Amsterdam: Rodopi, (1994), p. 3. Bergmann had just published his paper, “A Positivist Metaphysics of Consciousness,” Mind, N.S. 54, (1945), pp. 193-226. As Bergmann told the story, Carnap, journeying from Chicago to California by train, made a stop in Iowa City to make clear to Bergmann his displeasure. “Leave the metaphysics to Feigl,” Carnap advised Bergmann.

187 is this, The Positivist and the Ontologist, a book by Herbert Hochberg.7 It contrasts Carnap who is taken to be the positivist and Bergmann who is taken to be the ontologist. The other title is this, “From Positivist to Realist: The Philosophy of Gustav Bergmann,” a study by William Heald.8 This title makes much the same point as the first. It does allow that Bergmann started as a positivist, something Hochberg also allows, though his title does not. But Heald argues, as Hochberg also argues, that Bergmann at some point ceased being a positivist and became an ontologist of some sort, in particular, a realist, in several meanings of ‘realism,’ a defender of universals (against the nominalists and tropians [such as Wilfrid Sellars, who was for some years his colleague at Iowa]), a defender of the existence of unobserved material objects (against the phenomenalism or at least the supposed subjectivism which it is supposed was the implicit ontology of the early positivists [Carnap fell unhappily into the expression “methodological solipsism” in his Aufbau9]), and a defender of the exis7

H. Hochberg, The Positivist and the Ontologist, op. cit. William Heald, “From Positivism to Realism: The Philosophy of Gustav Bergmann,” Books at Iowa, 56, (April, 1987), to be found on the internet at: http://www.lib.uiowa.edu/speccoll/Bai/heald.htm 9 Rudolf Carnap, Der logische Aufbau der Welt, Berlin: Weltkreis-Verlag, (1928). Trans. Rolf George, The Logical Structure of the World; with, Pseudoproblems in Philosophy, Berkeley: University of California Press, (1967). See § 63. Although it seems often to be thought otherwise, starting with a phenomenological method, one which takes as its starting point the ordinary world with which one is acquainted in ordinary experience, does not commit one to a phenomenalism which is subjectivistic. To say that one begins with what one is acquainted with is not to say that those entities are somehow mind-dependent. If one has a state of affairs given to one that one describes as “This is green” then there is nothing about that state of affairs that makes it mental or makes it ontologically dependent on the one (whatever or whoever that is) to whom it is given: it just is what it is, something that is green. Thus, to give one example of misreading Carnap as a subjectivist, see A. Coffa, “Idealism and the Aufbau,” in N. Rescher, ed., The Heritage of Logical Positivism, Lanham: University Press of America, (1985), pp.133-156. Coffa quotes from one of Carnap’s unpublished lectures: “Quite generally, everything that we talk about must be reducible to what I have experienced. Everything that I can know refers either to my feelings, representations, thoughts and so forth, or it is to be inferred from my perceptions. Each meaningful assertion, whether it concerns remote objects or complicated scientific concepts, must be translatable into a statement that speaks about contents of my experience and, indeed, at most about my perceptions” (Coffa, pp. 139-140). This is a clear statement that the language L that one is to use in laying out the 8

188 tence of logical objects (against the early Wittgenstein who held that the logic of the ideal language L represented things that do not exist: “My fundamental idea,” he said, “is that the ‘logical constants’ are not representatives; that there can be no representatives in the logic of facts.” (Tractatus, 4.0312)10—and Wittgenstein has been followed in this by almost everyone logic of science is to be interpreted in conformity with a Principle of Acquaintance. Yet Coffa takes it to show that Carnap holds that “all there is is reified experience” (p. 139). But nothing in this passage shows a commitment to a subjectivistic idealism. Of course, in the language that I understand, the basic or primitive descriptive constants will be interpreted into things and properties and relations that I have experienced, that is, that have been given to me in the world as I experience it. But from the fact that an object or property has occurred in the world as I have experienced it, it simply does not follow that that entity is somehow dependent, logically or ontologically, on me as a subject, or on my experiencing of them, or on the experiencing of them. Carnap’s remark that sentences of L must so to speak be “translatable into statements about the contents of my experience” simply states PA, nothing is implied to the effect that those “contents” are mind-dependent. Carnap may have his problems, but one cannot infer from this passage (as Coffa attempts to infer) that he was committed to a subjectivistic idealism. It remains true that Carnap spoke in his Aufbau of the experienced given with which he proposed to interpret his L as the “autopsychological” and, even more misleadingly, as “solipsistic” (see Chapter C, § 63). This way of speaking was unfortunate; it really is misleading, and invites unfounded criticism. At the same time, Carnap attempts to disarm any criticism that he has fallen into a subjectivistic metaphysics: he goes on almost immediately to point out that “the given does not have a subject” (§ 65). It is not self-contradictory to say of any fact or any “content of experience” with which Carnap begins that it exists unperceived. In fact, at the start of Carnap’s analysis, there is no concept of a self or subject, so at that stage it is simply meaningless to say of any of the basic facts either that it belongs to a self or that it does not. The problem of how to analyze the language of psychology does not arise until much later (Part IV, § 106 ff). And even then, the self at which one arrives is no simple substance, no Cartesian ego, and certainly not an active subject of the sort that would be required by and would delight a neo-Kantian or an idealist. Carnap is in philosophical agreement on this matter of the “subject of experience” with many others from Hume on, among them notably Nietzsche; compare the latter’s comment that “With regard to the superstitions of logicians [that is, the metaphysicians], I shall never tire of emphasising a small, terse fact, which is unwillingly recognised by these credulous minds—namely, that a thought comes when ‘it’ wishes, and not when ‘I’ wish; so that it is a perversion of the facts of the case to say that the subject ‘I’ is the condition of the predicate ‘thinks’.” Friedrich Nietzsche, Beyond Good and Evil, trans. R. J. Hollingdale, London: Penguin, (1990), § 18. 10 L. Wittgenstein, Tractatus Logico-Philosophicus, trans. D. F. Pears and B. F. McGuiness, with forward by B. Russell, London: Routledge, (1975).

189 else). Now, I think that it is quite correct to say that Bergmann became an ontologist, and that his final philosophy, his final ontology, was realist in all these senses of ‘realism,’ and indeed was its own unique form of wellargued and defended realism. But if Bergmann became an ontologist, I think it is misleading to say the he ceased to be a positivist—not only in the sense that he denied the existence of the transcendent entities that so many of the philosophers of the past felt forced to introduce, be it God or angels or absolutes or Ones or Platonic forms, but also in the sense that there is an important way in which his ontology, even in its late phases, remained positivist. Above all, his conception of reason as a way of knowing is that it is a human reason and is essentially positivist. What is reason? Surely it is the capacity to know or to grasp the reasons for things. Now consider a slim volume by Hans-Georg Gadamer. It is entitled Reason in the Age of Science.11 One notes already in the title of this essay the contrast that is being drawn between reason on the one hand and science on the other. One is inclined to say that it is surely science that discovers the causes of things. In that case, the contrast Gadamer is making really does not exist. Or, rather, Gadamer is presupposing that there are causes of things that the reason which is science cannot grasp and that it is the task of a deeper reason to grasp the deeper cause or causes— or, perhaps, the cause of the causes, that is, the real or underlying cause of the superficial causes that science locates. Philosophy, we are told, “has to do with the whole.”12 Science deals with the parts, philosophy with the whole. Well, the planetary system is a whole; for that matter, so is a basket of kumquats. However: ...this whole [the whole with which philosophy deals—that is, the whole with which philosophy as Gadamer conceives it is supposed to deal—this whole] is not merely, as is true of any other whole, the whole comprised of all its parts. As the whole, it is an idea that transcends every finite possibility of knowledge, and so it is nothing we could know in a scien13 tific way.

Yet we can still speak of the scientific nature of philosophy: 11

Hans-Georg Gadamer, Reason in the Age of Science, trans. Frederick G. Lawrence, Cambridge: MIT Press, (1981). 12 Ibid., p. 1. 13 Ibid.

190 ...philosophy can justly be called scientific because in spite of every difference from the positive sciences, it still possesses a binding proximity to them that separates it from the real of the world view based on strictly 14 subjective evidence.

There is an objective reality to the whole beyond what any positive science can grasp, and this reality can be known through that reality grasping the mind and binding it to it. There is, we are told, a “natural inclination towards philosophy”15 that science cannot satisfy yet resists it. Science makes progress “by renouncing comprehensive knowledge in the grand Aristotelian style.”16 Inasmuch as it [science] subjected what is observable to the quantifying methods of mathematics, empirical science discovered a new notion of natural law and it moved forwards to scientific knowledge in all direc17 tions by means of experiment and hypothesis.

Note that this is just wrong: Galileo and Newton of the Principia used mathematics. They also used hypotheses and experiments. But Newton in the Opticks used experiment without mathematics. And Bacon uncovered the logic of experiment, and it was not mathematical. The logic of experiment, whether in Bacon or Newton or Hume or Mill, is a matter of hypothesis testing. These hypotheses concern the regularities that describe the behavior of things in the world, the world of ordinary experience.18 The hypotheses aim to capture the general patterns of how things of the world work—from stones to stars to oysters to shadows to people.19 If reason is the capacity to grasp the causes of things, then sci14

Ibid. Ibid., p. 139. 16 Ibid., p. 144. 17 Ibid. 18 This was the view of Hume and John Stuart Mill. It was defended by the positivists of the Vienna Circle. Carnap came to qualify it, but Bergmann remained committed to defending it; see his Philosophy of Science, Chapter One. For an extended defence of the idea, see F. Wilson, Explanation, Causation and Deduction, Dordrecht: D. Reidel, (1985), and Laws and Other Worlds, Dordrecht: D. Reidel, (1986). 19 It is well to emphasize that whether one is a positivist or not, the experimental method is one that aims to uncover regularities, general patterns of behavior: it aims to eliminate hypotheses by putting them to the test; those for which there is a counterexample are rejected as false and thereby eliminated; it is empirical regularities—matter15

191 entific reason is the capacity to grasp the regularities among ordinary things—regularities or general patterns that are to be found in the ordinary world. These regularities constitute the causal structure of the world of ordinary experience. Thus, for the positivists, reason as the capacity to grasp the causes of things is the capacity to know the regularities that describe the behaviour of things in the world of ordinary experience.20 Thus, the reason of the positivists, in its aim of understanding the things of the world of ordinary experience, does not depart from the world of experience. This is in contrast to the reason of Aristotle, which sought to penetrate to a reality of causes that lie outside the world of ordinary experience. And also, of course, the reason of the positivists is in contrast to the reason of which Gadamer speaks—the reason which he assumes a priori and for which he offers no defence.21 Science aims to grasp regularities. The methods of science aim at discovering those regularities. Science does this by putting hypotheses about the regularities to the test, the test of experiment.22 These methods define scientific reason, the reason of the positivists—the reason of common sense. Scientists put hypotheses to the test. There are always a number of possible hypotheses. Which hypothesis it is that is put to the test is not one that is merely randomly chosen. There is method to the method of science. For the experimenter, there is given a range of possible hypotheses, and experiments, whether natural or artificial, take place, in which hypotheses within the range are tested; if one fails the test it is eliminated as falsified; of-fact generalizations—that have the logical form that permits of elimination by counterexamples. The non-positivists and non-Humeans want to add something to this, before a regularity can be counted as a law. But that is not what the method of experiment aims at discovering. 20 As Bergmann notes in his essay on “Logical Positivism,” it is one of the marks common to several versions of logical positivism that a positivist accepts the Humean account of causation—to know a cause is to know a regularity. See his “Logical Positivism,” in his Metaphysics of Logical Positivism, first edition, New York: Longmans, Green, (1954), pp. 1-16, at p. 2. 21 See F. Wilson, The Logic and Methodology of Science in Early Modern Thought: Seven Studies, Toronto: University of Toronto Press, (1999). 22 For the logic of the scientific method, and of experiment in particular, see F. Wilson, The Logic and Methodology of Science and Pseudoscience Toronto: Canadian Scholars Press, (2000).

192 what remains in the end as uneliminated is accepted as true—accepted tentatively at least, until the next round of experiments. This conclusion that the uneliminated hypothesis is true is warranted provided one accepts the Principle of Determinism for that area, the Principle that for the phenomena in question there is a sufficient (or necessary and sufficient) condition that is regularly connected to it, and the Principle of Limited Variety that this condition can be found within a specified range, that it is one of the species within a given genus. These Principles guide the researcher as he or she attempts to locate order in the world.23 These Principles are laws about laws, as Mill put it.24 Bergmann’s philosophy of science is through and through positivist. So much so that, as we have remarked, it evoked intemperate responses among those who think it is bad to be old fashioned.25 But unlike most, even among the positivists, he discussed laws about laws. To be sure, he concentrated, which for his purposes in the philosophy of psychology was perfectly reasonable, on one such law, the composition law that is present in many theories, and in particular in classical mechanics, with its law for the vector addition of forces in complex mechanical systems.26 Bergmann emphasizes that such a law is in fact a 23

For details of this account of science, see F. Wilson, The Logic and Methodology of Science and Pseudoscience. This study makes clear the relation of Kuhn’s notion of a “paradigm” to the account just sketched of theoretical structure in terms of the Principles of Determinism and Limited Variety understood as requiring formulation in higher-order logic. See also the following studies by F. Wilson: Hume’s Defence of Causal Inference, Toronto: University of Toronto Press, (1997), which has a discussion of aspects of the confirmation of such theories; and has a discussion of such theories and Kuhn’s notion of “revolutionary science”; The Logic and Methodology of Science in Early Modern Thought: Seven Studies, which gives something of the historical context for the emergence of such theories and of modern science; and Empiricism and Darwin’s Science, Dordrecht: D. Reidel, (1999), which applies this logic of theories to an analysis of Darwin’s theory in particular. The relevance of this account of the logic of theory structure to the logic of scientific explanation is explored in F. Wilson, Explanation, Causation and Deduction. See also F. Wilson, “A Note on Hempel on the Logic of Reduction,” International Logic Review, 13, (1982), pp. 17-29. 24 As Mill put it, “it is a law that there is a law about everything”; see John Stuart Mill, System of Logic: Ratiocinative and Inductive, London: Macmillan, (1974), Bk. III, ch. v, § 1. 25 See the review by Putnam mentioned in note 3. 26 Bergmann, Philosophy of Science, Chapter Three.

193 law—a law about laws, a generalization about generalizations, a regularity about regularities. The Principles of Determinism and Limited Variety in a given area are laws of this sort. The basic form of a law is that of a universal conditional formulated in first order logic. Laws about laws have to be formulated, in their simplest form, in second order logic, but all that is part of the logic of Principia Mathematica that the positivists and, following them, Bergmann accepted as giving the basic logical structure of scientific theories and, indeed, the logical structure of the world.27 The central core of modern science is observation and experiment.28 The point is that the logic of modern science is given by the logic of elimination, guided by the Principles of Determinism and Limited Variety.29 To repeat, the central core of modern science is observation and experiment. It is not mathematics. Mathematics is of course nice if you have it, but it is not essential. But mathematics or the Principles of Determinism and Limited Variety, or whatever, it is simply not the case, contrary to Gadamer, that modern science “subjects the world” to this form; rather, it discovers that certain sorts of mathematics fit the world, and it discovers that there 27

Carnap, in his Introduction to Symbolic Logic and Its Applications, New York: Dover, (1958), gives several examples of scientific theories the logical structures of which require for their formulation higher-order logic. But when he came to try to study what he called logical probability, Carnap restricted himself to languages formulated in first-order logic, which foreclosed the possibility of his studying the confirmation of any theories requiring higher-order logic, and therefore prevented him from dealing with the confirmation of any Principles of Determinism and of Limited Variety. (In fact, it was even more restrictive: he used only first-order logic with monadic or non-relational predicates. That is a pretty impoverished language for attempting to analyze the logic of scientific theories. But many would argue, among them Bergmann, that the whole notion of “logical probability” is empty. See G. Bergmann, “Frequencies, Probabilities, and Positivism,” Philosophy and Phenomenological Research, 6, (1945), pp. 26-44. It is perhaps worth noting that Carnap got the basic idea for his “logical probability” from the Tractatus. Carnap develops this basic idea in formidable detail and with a symbolic notation that is at once confusing and fascinating; but the core idea is Wittgenstein’s and Carnap’s contribution is the simple one of elaborating on this using the tools of elementary combinatoric arithmetic. See G. Bergmann, “Some Comments on Carnap’s Logic of Induction,” Philosophy of Science, 13, (1946), pp. 71-78. 28 For the historical context, see Fred Wilson, The Logic and Methodology of Science in Early Modern Thought: Seven Studies. 29 See Fred Wilson, The Logic and Methodology of Science in Early Modern Thought: Seven Studies, Study One.

194 are certain experiment-guiding Principles that hold of the world: these things, it must be emphasized, are discovered and are not imposed. As Bergmann recognizes, the laws about laws that guide research are matters of fact about the world, matters of fact that we discover. Gadamer’s contrary suggestion that they are imposed on the world is just wrong: it is just wrong to suggest that modern empirical science is an “attack on nature, which it subjects to a new but only partial mastery.”30 It is Gadamer’s further idea that the new science has pretensions to comprehensiveness that in the end, since it is restricted to the observable, it cannot fulfill, not in the way that the older metaphysics aspired to comprehensiveness. What the old science [of the Greeks], crowned by metaphysics, had provided was a whole orientation to the world, which brought the natural experience of the world and its linguistically mediated interpretation of 31 the world. Modern science could not provide this.

There is, then, the natural human inclination to philosophy, to philosophical knowledge, knowledge of the whole as a whole that objectively determines its parts, an inclination that science cannot fulfill but which with its own aspirations works to frustrate. Philosophy must recover itself, and put science in its correct place as less than humanly fulfilling knowledge, if we are to satisfy our natural inclination to knowledge about the deeper and more genuine and real causes of things. Genuine reason must recover itself and take its place as more genuinely human than mere empirical science. Only then, with metaphysics once again re-installed as the only genuine source and object of reason, will our human cognitive aims find fulfilment. Now, there is no question that science, that is, empirical science, works: it does discover causes. We can build efficient steam engines, we can control childbed fever in women, we can send text messages wirelessly. But this, Gadamer proposes, is not enough: there is, he suggests, something more. But nowhere does he give any justification of the claim that there is such knowledge beyond scientific knowledge or that there is a natural human inclination to knowledge of the whole as such: Gadamer merely asserts. 30 31

Gadamer, (1981), p. 145. Gadamer, (1981), p. 144.

195 It is perhaps needless to say that this is hardly a defence of philosophy or of metaphysics: it is, to the contrary, contemptuous of genuine thought: for, it takes it that thought is not necessary, mere assertion will do. And then science will be put in its place, the role of reason in its best sense restored. A little more assertion and the trick will be done. One here is compelled to recall Nietzsche’s comments on metaphysics. Nietzsche32 ridiculed the love of truth as a motive for philosophers.33 They were in fact moved by such things as the need to secure a university chair, or even to become a Philosopher-King, someone of world-historical importance. But mostly Nietzsche’s argument was on the one hand the positivist argument that transcendental metaphysics is illusion and, on the 32

We cannot underestimate the impact of Nietzsche; for a just evaluation, see Thomas Mann, “Nietzsche’s Philosophy in the Light of Recent History,” in his Last Essays, trans. Richard and Clara Winston and Tania and James Stern, London: Secker and Warburg, (1959), pp. 141-177. The impact extended to the Vienna Circle. Herbert Feigl, in his essay “Moritz Schlick: A Memoir,” in Eugene T. Gadol, ed., Rationality and Science, pp. 55-82, indicates (p. 62) the importance of Nietzsche for Schlick’s philosophy. Schlick himself invokes the authority of Nietzsche in his essay “On the Meaning of Life,” in his Philosophical Papers, vol. II, pp. 112-129, at p. 113, p. 125. Carnap refers to Nietzsche’s Will to Power three times in the Aufbau, citing (in § 65 and 67) Nietzsche’s views on the self and the subject (there is no primitive self, no Cartesian ego) and no primitive activity (in § 163); he agrees with Nietzsche, in other words, that there is no metaphysical subject on which the world is somehow dependent and no unanalyzable activity that brings the world as we know it into being and endowing it with a necessary structure. Both Nietzsche and Carnap in effect appeal to the same positivist principles to exorcise from the world these holdovers from obscurantist metaphysics. It is worth noting that Heidegger too reads Nietzsche as arguing for a positivist position; see M. Heidegger, Nietzsche, 3 vols., trans. David F. Krell, New York: Harper and Row, (1979), vol. 1, § 20, p. 151ff. Not that Heidegger has any deep understanding of either science or positivism. Some of his objections to positivism are discussed below. We should say this, however, at this point: That when Carnap drifted away from the early positivist position on logic (in his 1936 essay on “Testability and Meaning,” Philosophy of Science, 3, (1936), pp. 419-471, and 4, (1937), pp. 1-40, he left himself open to the sort of criticisms Heidegger offered. Whether these amount to revealing Being is another matter. We discuss these issues below. 33 See Friedrich Nietzsche, Beyond Good and Evil, Part 1, on the “Prejudices of Philosophers.” Thus, he writes that “It has gradually become clear to me what every great philosophy up till now has consisted of—namely, the confession of its originator, and a species of involuntary and unconscious auto-biography; and moreover that the moral (or immoral) purpose in every philosophy has constituted the true vital germ out of which the entire plant has always grown” (§ 6).

196 other hand, the argument that these illusions were disguised wishes the aim of which was to enchain humankind with ostensibly objective duties.34 These duties were not, when it came down to justification, rules that enabled people to live together, though they are that. The will to power is the will to command others. One commands others in the enterprise of satisfying one’s own instinctual urges. One commands them not personally but through the illusion of objective duties. German romantic metaphysics provided the rationale once the illusions of religion had lost their force. Somehow, these metaphysicians said, there is an objective self or being that is beyond the world of ordinary experience that commands us. There is Being—majuscule “B”—through immersion in which we are to find out the truth of things, the deeper truth that lies behind the world of ordinary sense, and as part of that our real value and meaning. It is precisely this same sort of thing that the positivists argued against, this sort of dogma made pretentious and passing for profundity, the attack on science that proposes to put it in its place and substitute for it what amounts to mere fantasy. Gadamer asks us to go back to the way philosophy was done in the ancient world: we are to let ourselves be immersed in what amount to the forms or essences of things, those forms or essences that lie beyond the world of ordinary sense experience but define what is real and true in this ordinary world. He laments the apparent passing of this way of thinking ourselves into things, but clearly thinks it is a way of thought that can be recovered. Indeed, it is apparent that he thinks that he has somehow himself recovered it, and that he at least has immersed himself in it. But notice once again how he simply assumes that such an approach to the way things are makes sense, how he simply as34

See Beyond Good and Evil, § 11, on Kant. Nietzsche rightly points out that the Kantian appeal to categories is no more explanatory than an appeal to dormitive powers. He goes on to indicate how it became fashionable for German philosophers to find faculties for discerning things transcendental. But it was all illusion: “Enough however—the world grew older and the dream vanished” (p. 17). He has already remarked ( § 5) that “The spectacle of the Tartuffery of old Kant, equally stiff and decent, with which he entices us into the dialectic by-ways that lead (more correctly mislead) to his ‘categorical imperative’—makes us fastidious ones smile, we who find no small amusement in spying out his subtle tricks of old moralists and subtle preachers.” Compare the argument in F. Wilson, “Science and Religion: No Irenics Here,” Metaphysica, 7, (2006), pp. 159-170, and also in P. Kurtz, ed., Science and Religion: Are they Compatible?, Buffalo: Prometheus Books (2007).

197 sumes that there are the forms or essences and that the mind does have the capacity to rise above sense experience and grasp these deeper structures. There is no argument. There are no instructions as it were as to how we are to achieve this way of knowing. In sense experience, the world as it were grabs hold of one; the sensible world so grabs hold of us that cognitively we find ourselves immersed in that world. There is no avoiding the world of sense and our knowing of it. It and our experiencing of it are there to be described. But Gadamer would have us go beyond this world to one that as it were lies outside this ordinary world, encompassing it as a whole but lying beyond or behind it, and from which this ordinary world emerges. But is there such a world? We cannot take seriously the claim that there is this deeper reality, not until one has some sort of argument for the existence of such a world beyond the ordinary, or a technique that enables those of us who are not immediately aware of it, to focus our cognitive capacities so that we too can grasp it. But we are not given any of those things: Gadamer simply assumes a priori that the world beyond our world that determines the shape of our world does exist and that some of us, him at least, grasp that world. But in the absence of argument, the assumption, we must repeat, can amount to nothing more than fantasy. In the absence of anything like modern science, in the absence of the understanding of the world achieved by the latter, such fantasy can perhaps be excused. But not now: in the present age it is pretentious fantasy disguised as metaphysical profundity and deserves the contempt with which the positivists, following Hume and Nietzsche, argued for its dismissal.35 35

Compare Nietzsche’s remark: “—what excellent tools for observation we have in our senses! ...We have science these days precisely to the extent that we have decided to accept the testimony of the senses, to the extent that we have learned to sharpen them, and think them through to the end.” One is reminded here of Bergmann’s comment that science is the long arm of commonsense. Nietzsche continues: “Everything else is deformity and pre-science: I mean metaphysics, theology, psychology, epistemology. Or formal sciences, a system of signs: like logic and that application of logic, mathematics. They do not have anything to do with reality, not even as a problem...” Twilight of the Idols in F. Nietzsche, The Anti-Christ, Ecce Homo, Twilight of the Idols, ed. A. Ridley and J. Norman, trans. J. Norman, Cambridge: Cambridge University Press, (2005), Chapter on “‘Reason’ in Philosophy,” § 3, p. 168. One is reminded also of Hume’s remark at the end of the Enquiry concerning Human Understanding: “When we run over libraries, persuaded of these principles, what havoc must we make? If we take in our hand any volume of divinity or school

198 Gadamer is part of what might be called the Aristotelian tradition.36 This tradition did not begin by looking about the world and asking what sorts of things do exist and then proceed to categorize them. Rather, it argued on an a priori basis that there were substances and accidents, and that each substance had a nature or essence which determined the kind of substance it was and which it shared with other substances of the same kind; and if there were things we knew that did not fit these patterns—e.g., relations—then they were declared not to exist. Far from being a cautious sort of empiricism, as its defenders so often claim, the Aristotelian tradition proceeded on the basis of dogmatically asserted a priori principles, which they disguise as insights into the a priori metaphysical structure of the world. Nor could it do otherwise. It had its own a priori thesis about knowledge and this was imposed on the world in a way that simply denied the phenomenological detail that did not fit the patterns. This dogma about knowledge was that “like knows like” and this was interpreted to mean that the knowing mind became identical with the object known. Since the mind could not literally become the material object presented in sense, sense could not be a form of knowledge; what came to be known, or rather, what came to be in the knowing mind, was the nature or essence of the sensed object. This doctrine of the two eyes—the eye of the senses, that is aware of sensible particulars which the object known presented to the knower, and the mind’s eye, which knew the universal, the nature or essence in the material object known—this doctrine of the two eyes is the most pernicious feature of the Aristotelian legacy.37 It prevented philosophers from recognizing that in sense experience we are presented not only with particulars but also with their properties, and not just particulars and properties but also the facts of the particulars having certain properties; and not only properties but also relations and the facts of various particulars standmetaphysics, for instance, let us ask, Does it contain any abstract reasoning concerning quantity or number? No. Does it contain any abstract reasoning concerning matter of fact and existence? No. Commit it then to the flames: for it can contain nothing but sophistry and illusion.” D. Hume, Enquiry Concerning Human Understanding, ed. L. A. Selby-Bigge, rev. P. Nidditch, Oxford: Oxford University Press, (1975), p. 165. 36 For an extended discussion of the Aristotelian tradition, see F. Wilson, The Logic and Methodology of Science in Early Modern Thought: Seven Studies, Study One. 37 See R. Grossmann, “Conceptualism,” in E. B. Allaire et al., Essays in Ontology, The Hague: Nijhoff, (1964), pp. 30-39.

199 ing in certain relations. When the empiricists attacked Aristotelian natures and essences, they also attacked the mind’s “eye” that saw these things. That left them with the eye of the senses that grasped only sensible particulars. The resulting nominalism was, of course, untenable, as Russell for example argued, and one can again and again feel Locke, Berkeley and Hume struggling to affirm phenomenological fact while still gripped by this uneliminated Aristotelian dogma about the eye of the senses. However, as the rejection of substances forced an examination of the problem of relations, so the examination of the latter permitted a re-examination of the relation of a mind to what it knows. This elimination of the dross of the Aristotelian tradition permitted philosophers to accept, on the one hand, without dogmatic distortion phenomenological accounts of what is given in experience,38 including acts of experiencing themselves; and also permitted them, on the other hand, to develop an account of experiencing adequate both to phenomenological detail and the philosophical demands of categorial analysis. It was the virtue of the positivists that they began with a phenomenological method. They began, as did the empiricists, and as did Mach, and as did Mill and Hume, with the world of ordinary experience. This they took for granted: philosophy began with its description. The language they proposed to use for its description was that of Principia Mathematica, that is, an ideal language with the syntax of formal or symbolic logic together with descriptive predicates giving semantic or cognitive content by designation rules or rules of reference through which they are applied, on the one hand, to individuals in this world of ordinary sensible experience and, on the other hand, to properties of those individuals and relations among them. This was the empiricist’s language L. There were disagreements, naturally enough, about the domain of individuals. Did it include not only sense impressions but also physical objects—that is, ought one to have a phenomenalistic base or a physicalistic base? Were the constants to be taken as referring to experiential wholes, as the gestalters were inclined to suggest, and as Carnap followed them in this in his Aufbau? or was one to take the base more atomistically, beginning with individual parts of 38

See F. Wilson, “Acquaintance, Ontology and Knowledge,” The New Scholasticism, 54, (1970), pp. 1-48, and “Effability, Ontology and Method,” Philosophy Research Archives, 9, (1983), pp. 419-470.

200 these whole, as Russell and Mach would have it? But if there was disagreement about exactly how best to describe the world of ordinary sensible experience, there was agreement that if some entity was to be philosophically acceptable, descriptions of it had to be expressible in this language that described the world of ordinary experience. If some entity was proposed, then it was incumbent upon the one who proposed it to make clear how it was part of the world of ordinary experience. This world could contain things very small, too small to see, such things as germs or even atoms, and things that are very far away, too far away to see, such things as mountains on the moon and distant stars and galaxies. Germs and stars could be accommodated; one knew how to fit them into the world of ordinary experience, even if one did not understand precisely the logical form of the relations by which that accommodation was effected. What, precisely, is the logic of microscopes? There were also the efforts of these philosophers to use the techniques of the new logic to understand old problems, e.g., the nature of measurement. Some of the discoveries were profound. They discovered, for example, that the old distinction between quality and quantity was mistaken: quantities turned out to be qualities differing from “mere” qualities only in their relations that order them in certain ways.39 Discoveries such as the last were impossible under the Aristotelian system, which denied the reality of relations. It was the new logic that was important here. But it was also the phenomenological turn, which one finds in James and Russell and later in the gestalters, that allowed that relations are among the entities we experience. But all were agreed that unless some proposed entity could be accommodated, somehow or other, in the world of experience and in the L that describes this world, then it could be rejected as philosophically inadmissible—it was “metaphysics,” as they said, “pseudo-statements,” and “meaningless.” Something had literal or philosophically admissible meaning only to the extent that it could be fit into L. Otherwise, it was devoid of “cognitive content.” Some positivists went through a series of logical contortions trying to give a precise meaning to the verification principle, one that could be used 39

See G. Bergmann, “The Logic of Measurement,” in Proceedings of the Sixth Hydraulics Conference, Iowa City: State University of Iowa Publications, (1956), pp. 1934.

201 almost a priori to determine what was acceptable and what was not.40 The ingenious were, as always, able to find counterexamples, and the exercise became a sort of scholasticism of the worst sort.41 But there can be no way that simply a priori one can determine once for all which statements have cognitive content and which don’t. One has to examine them as they arise, and give each case its own analysis. Some are obvious: not only is there 40

It is interesting to note how Carnap deploys what amounts to the verifiability principle in The Logical Syntax of Language, London: Routledge and Kegan Paul, (1937); see the § “Philosophy Replaced by the Logic of Science,” pp. 277ff, and “The Logic of Science is the Syntax of the Language of Science,” pp. 281ff—these are followed by “Pseudo-Object Sentences,” pp. 284ff, and “Sentences without Meaning,” pp. 288ff. Carnap insists that the syntax of one’s philosophically perspicuous language L is constituted by the symbolic logic of Principia Mathematica, but there is little about the interpretation of L. But as Bergmann saw, the latter is necessary, that is why we need PA. If all you have is syntax, why not allow as meaningful a formalization of the language of Aristotelian essences? Carnap is too much a formalist: if all one has is syntax then one has little ground for excluding any sort of metaphysics, so long as someone who is just as good at formalization can invent a syntax for the language in which that metaphysics is expressed. One can exclude such “entities” only if, besides syntax, one can appeal to experience, to Bergmann’s PA. Carnap really does know this, but in the Logical Syntax of Language he is so caught up in his work, perfectly sound, in formalizing the language of science that he cannot recognize the need for the appeal to experience. 41 Bergmann gave the following rough definition of a logical positivist: he or she (a) holds a Humean view on causation and induction; (b) insists on the tautological nature of logical and mathematical truths; (c) conceives of philosophy as logical analysis, that is, the clarification of the language we speak in everyday life; and (d) that such analysis leads to the “rejection of metaphysics” in the sense that, e.g., the points of dispute among the traditional forms of idealism, realism, phenomenalism, could not even be stated, or, at least not be stated in their original intent, in a properly clarified language. Bergmann refers to these as “diplomatic formulae.” (“Logical Positivism,” p. 2). It is clear that he thinks that the technical literature about the “verifiability principle” is of little interest. It is important to note, however, that Bergmann qualifies point (d) regarding the rejection of metaphysics. For, he also goes on to point out that it is, or ought to be, part of the positivist’s job, as a philosopher, to deal with the traditional issues, recovering their “commonsense core,” by construing them as positions about the nature of the philosopher’s language L, its syntax and/or interpretation (p. 7). He suggests that the proper L has the syntax of, roughly, Principia Mathematica, (p. 7) and that its proper interpretation is by means of the “principle of acquaintance,” that is, its undefined descriptive constants refer to entities that are “phenomenally given” (pp. 6-7). One may reject metaphysics when it is done in the old way, but remain a metaphysician (ontologist) when one takes up ontological issues as they come to be re-constructed.

202 no God but discourse about him or her is indeed devoid of cognitive content.42 Others are not so clear. Are there universals? for example. Bergmann did not waste his time on the obvious cases. What he tried to do was take the problematic cases and try to find where they were nonsensical and where they were legitimate. The problematic cases had their own peculiar “feel” as it were. He saw it as the task of philosophy to uncover the real content of these puzzling cases. He argued that they, most of them at least, could be understood as trying to say in what Carnap called the material mode of speech what is correctly said in the formal mode of speech.43 One could take the puzzling cases and re-formulate them as providing answers to questions about the proper empiricist’s language L, that is, its syntax and its interpretation (semantics).44 42

A brief history of the vicissitudes of the “verifiability principle” can be found in C. G. Hempel, “The Concept of Cognitive Significance,” Proceedings of the American Academy of Arts and Sciences, 80 (1951), pp. 61-77. The outcome, Hempel suggests, is that the principle can’t do what it was intended to do, namely, eliminate metaphysics. He ignores the fact, emphasized by Bergmann, that even a rough formulation will do the required job. C. Misak, Verificationsim, London: Routledge, (1995), follows Hempel in his history and in his pessimistic conclusion that metaphysics escapes unscathed; see her Chapter Two. She ignores Bergmann’s discussion in his “Comments on Professor Hempel’s ‘Concept of Cognitive Significance’,” in his Metaphysics of Logical Positivism, pp. 253-267. Hempel raises the problem of translating contrary-tofact conditionals into a positivistic L modelled on P.M. Bergmann offers a solution to the problem in his “Comments.” Misak may think it wrong, but it does deserve a mention and an indication of her reasons for rejecting it. Misak also raises the old canard about statements regarding the past or future being unverifiable. Bergmann discussed this issue too, in his “Remarks on Realism,” (in his Metaphysics of Logical Positivism, pp. 153-175), which again she fails to discuss. One might also mention that she fails to discuss Bergmann’s essay on “Logical Positivism,” which gives a rather different slant on logical positivism than one finds in Hempel. A greater effort on her part to deal with the literature on the topic might not have changed her conclusions, but it would have made her discussion historically more adequate and certainly more nuanced and fair. 43 R. Carnap, The Logical Syntax of Language, pp. 298ff. 44 Carnap misses the point about the philosophical problems of universals. In the Logical Syntax of Language, § 76-77, he discusses “universal words”, but besides pointing out that the material mode of speech is a source of confusion, he slides into a conventionalism. “Universal words” occur in language as variables; thus, ‘x’ in “for all x, ...” is a universal concept. Understood as the concept “object” it is in the material mode of speech and is, as Wittgenstein said, a “pseudo-concept,” one that is apt to be philosophically misleading. But if one puts it into the formal mode of speech as

203 “variable,” one is speaking correctly, in a way that does not generate the problems characteristic of philosophical puzzlement. Carnap (Logical Syntax, p. 295) cites the Tractatus, where we find the following assertions: 4.1271: “... the variable name ‘x’ is the proper sign for the pseudo-concept object”; and 4.1274: “To ask whether a formal concept exists is nonsensical,” and 4.121: “Propositions show the logical form of reality. They display it,” and 4.1212: “What can be shown, cannot be said.” But, Carnap states that “I do not share Wittgenstein’s opinion this method of employing universal words [in the formal mode of speech] is the only admissible one” (Logical Syntax of Language, p. 295). There is also the material mode of speech (pp. 295-6). But though the material mode is admissible, its use can be philosophically misleading: “The use of the material mode of speech leads...to a disregard of the relativity to language of philosophical sentences; it is responsible for an erroneous conception of philosophical sentences as absolute” (p. 299). Thus, Carnap holds, contrary to Wittgenstein (and Bergmann) that logical form, e.g., particularity, is conventional. However, two points need to be made. Both turn on the fact that Carnap allows the variable ‘object’ to range over various things, from numbers to conditions like friendship; what counts as an object is a matter of convention, the meaning of the universal term is established by a decision. The meaning of ‘object’ is given, not by its reference or connection to the world, but simply by the syntactical rules (L- or logical rules) that connect it to other elements of logical form. It gains whatever cognitive meaning it has by these syntactical rules which as one says “implicitly define” it. First, Carnap’s method, unlike Bergmann’s, does not lead to the solution/ dissolution of traditional philosophical problems. Thus, for example, there is a traditional problem, discussed at length by philosophers such as Moore, Russell, and Stout, about whether the characteristics of things, that is, sensible things, are universal or particular. This deals with objects as given in sensible experience. Can the same property be in two different things, and therefore be a universal, or are the properties of things as particular as the things of which they are the properties? To deal with this issue one must consider the world as given in sensible experience and talk about the interpretation of L into that world. Bergmann of course does this for his L via his appeal to PA as a rule for interpreting descriptive constants. Carnap in contrast simply ignores the fact that L must be interpreted, and therefore ignores the hook-up of L to the world. In particular, he fails to give any special consideration to the world as given in sensible experience. This in turn means that, in contrast to Bergmann, he has chosen in his purely syntactical approach to problems a method that so narrows the philosophical discussion that he avoids the issue—the philosophical problem—raised by Russell, Moore, and Stout. Bergmann’s method with its appeal to PA generates solutions to traditional philosophical problems, some of them anyway, where Carnap’s syntactical method merely avoids those issues. Second, Carnap does not notice that when he allows the term ‘object’ to have its sense conventionally determined by its syntactical connections to other logical words, he is implying that there is no real distinction between formal and material properties (to speak as the Tractatus does on these things), and thereby creates a slippery slope that leads to the conclusion of linguistic nominalists such as Sellars that all properties

204 are conventional. (See Sellars’ “Naming and Saying” in his Science, Perception and Reality, London: Routledge and Kegan Paul, (1963), pp. 225-246, and “Meaning and Inference,” Mind, n. s., 62, (1953), pp. 313-338.) Bergmann, in contrast, insists that material properties are objectively there in the things the world presents to us, and further insists that the properties such as particularity which are represented in L as the logical forms of things are also objectively there in things. (See his “Ineffability, Ontology and Method,” in his Logic and Reality, Madison: University of Wisconsin Press, (1964), pp. 45-63.) Wittgenstein said that formal properties such as particularity do exist, objectively in the world, but are ineffable and such that discourse about them is meaningless. Carnap took the meaninglessness seriously, and concluded that they obtain their meaning not by reference to an objective feature of the world given to us in experience but simply to other logical or formal concepts by the logical rules in which they occur. Bergmann, in contrast, accepted Wittgenstein’s point that logical form is objectively there in the world but rejected the claim that talk about such entities was meaningless and found a way to make them effable after all. This realism about logical form is one of the ways in which Bergmann is a realist. Then there is Sellars, who accepts Carnap’s view on logical form and then extends it to all properties, those that are material in Wittgenstein’s sense as well as those that are formal: all sameness and difference, material as well as formal, is ineffable and therefore determined only conventionally. Nominalism about logical form is extended to become a nominalism about all properties. We discuss Sellars’ nominalism below. It should perhaps be noted, however, that Carnap’s idea the syntactical categories receive their meaning from syntactical rules that “implicitly define” them has its roots in the Tractatus. That the categories represent is there in the Tractatus, but also present in that very frustrating work is the idea that they receive their meaning from the rules in which they occur. Thus, for example, the Tractatus tells us that: 3.334: “The rules of logical syntax must go without saying, once we know how each individual sign signifies.” What this says, in effect, is that once we know the type of a name, we know the formation rules for sentences containing such names; in other words, the rule that a-signs combine with f-signs in sentences of the form ‘fa’ is part of the meaning of what it is to be an a-sign—the type rule follows from the type distinction. Meaning in the sense of reference does not enter into syntax: 3.33: “In logical syntax the meaning [reference, Bedeutung] of a sign should never play a rôle.” But the syntax is not irrelevant to the representation: 4.12721: “A formal concept is given immediately any object falling under it is given”; and: 4.1272: “Thus the variable name ‘x’ is the proper sign for the pseudo-concept object”; and: 4.121: “What finds its reflection in language, language cannot represent... Propositions show the logical form of reality”; and: 4.1212: “What can be shown, cannot be said.” The syntactical categories do have an objective reality that is reflected in language. So, while the categories are implicitly defined by the rules of syntax, it is also true that they represent something that is objectively there in the reality that propositions picture. Carnap picks up on the contextualism and ignores

205 Some cases one could not re-formulate. This is true of the Aristotelian notion of nature or essence. Here is something that some philosophers, from Aristotle to Gadamer, have claimed to exist. They also claim to be acquainted with these natures or essence or the Being from which they are supposed by some to arise. But they have not been successful in convincing people how to have this experience. In fact, it should be a very common experience. Natures or essences are intended by those who defend them to provide necessary connections among events. Event A occurs, followed by event B. So far as sense is concerned, these events, while not separate, are separable, logically and ontologically: each is what it is and not another thing, either wholly or partially, and does not in itself implicate any other event. In particular, our A does not in itself in any way imply B. That is, this is what there is so far as sense experience is concerned. A causes B, but so far as sense is concerned, it is one event folthe representative function. Carnap’s tendency to ignore reference in favor of implicit definition appears already in the Aufbau, where, for example he not only has observation terms designating or representing what is given in experience but also proposes to define them contextually, that is, give their meaning through axioms that “implicitly define” them. Thus, Carnap suggests that the visual field is the unique sense modality having exactly five dimensions; we then pick out colours as a three-dimensional subspace. (Carnap, The Logical Structure of the World, § 88-91). M. Friedman, Reconsidering Logical Positivism, Cambridge: Cambridge University Press, (1999), p. 99, notes the contextualist or holist element in Carnap’s thought, but seems to welcome it as a shift away from the empiricist emphasis on reference—which it is—but does not note the fact that Frege had already raised serious objections to it. We take up these points below; see the passages from Frege cited by notes 121 and 122. Oddly, Friedman does not mention the Fregean objections to the whole notion of “implicit definition.” Perhaps we are by now less surprised that neither does he mention Bergmann as objecting from a positivist perspective to this drift in Carnap’s thought from empiricism to holism. Friedman also misses the important essay by L. Pinsky, “Positivism and Realism,” Mind, N.S., 63, (1954), pp. 495-503. Again, oddly, for a work that aims to give something of the history of the Vienna Circle, Friedman does not discuss Bergmann’s memoir of the Vienna Circle, “Memories of the Vienna Circle: Letter to Otto Neurath,” first published in 1993 (see note 4). Neither is this memoir mentioned in Friedman’s discussion of the history of the origins of the division between philosophy done in the style of the Vienna Circle (“analytic philosophy”) and philosophy after the style of Heidegger (“continental philosophy”), A Parting of the Ways: Carnap, Cassirer, and Heidegger, Chicago: Open Court, (2000).

206 lowing another. As Hume said, following Sextus, all there is to causality is regularity. But according to the Aristotelians, there is more: there is a necessary connection effected by the nature or essence of the substance that is changing from A to B. Since causation is everywhere, so should there be these necessary connections, and since we know that one event, say A, causes another, B, it should be part of our common experience of the world that we experience one way or another that necessary connection. But, try though they have, the Aristotelians have not been able to exhibit this experience. They have tried. For example, they have tried to find the source of our idea of necessary connection in volition. Locke suggested this, Reid took it up, but Hume had already disposed of it: there is no necessary connection to be found here any more than anywhere else in the world. So necessary connections, and the natures or essences that are supposed to effect them, find no place in the empiricist language L, nor, therefore, in Bergmann’s L.45 45

The framework accepted by the early positivists and by Bergmann throughout his career was rejected by Carnap in 1936, when he published his essay on “Testability and Meaning.” Carnap thought he had a problem with disposition concepts. If one defined ‘soluble’ as (a) x is S = Df x is in water e x dissolves then it would seem (given the usual logician’s meaning for the horseshoe, i.e., material implication) then a piece of iron i which is never put in water will be soluble, since i is S will be true since the conditional i is in water e i dissolves is true by virtue of a false antecedent. But iron is not soluble, so (a) cannot provide the correct translation of ‘soluble’ into L. Or so the argument goes. Carnap proposed to meet this objection by “introducing” the disposition predicate ‘S’ into L by means of what he called a “reduction sentence”: (r) (x)[ x is in water e (x is S = x dissolves)] Now, if x is not in water nothing follows, whereas if x is in water and dissolves then x is soluble while if it does not dissolve then it is not soluble. The paradoxical case is eliminated. However, the paradox is eliminated only at the cost of abandoning the positivist principles. The elimination of the paradox thus comes at considerable expense— namely, at the expense of the positivist Principle of Acquaintance and the possibility of a phenomenological method, which is to say, at the expense of what amounts to the central core of the positivist position, at least as Bergmann saw positivism. The point is that (r) does not provide for the elimination of ‘S’ = ‘soluble’ from

207 L. It is a new primitive term. This is on the one hand. On the other hand, it receives no designation rule, no rule of reference tying it to the world as we ordinarily experience it. So, solubility is now construed as a primitive term, which, if we take it to represent something in the world, represents an unanalyzable power or disposition the presence of which effects a change in the world as we experience it (if x is in water then x will dissolve just in case that x is S) but which is not an entity of that world. Introducing terms by reduction sentences amounts to re-instating in the world the unanalyzable powers of the Aristotelians. Carnap, the formalist, was in effect providing a formal language, a syntax, for the metaphysics of essences. The author of the Logical Syntax of Language had found then means by which a metaphysics of Aristotelian essences could be legitimated. Introducing terms by reduction sentences amounted to providing a way for the Aristotelians to claim legitimacy for the unanalyzable powers that it was the point of the Principle of Acquaintance to exorcise. Recall Bergmann’s informal characterization of a positivist (“Logical Positivism,” p. 2), where one of the marks of a positivist is the acceptance of a Humean or regularity account of causality. In his 1936 essay, Carnap is rejecting this as an adequate account of causality—Aristotelian real connections, given by reduction sentences, are also needed. In other words, in 1936 Carnap gave up the positivist program. This was the upshot of his formalism, which argued that the cognitive meaning of a term is given by its syntactical connections to other terms. Unlike (a), or at least the universal conditional (a') (x) [ x is S = (x is in water e x dissolves) which (a) licences, which is analytic, (r) has the form of a synthetic statement. But Carnap lays it down that the role of (r) is to determine the meaning of ‘S’. It is therefore also a priori. In effect, then, a reduction sentence is a synthetic a priori statement analogous to an Aristotelian real definition. Carnap was later to generalize this idea of synthetic a priori statements in the language L he proposed. These sentences were not tautologies (like “p or not p”) nor definitional truths (like (a')), but somehow necessary. These statements he referred to as “meaning postulates;” see Carnap, “Meaning Postulates,” Philosophical Studies, 3, (1952), pp. 65-73. Reduction sentences were but a special case of meaning postulates. But what distinguished meaning postulates from other synthetic sentences seems to be nothing more than the fact that in the construction of L they occur on a page that Carnap has labelled “Meaning Postulates.” This is hardly a strong way to demarcate them as necessary truths. Quine, in his famous paper, “Two Dogmas of Empiricism,” Philosophical Review, 60, (1951), pp. 20-43, criticized what he took to be the empiricist’s notion of necessary truth. But he never attacked the principle that tautologies like “p or not p” are necessary, nor the principle that statements true by explicit nominal definition (like (a'), deriving from (a)) are necessary. That is, necessary truth as syntactical was not really called into question. Neither were the truths of mathematics like 3 + 4 = 7. But he did attack the doctrine of necessary truth as Carnap spelled it out: this was the real object of Quine’s critique. This was the idea that one can have necessary truth by mere convention, that is, the idea that by a mere decision, or even by community

208 agreement, for example, by an act of labelling as a “Meaning Postulate,” or by there being an agreed upon list of such “postulates,’ one could transform a statement into one that is both true and necessary—so transform it even if, like reduction sentences (r), it contains empirically meaningless terms and is therefore in a straightforward sense cognitively empty. It was a reasoned and successful critique: the doctrine of reduction sentences and the more general notion of meaning postulates hardly survived. Many took it that the syntactical notion of necessary truth was also demolished. Whether it was the object of Quine’s attack or not, that syntactical notion was successfully defended by Bergmann, in his essay on “Two Cornerstones of Empiricism,” in Metaphysics of Logical Positivism, pp. 78-105. For an extended discussion of the doctrine of reduction sentences and of meaning postulates, see F. Wilson, “The Notion of Logical Necessity in the Later Philosophy of Carnap,” in A. Hausman and F.Wilson, Carnap and Goodman: Two Formalists, The Hague: Martinus Nijhoff, (1967). Bergmann offered an alternative solution to the problem of disposition predicates, arguing that what was crucial was not the syntax but rather the psychological context in which they are used. See his “Comments on Prof. Hempel’s ‘Concept of Cognitive Significance’.” Bergmann’s point is elaborated upon in F. Wilson, “Dispositions: Defined or Reduced?” Australasian Journal of Philosophy, 47, (1969), pp. 184-204. The latter argues that a predicate defined by (a) counts as a disposition predicate provided it also appears as a concept in a statement of law, such as: (L) (x)( x is sugar e x is S ) It counts as a disposition when predicated of an object if that predication is justified by derivation from a law like this. Since there is no law that iron is soluble, S cannot be predicated as a disposition of an iron object such as i. In fact, given the laws, object i will have “insoluble” predicated of it as a disposition. The fact that the untested object has “S” predicated of it is harmless; what counts is predication justified by a law. The contexts provided by laws like (L) successfully serve to pick out those predications which are predications of dispositions. (See also H. Hochberg, “Dispositional Properties,” Philosophy of Science, 34, (1967), pp. 10-17.) As for what distinguishes those generalizations which are laws from those that are not, this, too, is a matter of psychological context. Generalizations are reckoned as causally necessary just in case they are subjectively treated as necessary. This is how one accounts for causal necessity in a world in which objectively there are no necessary connections, only regularities. For a discussion of this Humean appeal to psychological context as the origin of our “idea” of causal necessity, see F. Wilson, “Hume's Theory of Mental Activity,” in D. F. Norton, N. Capaldi, and W. Robison (eds.), McGill Hume Studies, San Diego: Austin Hill Press, (1979), pp. 101-120. If it is a matter of psychological context that distinguishes nomic statements and disposition concepts from others, then, as Bergmann argued, one cannot neatly capture these concepts in a formal language. The attempt to do so by “reduction sentences” would predictably fail. So will other attempts to formalize nomic necessity. Hume would have seen this. Carnap did not. He never gave up the dream that everything could be captured in a neatly formalized L. And so, late in his career he wrote that:

209 “Once the problem of the explication of nomic form has been solved and a logic of causal modalities has been constructed, it will be possible to use these modalities for explication of subjunctive and, in particular, of counter-factual conditionals. Presumably, it will then also be possible to introduce disposition terms by explicit definition;” R. Carnap, “Replies and Systematic Expositions,” in P. Schilpp, ed., The Philosophy of Rudolf Carnap, LaSalle: Open Court, (1963), pp. 859-1013, at p. 952. Carnap never gave up the idea—the vain idea—of formalizing everything. Every attempt to formalize causal modalities has turned on taking some modal casual concept as primitive but as uninterpreted into entities of the world with which we are acquainted. These unanalyzed modalities work no better than reduction sentences, and like the latter are but another way of injecting Aristotelian objective necessary connections back into the world from which they had been exorcised by the empiricists and the (early—pre-1936) positivists. In fact, we have to recognize the significance of “Testability and Meaning” for what it is: 1936 marks a radical rejection of the positivist program of eliminating such metaphysical entities as Aristotelian objective necessary connections and such metaphysical aspirations as those of Gadamer. Bergmann, almost alone among the members of the Vienna Circle, maintained the positivist standard. So Hochberg’s characterization of Carnap as a “positivist” in his The Positivist and the Ontologist, is doubly misleading: it suggests that Bergmann ceased to be a positivist, which is wrong, and it suggests that Carnap remained a positivist, which is also wrong, at least so far as the post-1936 Carnap is concerned, since he rejected both the Humean account of causality and the basic positivist appeal to a Principle of Acquaintance as a criterion for interpreting one’s ontologically perspicuous ideal language L. One might add that most discussions of Carnap ignore these points. See, for example, J. H. Fetzer, “Reduction Sentences ‘Meaning Postulates’,” in N. Rescher, ed., The Heritage of Logical Positivism, pp. 55-66. Fetzer simply ignores critiques of Carnap such that of Bergmann, and of those, such as Wilson and Hochberg, who have pursued Bergmann’s line of criticism. Such neglect to deal with the literature is neither good scholarship nor good philosophy. One further thing about Quine’s “Two Dogmas” essay. One “dogma” was that of “analyticity.” This we have just commented upon. The other also deserves notice here; we shall comment on it further below. This other “dogma” was that philosophy should begin with a Principle of Acquaintance. What Quine actually argued was that acquaintance could not by itself provide a foundation for empirical knowledge. Having disposed of that he took it that he had ruled out a Principle of Acquaintance for ontology. But this does not follow: one can successfully appeal to acquaintance in ontology even though that acquaintance is not incorrigibly veridical. Bergmann replied to Quine on this issue, in his “Two Cornerstones of Empiricism,” where he defended the use of a PA in ontology. He made it clear that acquaintance as providing a foundation for empirical knowledge was not the real issue. Quine infers that lacking an incorrigible foundation, knowledge must depend on coherence as its criterion of acceptability. This is a not unreasonable inference. (But

210 The denotation of the language of essences or natures as devoid of cognitive content was one of the main points of the early positivists. It was used in particular to effect the exclusion as meaningless all discourse about the role of final causes. There was little need to make the point with regard to physics: it had been made by Galileo, when he insisted that physicists seek regularities in nature and not the forces the Aristotelians supposed there to be behind them. But biology still had its defenders of teleology and finality—think of Dreisch. The positivists argued that such discourse lacked cognitive content—it could not be cashed out in terms of the world of ordinary experience, where phenomenological description can locate no finality. The positivists in effect worked to cleanse biology of the dross of Aristotelianism. They made the same point with regard to the human sciences. Explanation in the Geisteswissenschaften had exactly the same form as explanation in the Naturwissenschaften: there was no more role for teleology in psychology than there is in astronomy. Bergmann’s philosophizing began with this point—this positivist point—and he never abandoned it. There are two points that need to be made. First, you don’t need a precisely formulated thesis about the Verifiability Principle to argue that the natures and essences that are supposed to ground the final causes lack cognitive content. To attack the Principle on grounds that it never was precisely formulated and to argue that positivism failed for that reason is to miss the real issue, and ends up a way of excusing those who defend na-

still not correct; see F. Wilson, The External World and Our Knowledge of It: Hume’s Critical Realism: An Exposition and A Defence, Toronto: University of Toronto Press, forthcoming). Quine however further infers that this implies a holistic account of meaning, that is, in effect that it implies a denial of the logical independence and ontological separability of the entities in one’s ontology. But this latter inference, from coherence as a criterion for epistemological acceptability of propositions, that is, from a holistic account of acceptability, to a holistic account of meaning and a denial of logical atomism, is illegitimate. We shall be looking at the validity of Quine’s inference shortly, together with the views of such thinkers as W. Sellars who make similar inferences. We shall argue that the inference is invalid, and shall defend the ontological independence of entities in a positivistic metaphysics of the world as we experience it. But the basic case can be found in Wilson’s 1970 essay, “Acquaintance, Ontology and Knowledge,” The New Scholasticism, 54 (1970), pp. 1-48.

211 tures and essences and teleology and finality.46 And second, to repeat, Bergmann never departed from this sort of positivism. It is within this framework that Bergmann develops his ideal language L. His purpose is, or at least became, ontological.47 46

This is a common line of thought. Thus, Ayer, who worried about the form the Principle of Verifiability might take, is deemed a failure because he never got it quite correct. But he didn’t need that precision for his critique of religion. However, since he failed in his search for a precise formulation, his critique of religion is assumed also to have failed. That is far from being so, but it is what one hears from the defenders of theology. And so a book devoted to evaluating Ayer as a philosopher, Fact, Science and Morality: Essays on A. J. Ayer's Language, Truth, and Logic, ed. Graham Macdonald and Crispin Wright, Oxford: Blackwell, (1987), devotes considerable space to the attempts to make the Verifiability Principle precise, but ignores the critique of religion. It is the latter that has real social significance, but contributors to the volume prefer the scholasticism that creates the debates about the niceties of how to make precise Ayer’s Principle. See F. Wilson, Review of Fact, Science and Morality: Essays on A. J. Ayer's Language, Truth and Logic, Teaching Philosophy, 11, (1988), pp. 179-81. 47 Actually, little things—things like atoms or quarks—gave some positivists some trouble. If we begin with acquaintance how can we have terms that refer to unobservables? So they distinguished theoretical terms from observation terms, and created for themselves the problem of how to give meaning to the former in a way that could be fit into the framework of a language L interpreted in conformity with a PA. There was a minor cottage industry that developed about the distinction and the “problem” of the meaning of the “theoretical vocabulary.” The distinction seemed to create more problems than it solved. The first attempt to bring in terms on each side of the distinction was with Carnap’s attempt to “introduce” disposition predicates by way of “reduction sentences” in his 1936 essay on “Testability and Meaning.” Disposition terms are undefined with respect to the observation vocabulary and are “theoretical” because they are “introduced” through their occurring in synthetic statements of fact deemed “necessary” by the one constructing the language. Later attempts along the same lines were made to “introduce” various terms other than disposition terms, e.g., the terms of atomic theory, by allowing other forms for the introducing sentences besides that of reduction sentences. These attempts all had the same philosophical problems as that of the simple reduction sentences, but usually additional problems of their own. The problems with the distinction were never really resolved; counterexamples were always to be found. Bergmann’s advice was that one not make the move that Carnap made in 1936; this advice, that one not allow undefined terms into L unless they met the requirements of PA, turned out to be wise. In fact, the same problem the “observation term / theoretical term” distinction attempted to address, the problem of unobservable entities such as atoms, arose al-

212 ready for physical objects: all is fine so long as the objects are observed, but as soon as they slip from sight how can the terms of one’s language any longer refer to them, and if meaning is given by rules of reference or designation then how could such language be construed as meaningful? Some took up this argument and inferred, not unreasonably, that, if the verifiability principle had such an obviously unacceptable conclusion, then that principle had to be rejected. Russell provided the obvious way out: the terms referring to unobservables, say those in the past, were not really terms but disguised definite descriptions—where sentences involving definite descriptions, while they have a grammatical form of subjectpredicate sentences, are to be construed as having a different logical form, specifically a complex quantificational form. Thus, consider: The F that is R to a where the definite description refers, or rather, seems to refer to an individual that is F and stands in the temporal relation R ( say, sometime before) to a, where ‘F’, ‘a’ and ‘R’ are all observation terms, that is, terms introduced into one’s L in conformity with the Principle of Acquaintance. This sentence seems to have a term “the F” which refers to an unobservable, and therefore the positivists should grant that the sentence cannot be verified n experience and is therefore cognitively meaningless. What Russell showed is that one can construe this sentence as having the logical form: There is at least one thing x that is F and at most one thing x that is F and is R to a This contains no terms that violate PA, yet, given its quantificational complexity, succeeds in being true (or false) if and only if an unobserved individual existing R before a really is (or is not) F. The logic of quantifiers succeeds in creating a reference, but only a reference of sorts, where there is no term that refers, and therefore no terms for which the issue of conforming or not conforming to PA arises. The logic of quantifiers enables one to refer (or rather “refer”) to things that are unobserved and to do so in away that is compatible with the positivist’s PA. This move obviates the need to distinguish two sets of descriptive constants, a “theoretical” vocabulary from an “observation” vocabulary. The only descriptive terms are those that are interpreted into the entities of the world of ordinary experience. That is, the only descriptive terms are those of the observation vocabulary. The remaining descriptive terms do not form a separate vocabulary; in fact they are not really descriptive terms, but rather definite descriptions for which the sentences that contain them are unpacked by means of complex quantificational expressions. Russell made the essential point. Bergmann was to expand on it sometime later in his “Remarks on Realism,” where he showed how one could be a realist about unobserved objects while yet adopting what he then called a “phenomenalistic” language, by which it is clear he meant no more than that the language had to be interpreted in conformity with PA. For further aspects of these issues, see F. Wilson, “Empiricism: Principles and Problems,” in W. Sweet, ed., Approaches to Metaphysics, Dordrecht: Kluwer, (2004), pp. 265-302. One can also use the apparatus of definite descriptions to refer to kinds of things that are unobserved, that is, to “refer to” such unobservable properties as “The sound

213 I propose to look at this ontology in part. I shall look at what some have found strange indeed, his ontologization of logic. After all, don’t we all know that logic is linguistic? I shall also look at two entities that are considered to be among the strange things Bergmann has introduced. One is the bare particular. The other is the even stranger, that property of the bare particular that he calls particularity. The latter enters as part of Bergmann’s ontologization of logic. In fact Bergmann, too, seems to find particularity strange. He struggles with bare particulars, but it is worse with particularity. the dog heard had a pitch that was too high for me to hear.” For such uses of the quantifiers, higher-order logic is required. This use of the logic of quantifiers is to be found in laws and theories that are laws about laws, that is, research guiding theories that include Principles of Determinism and of Limited Variety. The mixed quantificational and higher-order uses of logic enable one to create terms that “refer” to entities of kinds one hopes to discover but where one has not yet been successful. The term ‘the flu bug’ “refers” to an unobserved cause of a disease and it is up to the researcher to find the bug that is the flu bug, that is, the germ or virus that is such that the species of which, when it is present, causes the presence of the disease we call the “flu.” See F. Wilson, The Logic and Methodology of Science and Pseudoscience. For the relevance of these points to the logic of scientific explanation, those points about definite descriptions in particular, see F. Wilson, Explanation, Causation and Deduction, Dordrecht: D. Reidel, (1985). For details about how to refer to unobservable things like atoms and how to develop atomic theory in a way compatible with PA, see F. Wilson, “Empiricism and the Epistemology of Instruments,” The Monist, 78, (1995), pp. 207-229. It was actually Russell who showed the way; see his Analysis of Matter, London: Allen and Unwin, (1927). It is worth noting that if one defines the quantifiers as is some times done, that is, by construing the universally quantifier as an indefinitely long conjunction and an existential quantifier as an indefinitely long disjunction, then all the terms will satisfy PA but one will have no quantificational apparatus that can so to speak refer to unobserved entities. One will find oneself restricted in one’s ontology to the things that are near to one in space and time and given in immediate sense experience. Wittgenstein of the Tractatus reduced his quantifiers to long conjunctions and disjunctions. This commits him to the view that what exists is what is given in narrow sensible experience; there is nothing beyond the immediately given—nothing at least about which one can meaningfully speak: recall the remarks that “The world is my world: this is manifest in the fact that the limits of language (the only language which alone I understand) mean the limits of my world” (Tractatus, 5.62), and “I am my world...” (Tractatus, 5.63). In this sense Wittgenstein is a phenomenalist—a Berkeleyan phenomenalist but one without God—where Bergmann (recall his “Remarks on Realism”) is a realist (about unobserved objects and kinds of objects)—which is one of the several senses in which Bergmann is a realist.

214 The struggle is there by virtue of the method that he proposes for philosophy. Specifically, of course, Bergmann proposes a phenomenological method in ontology, one based on a Principle of Acquaintance (PA). This is essentially a positivist method.48 Given this method, Bergmann holds, as he therefore should, that both particulars and particularity are presented, that he is acquainted with them. At the same time he also holds that particularity has the further odd characteristic of being ineffable, and that its presence in a particular is an unsayable state of affairs, and further that it is something that is not a thing but yet is not nothing. In what follows we shall first discuss Bergmann’s philosophical method and gradually move to the cases of particulars and of particularity. In doing this we shall try to show that, Bergmann’s qualifications notwithstanding, things about logical form, and particularity in particular, really can be said. It was Kant’s suggestion, and before him Locke’s, that before undertaking research in ontology we ought to investigate our tools, to see what sort of job they can do, what sort of world they can reveal to us. That is, or so goes the suggestion, epistemology precedes ontology. Now, the maxim that one should check one’s tools before one starts work may be sound advice for a carpenter, but I cannot think of worse advice for a philosopher. Another maxim, equally appropriate for the carpenter and that might equally be recommended for the philosopher, might well be this: investigate carefully the nature of the material before you try to design a tool to work upon it; that is, for the philosopher, ontology must precede epistemology. But even this maxim is misleading in the case of ontology. For, to do ontology we must know what we are talking about. Epistemology thus seems to be both presupposed by and also justified by ontology. But that apparent circle should not surprise us: it is but one way of noting the delicacy of the metaphysical-ontological-epistemological enterprise. One point, however, is clear, and that is the starting point of this enterprise: the starting point of ontology is the world as we experience it, and the starting point of epistemology is our experiencing of the world, and, in fact, these 48

It is essentially the method that guides Carnap in his Aufbau. But it is also the method that clearly guides Russell in his Our Knowledge of the External World. It has its obvious forerunners in Mach, Mill, and Hume.

215 two starting points cannot be separated. Our experiencings of the world from which epistemology takes its start are themselves part of our experience, and part of the world as we experience it, and therefore part of the world that is the starting point of ontology. As for our ways of experiencing: they are manifold. Among the ways of experiencing the world, that we encounter in our experience, are our loves and hates, our perceivings and our rememberings, our sensings and our thinkings, our believings and our disbelievings, our expectations and our hallucinations, our scientifically methodical researches and our occult fantasizings. Bergmann is clear: the justified procedure for extending our knowledge of the world is to start from perception and memory and to explore the world by the methods of science: that is the pathway to truth. But that judgment—that perception and science are the measure of what is—presupposes that certain basic moves in ontology have already been made.49 It presupposes that the world that we experience has a certain ontology and that our various ways of experiencing it have been examined, located in their place in the world as we experience it, and their capacity to testify to the truth subjected to criticism and evaluated. All of which means that the starting point of all ontology can be none other than the phenomenological description of the world as we experience it and of our experiencings of that world. As Bergmann once put it: “Ontology is phenomenological. The rest is merely science. Such is the nature and such are the limits of human knowledge.”50 To suggest otherwise—as Jay Rosenberg has done—that science begins with epistemology, and that epistemology is exhausted by the methods of science,51 is simply to avoid all the serious work of philosophy. It is, of course, true that among the things that we find in experience, in the world as we experience it, are hallucinations, or, more generally, falsehoods. In fact, perhaps almost all our ordinary perceptions are false, as Locke, Hume, and other critical realists like Roy Wood Sellars, have, on the basis of the inferences of science, so cogently argued. That, however, 49

As Bergmann used to put it, science is simply the long arm of commonsense. Bergmann, “Realistic Postscript,” in Logic and Reality, Madison: University of Wisconsin Press, (1964), p. 339. 51 Rosenberg, “The Given and How to Take It,” Metaphilosophy, 6, (1975), pp. 303337, at p. 316. Rosenberg’s comment reflects a similar position adopted by his teacher Wilfrid Sellars. 50

216 is something we find out about the world we experience: recognizing that fact is one way of experiencing the world. Moreover, that atoms or whatnot that are there where our perceptions are false we know to be there because of what we discover through experience, and are parts—though perhaps unperceived parts—of the world we experience. For, our experience shows us that there are, in the world we experience, many parts that we do not perceive. Why should atoms not be among the unperceived parts of the world of our experience? But be all that as it may, the point is that, to say that ontology is phenomenological is not to deny the fact of error: error and its correction are processes that must be described by any phenomenological description of experience, that is, of the world that we experience and of our experiencings of that world. A full pheomenological description of the world that we experience cannot but include a description of falsehood and errors. This being so, it means that what is given to us in experience—what is presented to us in experience—includes what we might, later, or perhaps even at the moment, decide is not the case. It follows that what is given to us, what is presented to us, is not given or presented incorrigibly. It is a myth created by foundationalists that what is given is always given incorrigibly. Or perhaps better, it is a myth that was created by those who are concerned to refute something they call the “myth of the given.” There is, of course, one sort of experience, often referred to as sensing, or, by some, who wish to include inner awareness, as direct awareness, which is, in some sense, error-free, or, at least, is such that it makes no real sense to speak of correcting it. The entities that are given in the kind of experience called sensing and inner awareness are often in turn referred to as phenomena, or as phenomenal entities. There is a sense, then, in which phenomena might be said to be given incorrigibly. But phenomena are only some among the many sorts of entity that are presented, and sensing and inner awareness are only two of the many ways of experiencing the world. Moreover, phenomenology or the description of the world that we experience is not, or is not restricted to, the description of phenomena. Thus, while, on the one hand, when we offer a phenomenal description of something, what we describe must, in some sense of ‘must’, exist, on the other hand, when we offer a phenomenological description of something, it simply does not follow that what is thus described exists.

217 In addressing ourselves to the philosophy of Gustav Bergmann, it is important to be clear on this point. When one criticizes a philosopher, it is best to be clear on the method that that philosopher proposes to use in his or her attempts to solve those problems and perplexities that are typically philosophical. Rosenberg is one who has failed to follow this maxim. Bergmann proposes that the core of ontology—and of epistemology—is phenomenology.52 Ontology begins with what is presented to us, or, as he also says, synonymously, with what we are acquainted with.53 What we are acquainted with we know54—in the sense that we know what we are presented with, what is given to us: we can describe it phenomenologically. But, of course, we do not thereby know it in the sense that in knowing it we know incorrigibly that it exists. Acquaintance in this sense is not incorrigible, and it is not what some foundationalists at least have called “direct awareness,” where the latter is so direct that it is incorrigible. To be sure, Bergmann does speak of “direct awareness,” which is a way of experiencing that includes at least some of those sorts of experiencings that others have called “sensings.” “Direct awareness” is therefore in a way incorrigible: in some sense of ‘must,’ what is given in direct awareness must exist. But contrary to what Rosenberg suggests,55 when Bergmann speaks of ontology as phenomenological, or as beginning with what one is acquainted with, he is not speaking of direct acquaintance in the foundationalist’s sense nor of some incorrigible starting point. Failure to appreciate this point means that Rosenberg almost everywhere misunderstands Bergmann, and throughout his essay it vitiates most of his critique of the latter. To say that ontology begins with phenomenological description is not to say that the philosopher does not argue. Description does not exclude dialectics. Thus, Bergmann regularly argues that certain entities that are picked out in his phenomenological descriptions solve certain philosophical problems, and that other philosophers cannot solve these problems without invoking these entities. Such arguments must carefully be distin52

Bergmann, “Realistic Postscript,” p. 339, also p. 307; “Ontological Alternatives,” in his Logic and Reality, Madison: University of Wisconsin Press, (1964), p. 127; and “Strawson’s Ontology,” in ibid., p. 185. 53 Bergmann, “Ineffability, Ontology, and Method,” in Logic and Reality, p. 45. 54 Ibid. 55 Rosenberg, “The Given and How to Take It,” pp. 305-306.

218 guished from transcendental arguments. In the latter sort of argument, one argues, as Kant and the Kantians did, and as Plato and Aristotle and the Aristotelians all did long before, that certain things that exist within experience could not exist unless certain entities outside the world of experience also existed. One could add that at least these philosophers argued, they did not merely assert, as Gadamer does: they were not contemptuous of thought as Gadamer appears to be. But to give an example, Plato argued that, on the one hand, the things within the world of sense experience are all particular—he accepted a Principle of Localizaton for sensible entities—and argued that, on the other hand, these entities may often correctly be said to be the same. All entities in the world of sense are particular, so no entities within the world could account for those facts of sameness, and, therefore, there must be entities outside the world of experience that account for the sameness of experienced things. These objects are the Forms. These are not simply unexperienced parts of the world we experience, likes specks of dust too small to see, or atoms, which are even smaller; rather, the Forms transcend the world of ordinary experience. But then, if we are to identify things as the same, then we must have some knowledge of these Forms. This provides the basis for a further transcendental argument on Plato’s part, an argument to the effect that there must exist a special sort or form of knowing, which Plato calls “reminiscence,” that gives us knowledge of these Forms. Needless to say, this sort of knowing is not something that is described when one gives a phenomenological description of the world of experience, and, specifically, of the various modes of experiencing that are given to us in the world of experience. Aristotle offered similar arguments for the existence of substances. In contrast, Gadamer simply assumes, dogmatically, and without argument, that there are such transcendent entities and such extra-ordinary ways of knowing them. In contrast to Plato (and Gadamer), Bergmann systematically picks out certain entities in the world of experience, and then argues that these entities solve the problem of sameness that Plato thought could be solved only by introducing transcendental entities. Bergmann argues concerning entities in the world of experience; Plato argues for the existence of entities that transcend the world of experience. In fact, Bergmann proposes a method that conforms to a Principle of Acquaintance: not only is phenomenology the starting point of ontology-epistemology, it is also the end-

219 ing point, in the sense that no entity is introduced into one’s ontology unless it is given in experience,56 that is, unless it is picked out in one’s phenomenological description of the world of experience—ordinary experiencing, which is to say, either by sensory awareness of the world or by inner awareness. Rosenberg has wrongly concluded that, since Bergmann introduces arguments for his ontology, therefore he has abandoned phenomenology and the Principle of Acquaintance.57 This is connected to his other error, previously noted, that phenomenologically given entities are given incorrigibly. Phenomenological description is of phenomena; phenomena are known incorrigibly; what is incorrigible is known noninferentially; only the corrigible requires inference; but argument is inference; so, Bergmann’s arguing testifies that the knowledge he claims to have is corrigible and therefore not phenomenological. At least, that is what Rosenberg seems to think. One should perhaps add, however, that philosophers are human too, and are quite capable of convincing themselves that they are not acquainted with what is in fact presented to them. In this sense, dialectics may contaminate one’s phenomenology. Is there any guarantee that anyone’s phenomenology is not contaminated? The answer is No: there is no guarantee. The best we can do is be ever vigilant. As Allaire once put it, “One cannot do more, one must not do less.” 58 This does spot, however, an important role that dialectical arguments can play in the philosophical enterprise. Philosopher B might well succeed in directing the attention of S to a certain entity, but S’s own philosophical reflections convince him that these entities are philosophically irrelevant. Or perhaps S’s reflections may lead him to a different philosophical description of those presented entities. Thus, for example, some would, for their own reasons, refer to as “places” what others call “particulars.” In such circumstances, B’s dialectical arguments have the role of aiming to convince S that the entities to which B is attending have a certain philosophical relevance. Moreover, it is even possible that S’s reflections have had the result of convincing him that the entities to which B is attempting to direct his attention do not exist; in that case S’s reflections will have 56

Bergmann, “Acts,” in his Logic and Reality, p. 14. Rosenberg, “The Given and How to Take It,” p. 308. 58 E. B. Allaire, “Bare Particulars,” in Allaire et al., Essays in Ontology, pp. 14-21. 57

220 convinced him that he is not presented with what is in fact presented to him. In this latter sort of case, B’s dialectical arguments will have the further aim of challenging those arguments that have led S to deny the phenomenological claim of B. In this way, B’s argument does aim to convince S to, as it were, introduce a “sort” of entity into his world, and not just his ontology but the world as he phenomenologically describes it. But again, such dialectical argument is not a form of transcendental argument. It is not an attempt to introduce into his or her ontology an entity that is not presented but rather is an attempt to get one to recognize an entity that is presented.59 None of this is puzzling, but—again—clarity in these matters is important if we are to understand correctly one who insists upon a phenomenological method in philosophy. Thus, while ontology begins with the phenomenological, and in a sense ends there, finding its categories in the world as it is presented to us, these categories cannot simply be “read off” as it were from what is given: one concludes that aspects of the given constitute the categories of one’s ontology only after the arguments of one’s dialectical defence of that ontology. Everett Hall once misunderstood this in commenting on Bergmann’s ideal-language method. He suggested that it was Bergmann’s view that “experience comes already categorized, and the Ideal Language philosopher [i.e., Bergmann] needs only to square in its structure with this phenomenologically given structure.”60 He rejects Bergmann’s method with the remark that “experience seems to show the handiwork of Professor Bergmann’s categories.”61 Hall is commenting on an essay by Bergmann that is commenting in turn on Irving Copi’s views on a philosopher’s use of an ideal language.62 Hall rightly joins Bergmann in rejecting Copi’s view of an ideal language as a kind of Esperanto, only better, that we approach in ever more accurate form, converging in a quasi-mathematical fashion on the best. But Hall misunderstands Bergmann. Bergmann makes the point that according to his phenomenological method based on 59

See G. Bergmann, Realism, Madison: University of Wisconsin Press, (1967), p. 68. Everett Hall, What Is Value? An Essay in Philosophical Analysis, London: Routledge and Kegan Paul, (1952), p. 197n. 61 Ibid. 62 I. Copilowish (later Copi), “Language Analysis and Metaphysical Inquiry,” Philosophy of Science, 16, (1949), pp. 65-70. 60

221 PA, “...language can be known to be ideal [to be ontologically perspicuous] by comparing its logical structure with the ontological structure of the world, which must be known independently if the comparison is to be significant.”63 Bergmann does propose that “the ideal language...[is] complete only in the sense that it must show, in principle, the structure and systematic arrangement of all the major areas of our experience.”64 This is not to say, as Hall suggests it is to say, that one obtains one’s ontological categories simply by looking; this is simplistic and Hall rightly scorns such a simplistic method. But he is wrong about Bergmann: Bergmann’s method is not in this way merely simplistic. To be sure, it is Bergmann’s contention that the categories (such categories as particularity) are there, objectively there, to be found in experience; that is what it means to say that the method is phenomenological and is based on PA. But to stop here, as Hall does, is to miss an essential feature of Bergmann’s approach to ontology. It misses the dialectics: one does not find there, in experience, that just these features of the world as we experience it and not those other features solve the philosophical problems that derive from the tradition. One does not discover that certain features solve certain philosophical problems, as Hall suggests, simply by looking. One must argue, dialectically, that these features of the given, and not those, solve the problems. These dialectical moves are not given to one by phenomenological description. Hall misses this aspect of Bergmann’s method. One point that Bergmann has made and has emphasized is that “All awareness is propositional.” This thesis is one that Bergmann justifies by an appeal to the phenomenology of the world as we experience it. Bergmann has made this phenomenological claim a central pillar of his ontology and epistemology.65 It is a claim that is undoubtedly correct. In thought, in perception, in sensing, what is given to us are complexes. In other words, in order to give a phenomenological description of the whole of what is given to us in a perceiving, or in a thinking, or in any sort of experiencing, one must use a sentence. If we say that the whole of what is 63

G. Bergmann, “Two Criteria for an Ideal Language,” Philosophy of Science, 16 (1949), p. 70. 64 Ibid., p. 73. 65 Bergmann, “Intentionality,” in his Meaning and Existence, Madison: University of Wisconsin Press, (1960), p. 7.

222 given to us in an experiencing is the intention of the experiencing, then the intention of an experiencing is always propositional, represented by a sentence, and not by a name standing alone. But, of course, to say that—to say that the whole of what is given to us in an experiencing is a complex— is to say that the parts of such complexes are also given. Otherwise they would not be complexes. So, we are presented with complexes and also with the parts that are in those complexes. As Bergmann once put it, “To be presented (to me) is the same as to be the intention or ‘in’ the intention, of an act [of experiencing] (of mine).”66 This of course requires that we be able to distinguish the parts of the complexes that are given in experience. We must be able to attend to the parts—not, naturally, as separate entities (though they may be separable), but at least as parts. Otherwise what grounds would we have for including them when we give a phenomenological description of the complex that is the intention of the experiencing? All this is obvious enough. And it is equally obvious that, for one like Bergmann, who proposes to philosophize in accordance with a Principle of Acquaintance, the experiencings through which we distinguish the parts of the presented complexes must themselves be ordinary sorts of experiencings—experiences that are themselves experiences and not forms of experience, like Plato’s “reminiscence” of Forms, that are introduced into one’s epistemology solely on the basis of transcendental arguments. Moreover, if the complex I am presented with is a red square, then I can clearly distinguish the color and the shape. That makes it obvious that the experiencings in which we distinguish the parts of presented complexes are ordinary and are themselves experienced. So, when Bergmann insists that the parts of presented complexes are also presented, he is not thereby introducing a new sense of ‘presented’ or ‘acquaintance,’ contrary to what Rosenberg suggests,67 and certainly not any kind of acquaintance that is incorrigible in the foundationalist sense. Now consider a red square and a red circle. (Assume that they are the same shade of red.) In each of these two complexes we can distinguish a color. The color in each is what we represent in our phenomenological description by the word ‘red.’ The word ‘red’ is applied to each of the two 66 67

Bergmann, “Realistic Postscript” in Logic and Reality, p. 307. Rosenberg, “The Given and How to Take It,” p, 308.

223 complexes because the color in the one is indistinguishable from the color in the other.68 It is this indisdinguishability that accounts for ‘red’ applying to both complexes. That is, it is this fact that accounts for the sameness of the two complexes. Assuming that the distinct are distinguishable, it follows that the color of the one is the same as, not distinct from, the other complex. The color red is distinguishable from the shapes. It is also, of course, distinguishable from other shades of color. It is thus distinct from all these other entities. But the red in the one complex is not distinguishable from, it is the same as, the red in the other complex. A distinguishable entity that solves the problem of sameness by virtue of the fact that, as the same entity, it can recur in several complexes, is traditionally called a universal. Similar arguments will of course show that the other colors, the various shapes, and so on, are all universals. And we are therefore acquainted with universals: they are presented parts of presented complexes. Note how this approach locates Bergmann the ontologist as a philosopher who is squarely in the positivist tradition in which he first learned to philosophize. The problem of universals—how can words be general in their meaning?—is solved by such simple and unproblematic facts as that we can have two shirts that are the same color. The point can be made by reference to the technique of an ideal language L that was employed by the positivists and that Bergmann continued to the end of his career to use. Bergmann has an L with predicates like ‘Re’ (= ‘Red’). This predicate names or refers to or denotes red, the property red. In other words, it is interpreted into that presented feature of things.69 It applies, let us suppose, to this red shirt (which is called ‘a’) and that one (which is called ‘b’); that is, ‘Re(a)’ and ‘Re(b)’ are both true—they both correctly describe the entities presented to us. What ‘Re’ refers to is a presented feature of a and of b. This idea of interpretation into entities with which we are acquainted is taken as unproblematic. Indeed, it is taken a unproblematic not only by Bergmann but by other positivists such as Carnap.70 What Bergmann does 68

See F. Wilson, “The Role of a Principle of Acquaintance in Ontology.” Cf. Bergmann’s “Logical Positivism, Language, and the Reconstruction of Metaphysics,” in his Metaphysics of Logical Positivism, pp. 60-61, p. 63. 70 Cf. C. G. Hempel’s discussion of the idea of an “empiricist language” in his “Problems and Changes in the Empiricist Criterion of Meaning,” Revue Internationale de Philosophie, 2, (1952-53), pp. 52-53. 69

224 is take this unproblematic idea and use it to explicate the traditional claim that there are universals. That is, Bergmann argues, dialectically, that by adopting this ideal language he can solve/dissolve the traditional problem (and others, e.g., analyticity) where others cannot.71 The language and its interpretation are unproblematic.72 What are problematic are the claims of the other fellow; it is he or she who has the problems, and they cease to be problems once they are explicated. In this sense, it is always the other chap who has problems, not Bergmann. Bergmann’s task, as he proposes it, is to take those problems that the philosophers of the past have created and bequeathed to their successors, and so explicate them that they admit of commonsensical solutions. Insofar as Bergmann proposes solutions to traditional problems, they are not problematic as they were in the tradition as Bergmann has received it. Once again, this is a positivist stance, but note that it is also philosophical—and, indeed, metaphysical—recall the title of Bergmann’s first book, and one of his early papers (it proposed a “Positivistic Metaphysics of Consciousness”), where its appearance led to his being excommunicated by the orthodox positivists! It is philosophical insofar as, in contrast to what Carnap & Co. do, it takes the metaphysical tradition seriously, as presenting problems that are the core of the philosophical enterprise. In this sense, Bergmann is a positivist who is also a philosopher, rather than a positivist whose interest is the essentially non-philosophical area of Grundlagenforschung with respect to science.73 In a straightforward sense, then, the philosophical problem of universals is solved by pointing to the commonsense fact that for some shirts, two shirts do have the same color. At least, it is this fact that one argues is crucial to the solution. But insofar as this is the only fact to which one appeals, the specifically philosophical disappears for Bergmann. In this sense, for Bergmann there are, as the positivists said, no philosophical problems. However, Bergmann is not Carnap. If Bergmann is a positivist, 71

Cf. Bergmann, “Logical Positivism,” in his Metaphysics of Logical Positivism, p. 8; also his “Particularity and the New Nominalism,” in Meaning and Existence. 72 Cf. “Logical Positivism, Language, and the Reconstruction of Metaphysics” in Bergmann’s Metaphysics of Logical Positivism, pp. 30-77. 73 Cf. Bergmann, “Logical Positivism,” pp. 4-5; also his “Physics and Ontology,” in Logic and Reality, pp. 108-123.

225 then he also is or came to be a philosopher. To explain: Bergmann differs from Carnap in not dismissing as a non-question the question of which alternative solutions to a philosophical problem, e.g., that of universals, one ought to accept. For Carnap, what mattered was decision, not debate.74 As he would put it, the issue is practical, without cognitive content. Different solutions to traditional problems amount to different recommendations for an ideal language. The issue between them is to be resolved—dissolved rather than solved—by a simple choice of a convenient language. The issue is to be resolved by a decision, by stipulation and fiat, rather than by argument. For Bergmann, in contrast,75 the questions are to be settled by rational argument, by dialectics, even if he is also convinced—as his adoption of a phenomenological method or Principle of Acquaintance shows him to be convinced—that the solutions are to be found by appeal to what is unproblematic and uncontroversial. Hochberg has warned us against supposing that “...since one can buy two shirts of the same color, some form of Platonism must be adopted and all forms of nominalism must be rejected, on the basis of that fact alone.”76 This warning may well be just: Bergmann may well be wrong. Only practice of the method will reveal that; as Bergmann always said, the proof of any pudding is in the eating. Yet the warning is misleading, at least in its implied description of the phenomenological method. To adopt this method, and appeal to two shirts having the same color to solve the problem of universals, is not to appeal to that fact alone, even if that fact is the only fact to which one appeals. For, one must argue for the relevance of that fact to the traditional problems. That is, one must indicate that fact and also argue dialectically. Hochberg misses the latter.77 74

Cf. R. Carnap, “Empiricism, Semantics and Ontology,” Revue Internationale de Philosophie, 2, (1950). 75 Cf. his “Comments on Professor Hempel’s ‘The Concept of Cognitive Significance’,” in Metaphysics of Logical Positivism, p. 267. 76 H. Hochberg, “Russell’s Reduction of Mathematics to Logic,” in E. D. Klemke, ed., Essays on Bertrand Russell, Urbana: University of Illinois Press, (1970), p. 403. 77 As Bergmann put it in his discussion on “Logical Positivism,” the rejection of metaphysics is a negative step in the resolution of a traditional philosophical position, but one that is “partial” and “incomplete.” He goes on, “to complete it [the resolution], one must, also positively, identify the features of the ideal language that reflect those features of the ideal language that reflect those aspects of our experience one [the phi-

226 Insofar as Bergmann is, and Hochberg is not, prepared to appeal to such unproblematic facts in the solution of philosophical problems, the former remains closer than the latter to the positivism of the early Vienna Circle. But to return to the point, it is worth noticing that in discussion dealing with philosophical problems, one moves quickly from phenomenological description (as in “the red in the one complex is indistinguishable from the red in the other”) to dialectical argument (as in “the fact of this indistinguishability provides the solution to the problem of sameness”) to categorizations that are justified by those dialectical moves (as in “the red in the two complexes is a universal,” or “since we are presented with red, which is a universal, we are therefore presented with universals.”) The point is that such a transition does not imply that one has therefore somehow abandoned phenomenology for philosophy, where the latter is contrasted to phenomenology, perhaps identified with transcendental ontology: the entities described at the end of the dialectical argument are precisely those that are described in the phenomenological description with which one begins. Bergmann has, of course, made these points about properties and universals.78 The above discussion is misleading only in the brevity of the dialectic that it incorporates. For in fact, the dialectic can and must go on much longer than the above sketch suggests. Nor does Bergmann neglect that dialectic. He regularly was concerned to argue that those who deny that there are universals are mistaken. In particular, he was concerned to argue that the nominalism of one like Quine was mistaken. Let us say that the sentences that describe the complexes with which we are presented describe facts. Then, for Quine, facts are “ultimate and irreducible”: for Quine, the facts with which we are presented, in perception at least, are not really complex, they have no distinguishable parts. That makes Quine a “fact ontologist” rather than a “thing ontologist,”79 that is, rather than one who, like Bergmann, holds that presented facts have dislosophers] insisted when these engaged in the traditional dialectic...” p. 9. 78 Cf. “Ontological Alternatives,” in his Logic and Reality, pp. 132f. 79 Cf. Bergmann, “Strawson’s Ontology,” in his Logic and Reality, p. 181. See also E. B. Allaire, “Existence, Independence, and Universals,” in Allaire et al., Essays in Ontology, pp. 3-13.

227 tinguishable parts to which one must appeal in order to solve various philosophical problems. As Quine puts the point: One may admit that there are red houses, roses, and sunsets, but deny, except as a popular and misleading manner of speaking, that they have anything in common. The words ‘houses’,‘roses’, and ‘sunsets’ are true of sundry individual entities which are houses, roses and sunsets, and the word ‘red’ or ‘red object’ is true of each of sundry individual entities which are red houses, red roses and red sunsets; but there is not, in addition, any entity whatever, individual or otherwise, which is named by the word ‘redness’, nor, for that matter, by the word ‘househood’ ‘rosehood’, or ‘sunsethood’. That the houses and roses and sunsets are all of them red may be taken as ultimate and irreducible, and it may be held that McX is no better off, in point of real explanatory power, for all the occult 80 entities which he posits under such names as ‘redness.’

Suppose that we are presented with two red spots. We describe one using the sentence ‘This is red’ and the other with the sentence ‘That is red’. For Quine, ‘red’ is a common name. We need not rehearse the problems inherent in this notion,81 nor in the similar notion of a predicate being “true of” something.82 The crucial point is that the common name ‘red’ applies non-arbitrarily to this and to that. That is, there is something about each of the two spots as presented that justifies applying ‘red’ to both, as they are presented (which may not be veridically), rather than, say, ‘red’ to the one and ‘green’ to the other. This something is, of course, the fact to which Bergmann also appeals, namely, the fact that, with respect to color, the two spots are indistinguishable, that is, are the same. Nonetheless, the spots are two: they are distinguishable. Yet at the same time, according to Quine, neither do the two spots have any distinguishable parts. So, according to Quine, each spot is an unanalyzable and irreducible entity that is both the same and different from the other spot. The two spots in their simplicity, the pair and nothing else, ground the two relations of sameness 80

W. V. O. Quine, “On What There Is,” in his From a Logical Point of View, Cambridge: Harvard University Press, (1963), p. 10. 81 Cf. R. Grossmann, “Conceptualism,” in E. B. Allaire et al., Essays in Ontology, The Hague: Nijhoff, (1963). Also Allaire, “Existence, Independence and Universals,” and H. Hochberg, “Nominalism, General Terms, and Predication,” Monist, 61, (1978). 82 Cf. H. Hochberg, “Nominalism, Platonism and being True of,” Noûs, 1, (1969), pp. 413-419; and “Mapping, Meaning and Metaphysics,” Midwest Studies in Philosophy, 2, (1975), pp. 191-211.

228 and difference. Since the pair itself grounds these relations, they have been called internal relations. The nominalist has two internal relations, one for difference—here he or she agrees with the realist—and one for sameness—where he or she disagrees with the realist, who accounts for the sameness in question not by another internal relation, but by a shared part. Now, we contrast the complexity of a sentence with the simplicity of a name, that is, of a subject term or a predicate term. In this sense, for Quine the sentences ‘This is red’ and ‘That is red’, used to describe the spots, are misleading in their complexity. The spots, being simple, would be more perspicuously represented by names rather than by sentences: then the grammatical simplicity would reflect the ontological simplicity. Still, the names ‘this’ and ‘that’ won’t do, since that will serve only to represent the difference of the two spots. Rather, the names must be such as to, by themselves, represent both the sameness and the difference. The closest we could come would be ‘this-red’ and ‘that-red’, or, more generally, ‘red1’, ‘red2’, ‘red3’...and so on.83 Simple things which, without internal complexity, are both the same and different from each other, and are most perspicuously represented by internally complex names, have been referred to as “perfect particulars” by Bergmann and more recently by others as “tropes.” Thus, the phenomenologist uses sentences to describe presented facts, but these facts are held by the fact ontologist to be unanalyzable, and turn out, upon his or her view, to be tropes. The doctrine of tropes is the core idea of one version of nominalism, the thesis that there are no universals. A fact ontology like Quine’s is therefore nominalistic. The disagreement between Quine and Bergmann, between the fact ontology of the former and the thing ontology of the latter, is whether the presented facts really are simple, or whether such facts have distinguishable parts. But consider Quine’s examples: red houses, red roses, and so on; or, more simply, a red circle and a red square. These would be named 83

Cf. W. Sellars, “Naming and Saying,” in his Science, Perception, and Reality, London: Routledge and Kegan Paul, (1963), pp. 225-246. Sellars’ argument in this essay uses as its foil Bergmann’s ontology, and, indeed, is directed as negatively against this ontology as it is directed positively at a defence of tropes. Sellars considered Bergmann’s realist ontology of universals and particulars to be his main opposition. But if we are correct, then there are universals, and so Sellars’ trope ontology is wrong, and Bergmann’s realism is correct, at least in this regard.

229 by names like ‘this-red-house’ or ‘this-red-square’ or ‘this-red-round’. And surely, as a matter of phenomenological fact, no matter the device used to represent the red house or the red square in language, whether it be by (internally complex) names or by sentences, it is still the case that the red house and the red square are presented as complex and that by virtue of which ‘red’ applies to any such complex is part of and is presented as part of that complex. The nominalist uses names like ‘red1’ or ‘this-red,’ but these names have an internal complexity that represents the phenomenologically given complexity of what those names are used to represent.84 Construing the sentence ‘This is red’ as the complex name ‘this-red’ cannot deny the complexity of the fact represented by the sentence; all it does is represent it one way rather than another. This far from exhausts the dialectic of the universals vs. tropes is85 sue, but it does suffice to sketch, however broadly, certain crucial features. I have done this, not for its own sake, but rather to try to illuminate some other aspects of Bergmann’s ontology. First, however, another point about an ontology of tropes. We have two red spots. Call them “Socrates” and “Phaedo.” These are given in sense experience. Some arrive at a trope ontology through a commitment to a Principle of Localization, which asserts that whatever is given in sense is particular. If one accepts this Principle, as Plato for one did, then one is immediately committed to tropes: the spots are given in sense experience; so is the red in each; since what is given in sense is localized, the red in the one is localized in the one spot, the red in the other localized in the other. Thus, there are two tropes, the red in Socrates and the red in Phaedo. One might well be unhappy with the view that the two tropes in themselves ground both their sameness and difference.86 The dialectics being what 84

Cf. Bergmann, “Strawson’s Ontology,” pp. 185-86; cf. R. Ackermann, “Perspicuous Languages,” in M. Gram and E. Klemke, ed., The Ontological Turn, Iowa City: University of Iowa Press, (1974). 85 For an extended discussion of Bergmann’s views on tropes, see H. Hochberg, The Positivist and the Ontologist, Chapter Three: “Bergmann’s Realism and the Refutation of Bundle and Trope Ontologies”. Bergmann works out his position on tropes most fully in his Realism: A Critique of Brentano and Meinong, Madison: University of Wisconsin Press, (1967). 86 Cf. Plato, Theaetetus: Things known by sense are such that “each of them is different from the other but the same as itself...” (184a10). Sensible particulars are, of

230 they are, the concern is not unreasonable. One might therefore feel that it seems we call the tropes the same—they are both reds—without there being any ground for saying they are the same. This leads, as it did in Plato, to a transcendental argument: we call them the same, but there is nothing in sense that enables us so to call them; hence, there must be some third entity, not given in sense, but transcending it, to which the term ‘red’ applies primarily and in which both tropes participate, which participation justifies applying the term ‘red’ secondarily to the two reds given in sense. The argument may be put this way. The term ‘red’ is univocal; it applies, let us say, to the Red itself. The tropes Socrates and Phaedo are localized and therefore different. Each is not the Red itself. But in order to know that some entity is not something, say not F, then we have to know what F is. Thus, as the Phaedo would have it, in order to know that two pairs are not perfectly equal we need the concept of perfect equality. So, if we know of each trope that it is not the Red itself, then we must have a concept of the Red itself. This entity must be objective, since we all can come to agree about the reds that they are both instances of the Red itself. This objective ground of the sameness of the two tropes must be some entity, not localized and therefore outside of time and space, and not given in sense. This entity is the Form, the Red itself. This is Plato’s argument for the existence of Forms, one of them anyway. Aristotle uses a similar argument to establish the existence in things, i.e., for him “substances,” of common natures or forms or essences. The point is that such an argument is dialectically needed only because one is antecedently committed to tropes. And one is committed to tropes through a commitment to the Principle of Localization. But the doctrine of tropes is unacceptable: the properties of things are, as Bergmann argues, universals, and these universals solve the problem of sameness. So, since there are no tropes, there is no need for the Forms, and the argument for course, known by sense: thus, “the mind perceive[s] the hardness of something by means of touch”, and so on for other sensible objects. These are all localized; the ways in they are same and different are not found in sensible experience. For, “their being (that is, that they both are)” is known by the mind: this and “similarity, dissimilarity, identity and difference” are “in the class of things which the mind gets at by itself”; their “mutual opposition and...the being of their opposition” are in the same class, 186a1-c1, Plato, Theaetetus, trans. with an essay by Robin Waterfield, London: Penguin, (1987).

231 their existence collapses. Moreover, so does the Principle of Localization, which served as the premise from which it was concluded that the red in Socrates is different from the red in Phaedo. Since the red in Socrates is not different from the red in Phaedo but is something given in sense, it is simply false that what is given in sense must be localized. The Principle of Localization turns out to be mere a priori dogma. But it is a dogma with consequences. It is this dogma that generates the need for entities that transcend sense experience, and that leads to the transcendental argument for the existence of these entities. But we must know such entities so there is another transcendental argument for some sort of non-sensory way of knowing by which the forms of things are known. Not only do we need sensible experience (and inner awareness) as a way of knowing, we also need a sort of non-sensory intuition of these forms.87 But, since causal processes bring about things with different forms, these forms are among the reasons things are what they are, why they have the being they do, why these predicates and not those are correctly applied to them. The forms are thus the causes of the way things are, their being—the formal causes, in Aristotle’s way of speaking. But reason is the capacity to know the causes of things. That which grasps the forms of things is therefore rightly called reason. Our intuition of the forms is, in other words, a rational intuition. There are, in short, two worlds, that of sensible particulars, which are in space and time, and that of the forms, which transcend space and time; and then there two ways of knowing, one for each of these two worlds, the eye of sense on the one hand and reason as the eye of the mind on the other. All this comes from a dogma that can easily be seen to be unreasonable once one makes the positivist appeal to a Principle of Acquaintance as a rule for interpreting one’s ideal language L. This brings us back to Bergmann’s ontology. Consider once again our red circle. This is a whole, and is a complex whole. This whole has a color, it has a shape, and it has an area. The color so to speak fills or covers the area that the shape as it were surrounds. 87

Cf. Theaetetus: the sameness and difference of things, their ways of being, are “got at by the mind itself” (184a3); with regard to their being, “the mind by itself has the job of reaching a decision by reviewing them and comparing them with each other” (186b10).

232 This area is part of the presented complex, and it too is presented. Now, it is clear about areas that, at least so far as we can tell, for each perceived complex there is exactly one area that is part of it, and that the same area is never part of two distinguishable complexes. This is a general fact about perceived complexes: colors, shapes, and indeed all parts of such complexes, save areas, recur, while areas, alone among the parts of complexes, do not recur. If the problem of difference is that of discovering that by virtue of which two presented complexes are correctly judged to be two, then, as Bergmann argues,88 dialectically, areas can reasonably be claimed to solve that problem: one complex is distinguishable from another complex just because it contains an area distinguishable from every other part in it and distinguishable from every part, including the area, that is in the other complex. But something that is in a perceived complex, and that is not a universal, and that solves the problem of difference, is traditionally called a particular. So areas are particulars.89 And since areas are presented, so are particulars: particulars are given in perception. Such at least is the case, which does not seem unreasonable, that Bergmann makes. However, we need also to note that red (that is, the specific shade of red that we are presented with) turns out itself to be complex: red has as a part the feature of being a color—being a color is a quality that red has, just as red is a quality of the spot we are considering. This distinguishable feature of (the shade of) red is shared by (this shade of) red with all other shades of color; it recurs in blue, green, scarlet, etc. Similarly, the property of being a shape recurs in all the specific shapes, that is, in all the square shapes, in all the circular shapes, in all the pentagonal shapes, and so on. We are, in short, distinguishing properties and properties of properties. Now, just as color is a distinguishable property that recurs in, and is common to, all colors, so there is a distinguishable property that recurs in, and is common to, all areas, to wit, the property of being an area. This property is presented to one, just as the property of being a color is presented to one. It is by virtue of this property that we can identify areas as areas; it is this property that solves the problem of sameness with respect to areas, that is, it is this property that accounts for the facts that all areas are 88 89

Bergmann, “Synthetic A Priori,” in his Logic and Reality, pp. 288ff. Ibid.

233 the same, in respect of being areas, just as the property of being a color accounts for the fact that all colors are the same, in respect of being colors. It is by virtue if this property of being an area that we can formulate the rule for sentences like ‘This is red’ that the subject term is to be used to refer to particulars, i.e., areas. In such sentences, the subject-term, ‘This’, refers on each occasion of its use to a distinguishable area, but nonetheless on each occasion when it is used, it is used to refer to an area, some area or other. Or, if we use the customary standardization instead of the tokenreflexive device, where each ‘a1’, ‘a2’, ‘a3’, etc., names a distinct particular, i.e., area, then the fact that each of the ‘ai’ refers to an area is reflected in each of these names having that a-shape in common, while the fact that each refers to a different area is reflected in each such name having a different subscript. If we decide so to speak that the property that is shared by all and only particulars is the property of particularity, then, since areas are particulars, and since that property that all and only areas have in common is the property of being an area, it follows that the property of particularity is the property of being an area. The basis for identifying particularity with the property of being an area is, of course, the previous dialectical argument that areas are particulars: once the case is made that areas are particulars, that is, are the entities that solve the problem of difference, then the identification follows trivially. And it also follows, of course, that, since the property of being an area is presented, and is presented as being an area, so also we are presented with the property of particularity. I have said enough, I think, to have established that Bergmann’s method can reasonably be called positivist and that his ontology too can reasonably be called positivist, foreign though both the method and the ontology might be to Carnap or at least to the sort of philosopher that Carnap became. There is, however, one more feature of Bergmann’s ontology that should be mentioned. Bergmann’s L, as we have indicated, is essentially that of Principia Mathematica, unramified, and with descriptive constants added, interpreted in conformity to PA. Atomic sentences represent states of affairs in the world. Now, the atomic sentences of L are logically independent of one another: from one nothing follows about any other atomic sentence. But the atomic sentences represent states of affairs. But we take for granted

234 that L represents the way the world is. So, if the sentences that represent states of affairs are independent of one another, that fact must reflect the way the world is. It follows that states of affairs are logically, or, more appropriately, ontologically independent of one another. The logical independence of the sentences of L reflects the ontological independence of the entities they represent. The Tractatus puts the point simply: what holds for the sentences or propositions of L holds for the entities whose being they represent: 5.134: One elementary proposition cannot be deduced from another. 5.135: There is no possible way of making an inference from the existence of one situation to the existence of another, entirely different situation.

The ontological point is clear: “States of affairs are independent of one another.”90 But given that L represents nothing until it is interpreted, and since it is interpreted in conformity with the Principle of Acquaintance, it follows that the entities presented to us in experience are ontologically independent of one another. The entities must be such that there is nothing about any of them as they are presented to us that enables us to infer from one such entity anything about any other such entity. The Tractatus again puts it clearly: 3.1411: A proposition is not a blend of words.

So the entities in the states of affairs represented do not as it were blend into one another. And this is how they are presented to us. The entities presented to us are, as presented, such that none implies anything about any other thing. Each entity, as presented, is simply itself, that is, wholly itself, and not any other thing, not even partially any other thing. As Bergmann once put it, “Insistence on the self-containedness of the given” is “the very foundation” of positivism.91 He elsewhere insists that this selfcontainedness of what is given to us is a fact about the world, or, rather, about the entities of the world as they are presented to us, and forms a basic principle that makes possible Bergmann’s use of the formal logic of 90 91

Tractatus Logico-Philosophicus, 2.061 Bergmann, “Logical Positivism,” p. 8.

235 Principia Mathematica for his L. The interpretation of L must be such that the entities into which the primitive descriptive constants were interpreted were logically self-contained and independent of each other. The logical independence of sentences in L reflects this ontological independence of the things into which it is interpreted. His, or the positivist’s, way of interpreting the descriptive constants “could not be done” unless it were true that “There are several things with which we become acquainted if they are once presented to us. If one such thing is presented to us again, we recognize it.”92 One important way to make this point is made in the Tractatus: 2.0223: If two objects have the same logical form, the only distinction between them, apart from their external properties, is that they are different.

So two areas, besides being areas, that is, particulars, are simply different; so are two colors, besides being colors, simply different and so are two shapes, besides being shapes, are simply different. In no such case is the being of thing, that is, the being of the thing as it is in itself, constituted by some relation it has to some other thing. So, for example, it is not part of the being of this shade of red that it is closer on the spectrum to this shade of yellow than it is to that shade of blue: these shades, apart from being colors, are in themselves simply different. Similarly, it is not part of the being of red, which it shares with yellow in its being, that they exclude one another, that is, that the exemplification of the one by an area precludes the exemplification of the other: again, the shades of color are, in themselves, simply different. To be sure, the colors are located as they are on the spectrum, and they do mutually exclude the exemplification of the other. But these are external properties, not properties that constitute the being of the things as they are in themselves. To hold otherwise, to hold for example that color incompatibilities are necessary or essential truths, is, as Bergmann points out, to fall into idealism, after the fashion of Hegel or pragmatism (instrumentalism) after the fashion of Dewey.93 We have seen this already in another form. It is simply another way of making the Humean point that causality is simply regularity, or, in other 92

Bergmann, “The Revolt against Logical Atomism,” p. 44. Bergmann, “Logical Positivism,” pp. 13-14, and “The Revolt against Logical Atomism,” p. 45. 93

236 words, that there are no objective necessary connections. That is, among the things in the world there are objectively, among the things in themselves, no connections that could justify any inference from one to another. 5.136: There is no causal nexus to justify such an inference [that is, from one atomic state of affairs to another]. 5.1361: We cannot infer the events of the future from those of the past. Superstition is nothing but the belief in such a nexus.

Given that our or, rather, Bergmann’s ontologically perspicuous language L is interpreted in conformity to PA, we can say that the absence of such a nexus, the absence of such an objective necessary connection, is itself something grounded in the world as it is presented to us: each entity of the world is so represented in L as independent, since each such entity, as presented to us, is presented as an entity that stands in no such objective causal relation to any other such entity. There is no causal nexus in the ontology of the world because we are not acquainted with any such objective necessary connection: such a necessary tie, such a nexus, is not presented to us. Bergmann held constantly to the view that there are no objective necessary connections, and that this is reflected in the perspicuous language L that represents the ontological structure of the world.94 Heidegger has challenged any philosophy that uses the system of “mathematical logic” as a tool in ontology. One can see how much Berg94

Bergmann, in his Philosophy of Science, Ch. I, criticizes those who would introduce certain ways of handling dispositions or disposition predicates. He has in mind, among other things, the use of “reduction sentences” by Carnap and his followers. These philosophers are in effect (re-introducing Aristotelian forms or natures or essences as unanalyzable dispositions. They are, whether they know it or not, opting for a non-Humean view of causation. One reviewer of Bergmann’s book (Hilary Putnam) wonders why this is such a sin. It certainly is a sin for one to present oneself as an empiricist, as do many of those who adopt such a view of dispositions, and also to hold that non-Humean view of causation, as many self-styled empiricists also do. It is also a sin for one who holds such a non-Humean view of causation (as Putnam would permit one sinlessly to do) to also pretend (as Putnam seems to do) to accept that the world is such that the logic that yields its structure is one in which atomic sentences are logically independent of one another. But Putnam is certainly not alone in succumbing to the serpent in the garden nor alone in enjoying the fruit the serpent offers. Carnap is another.

237 mann is a positivist by looking at Heidegger’s views.95 At the same time, we can see how Bergmann offers an argument against the Heideggerian philosophy, which cuts to the core of the claims of the latter. In Bergmann’s work, the Heideggerian philosophy is not merely dismissed, as is done so often by so-called analytic philosophers, but effectively challenged. 95

There was in 1929 a conference in Davos, Switzerland, that brought together Heidegger and the neo-Kantian philosopher Ernst Cassirer. It was attended by Carnap, as a sort of representative of the positivists of the Vienna Circle. The conference and its upshot have recently been described in Michael Friedman, A Parting of the Ways: Carnap, Cassirer, and Heidegger. Carnap had studied under Cassirer and also Frege. Friedman tends to emphasize the impact of Cassirer and neo-Kantianism and to depreciate the impact of Russell. Cassirer, in his Substance and Function, LaSalle: Open Court, (1923), German original, (1910), struggles with the idea that any attempt to deal either with the world of sense or the world of science must go beyond Aristotelian logic and include a logic of relations. For Cassirer it really is a struggle. But for Russell, it is no struggle; see his Our Knowledge of the External World, London: Allen and Unwin, (1914); German ed., Unser Wissen von der Aussenwelt, Leipzig: F. Meiner, (1926) which makes this clear. Russell had mastered symbolic logic and within that the logic of relations, deriving from Peirce, and had also been deeply influenced by William James, who emphasized the presence of relations in the world as we ordinarily experience it. (James, in his Principles of Psychology, rather nicely referred to properties as “perchings” and to relations as “flights”.) Russell needed the logic of relations for his logicist programme in logic and the foundations of mathematics, but also for his attempt to provide a philosophically adequate description of the world of ordinary sense experience; it was James who gave him the phenomenological approach to the latter that made the use of relational logic indispensable. Carnap read Russell, and had seen the importance of the logic of relations (he sees that his Aufbau is “an attempt to apply the theory of relations to the task of analyzing reality” p. 7). But it was not just Russell’s logic that was important for Carnap; there was also Our Knowledge of the External World and also the Analysis of Mind, London: Allen and Unwin, (1921); German trans., Die Analyse des Geistes, Leipzig, (1927). These are clearly written, compared to the turgid prose of Substance and Function: Russell has mastered his material, Cassirer is still wrestling with it. The Aufbau clearly has these and other Russell texts (e.g., Analysis of Matter, London: Allen and Unwin, (1927)) as its inspiration, not Cassirer, no matter the importance of Cassirer for Carnap’s biography. (There are 23 references to Russell in the Aufbau, 3 to Cassirer.) Carnap had read Wittgenstein, too (there are 3 references to Wittgenstein in the Aufbau), but again, it is Russell’s program that he is inspired by, not Wittgenstein’s. Friedman makes rather too much of Cassirer and the neo-Kantian background and not enough of Russell. This is but another instance of the now not uncommon effort to downplay the significance of Russell, as a philosopher and as an influence. (By the way, the sanitorium at Davos is the setting for Thomas Mann’s The Magic Mountain.)

238 We should begin, however, elsewhere in Heidegger’s thought. And not with his “analysis” of Dasein which is in fact far from his central argument against any analytic philosophy that, like Bergmann’s, takes formal logic to be at the core of its ontology.96 We can in fact best begin our look at Heidegger by examining his views of science. We shall discover there an anti-Humean account of the structure of science that locates it is a context of supposed necessary connections that are rooted in turn, Heidegger claims, in the very structure of Being. As it turns out, Heidegger deals with science much in the way of Gadamer. This perhaps should not surprise us: Gadamer was a student of Heidegger. Specifically, Heidegger argues in his essay on “The Age of the World View,”97 that science, that is, the science of stones and planets, the science of physics and chemistry, seeks knowledge that finds its expression in the form of universal conditionals. The facts must become objective. The procedure must therefore represent the changeable in its change, bring it to a standstill, and nevertheless 98 let the motion be as motion.

Science seeks objectivity, that is, general agreement (that is the Kantian sense of ‘objectivity’, but there are, of course, other meanings to the term ‘objective’); it would have represented more careful thought had Heidegger not here used this ambiguous term. But he has his way of talking. The pattern explored is one of change, yet it is a general pattern, universal, and therefore timeless change brought to a standstill, but the standstill pattern or eternal, timeless pattern that describes motion, change. The science that seeks such a general pattern, Heidegger is saying, provides for itself a “blueprint” which effects the “demarcation of what na-

96

For criticism of other aspects of Heidegger’s philosophy, see F. Wilson, Socrates, Lucretius, Camus—Two Philosophical Traditions on Death, Lewiston: Edwin Mellen Press, (2001), Chapter Two; and also “Reflections on Kovacs’ Reflections on Death,” Ultimate Reality and Meaning, 25, (2002), pp. 225-239. 97 M. Heidegger, “The Age of the World View,” trans. Marjorie Grene, Boundary 2, 4 (1975-76), pp. 341-355. 98 Ibid., p. 344.

239 ture is to mean for the knowledge of nature that is sought.”99 We want to know what nature is like, so we have to demarcate a certain portion of the world as the object of study: this is what “nature” is going to mean for us. The knowledge that we are seeking is going to fit this pattern, this general description of that part of the world in which we are interested. This description that Heidegger gives is a not inaccurate description of normal science, though phrased in a way that is less than pellucid: the researcher in science, at least in periods of what Kuhn has called “normal science,” is guided by a Principle of Determinism and a Principle of Limited Variety, and the knowledge that he or she seeks is going to fit the pattern given by these Principles. Events, to be considered, must be “read into” this blueprint or form; events that we take to confirm or verify our hypotheses are only those that fit this pattern. For this part of nature only those events that fit the pattern are reckoned as natural; the rest (for these purposes) does not count as natural, it may even be treated as if it did not exist, that is, it can be ignored. It is only within the framework of this blueprint that a natural event be100 comes visible as such.

The hypothesis the acceptance of which concludes one’s research fits the antecedently given pattern. The design of nature...therefore contains its verification in the fact that for every step of its investigation physical research depends upon it in 101 advance.

Or again, ...the determination of the law occurs with a view to the ground plan of 102 the object area.

The experimenter’s research is guided by this plan, that is, in our terminology, by the relevant Principles of Determinism and of Limited Variety. [The hypotheses to be tested] are developed out of the ground plan of nature and are written into it. The experiment is that procedure which in its 99

Ibid. Ibid. 101 Ibid., p. 345. 102 Ibid. 100

240 arrangement and execution is borne and guided by the law basic to it, in order to produce the facts which confirm the law or deny confirma103 tion.

Heidegger, like Gadamer, supposes this logic of research is restricted to those sciences that employ mathematics in the representation of the world. Such sciences achieve exactitude not simply by virtue of being mathematical but because it is antecedently accepted that whatever law will be discovered will have a form of the relevant sort. This is so far fair. Put in more sober terms, all it means is that the Principles of Determinism and of Limited Variety that hold for that area assert that any relevant hypothesis will have a form that mathematically fits the pattern. Heidegger’s mistake is to suppose that the only way in which science can proceed through this sort of verification and hypothesis testing is if it is antecedently given that nature has this structure, that is, only if it is antecedently accepted that there hold for this part of nature Principles of Determinism and of Limited Variety that require the hypotheses to be mathematical. Now in this Heidegger is correct: the logic of experiment requires that such Principles hold for the area under investigation. But such Principles are by no means restricted to those that require laws of that form; there can be such Principles for areas where the laws are not mathematical. In Heidegger’s terms, there is no reason to suppose that there can be no natural sciences in areas of “inexactitude.”104 Heidegger makes the inference that since the patterns found in the biological and human sciences—the sciences of “living things”—do not have a mathematical form that permits “exactitude”, therefore they are different in kind from the physical sciences. But as we see, this is clearly wrong. Equally wrong is his condescending inference that because they are not mathematical the sciences of living things are harder: Of course, the delineation and the fixation of the object area of the historical disciplines remains not only different from but, as a task, much 105 more difficult than the realization of the discipline of exact science.

Even worse is his characterization of “busyness” as part of the essence of 103

Ibid., p. 344. Ibid. 105 Ibid., p. 347. 104

241 science106—a busyness that he clearly takes to exclude the reflective thought required for philosophy. This sort of thing—this name calling, if you wish—is, as I said, condescending, and not only is it condescending, but it is self-serving and narcissistic. What it says, in effect, is this: “look at what I, the humanistic philosopher, am doing: it is not science, and although I am only sitting in an armchair in an isolated forest hut, what I am doing is nonetheless much harder than the work of the physicist or chemist, and of course since it is humanistic it is in the end much more important.” And indeed, it really is important on Heidegger’s view, since, for Heidegger “metaphysics lays the foundations of an age.”107 It is thus not only condescending but also pretentious. It is, of course, just such pretensions that the positivists were determined to eliminate. In particular, Heidegger’s separation of the natural sciences and the sciences of living things was something that the positivists struggled against. It was something that Bergmann strongly opposed.108 This was another part of his positivistic commitment. A clear understanding of the logic of science establishes that Heidegger’s thought on these things, like that of Gadamer, is simply wrong and misguided. In fact, a careful reading of Hume or of John Stuart Mill would make this clear. But there is more to Heidegger than simply his confusion about the logical structure of science. After correctly arguing that work in the experimental sciences presupposes that there are Principles of Determinism and of Logical Variety that hold for the relevant area, he goes on to argue, this time incorrectly, that these Principles that guide modern science are given a priori to science, that they derive from and are founded on a source that science itself cannot know. It is a source and foundation that is beyond the realm of fact and of sense in which the hypotheses of science are verified and confirmed. It is, Heidegger suggests, to be left to metaphysics to bring to our awareness by reflective thought the “essential form” of an age. Metaphysics reflects upon a foundation of science that is beyond science but which is presupposed by science for its justification, a foundation 106

Ibid., p. 341. Ibid., p. 341. 108 Cf. Bergmann, “Holism, Historicism and Emergence,” Philosophy of Science, 11, (1944), pp. 209-211; or “Purpose, Function and Scientific Explanation,” Acta Sociologica, 5, (1962), pp. 225-238. 107

242 located in the “essential form” of the age. Here we have Heidegger invoking entities that are excluded by the positivist’s Principle of Acquaintance. Let us see if he gets caught in the positivist trap. For Heidegger natural science is the most significant feature of the modern world. The science he has in mind, we see, is a mathematized science, one for which, in our terms, the Principles of Determinism and of Limited Variety antecedently require that the generalization that describes the natural world be mathematical in form. Heidegger suggests that these Principles are themselves antecedently determined by metaphysics. Metaphysics lays the foundation of an age by giving the basis of its essential form through a particular analysis of the existent and a particular conception of truth. This basis dominates all the phenomena which distinguish the age. Conversely, it must be possible to recognize the metaphysical basis in these phenomena through sufficient reflection on them. Reflection is the courage to question as deeply as possible the truth of its 109 own presuppositions and the exact place of our own aims.

There is much in this passage that demands comment. We shall return to it in due course. But we can begin with the claim that it is metaphysics that “lays the foundation of an age” and is the source of the relevant Principles of Determinism and of Limited Variety. And the first point to be made in this context is that this is simply wrong, if by ‘metaphysics’ one understands something other than natural science. The Principles of Determinism and of Limited Variety are, as we have noted, laws about laws. Jointly they assert about a certain specific sort of system that there is a law that correctly describes the behavior of objects of that source and that this law is in a generically characterized range. This itself is, to emphasize, itself an empirical generalization, of a logical form essentially like that of the specific law it asserts to be there except that the description of nature that it gives is generic rather than specific (and therefore in terms of the logic of Principia Mathematica to be formulated in second-order rather than first-order logic). Such laws are confirmed when one discovers the specific law that they assert to be there, in the world: facts that verify or confirm the specific law also confirm the generic law. But the generic law is accepted as a 109

Heidegger, “The Age of the World View,” p. 341.

243 guide to research prior to confirmation of the specific hypothesis. What justifies this acceptance? If Heidegger is correct, then it is the metaphysical structure that so shapes the world that these generic hypothesis fit. But once one recognizes the empirical nature of these generic laws, one recognizes that they themselves are simple inferences from more general laws about laws. The Principles of Determinism and of Limited Variety apply to all systems of a specific sort. The more general law says that this specific sort is of a genus, and that all specific sorts of system under that genus have laws of the sort described by the two Principles. This more general law plays the role of what Kuhn called a paradigm,110 and we can refer to it with that term. So the paradigm predicts laws about specific sorts of system. Experiment confirms the specific laws that confirm the two Principles. And because the paradigm predicted those principles, the success of those predictions confirms the paradigm. The paradigm is accepted because it is confirmed through successfully guiding research.111 There is nothing in the nature of paradigms that requires them to be anything more than empirical generalizations, laws about laws, and certainly nothing that requires them to be grounded in something metaphysical that is different in kind from the natural laws of science. Heidegger simply misunderstands the nature or logic of scientific research; he fails to grasp the logic of the Principles of Determinism and of Limited Variety: these are taken by him, wrongly, to be something other than empirical generalizations. Since, on this misconceiving, they are not empirical, acceptance, Heidegger suggests, is to be grounded in something non-empirical. There is here a simple logical failure on Heidegger’s part, one that could have been prevented by a clear understanding of the logic of scientific theories and experimentation of the sort demanded by the positivists: the Principles are laws about laws, and being laws they are accepted on empirical laws. And one should know that one is going wrong as soon as one 110

See F. Wilson, The Logic and Methodology of Science and Pseudoscience. Bergmann makes this pattern clear in his discussion of composition laws, Philosophy of Science, Chapter Three. Newton’s composition law in mechanics together with the other axioms, all of which are laws about laws, leads to the prediction of the existence of Neptune and its orbit. Its orbit is a regularity or pattern. This pattern was confirmed by observation. These data that confirmed the orbit as predicted also confirmed the composition law that predicted that regularity. 111

244 is tempted to invoke non-empirical roots for the scientific framework. The positivists’ insistence upon verifiability should make this clear. And note that one does not need a precisely formulated principle of verifiability to make this point. In fact the relevant point is already in Hume. The ontological roots of science that Heidegger invokes are simply objective necessary connections that are to be rejected on Hume’s appeal to a Principle of Acquaintance, which is equally, as we have seen Bergmann point out, central to the positivist position.112 Heidegger’s thought clearly violates the strictures of the positivists, and moreover, like that of Gadamer, it is entirely without argument that he invokes entities that transcend the realm of the ordinary. The resolution to “question as deeply as possible” and to do so “with courage” turns out to be the audacity—arrogant, condescending and self-serving audacity— simply to assert, to assert to be sure with a tone of profundity and even at times a sense of vague moral uplift, but for all that to assert without justification. It is the audacity of a priori dogma. Heidegger too, like Gadamer, must be placed in the Aristotelian tradition. This comes out even more clearly in his comments on formal logic.113 His critique of formal logic turns out to involve those very objective necessary connections as does his account of science. Formal logic, he argues, is to be rejected because, if it is to be ontologically relevant (as Bergmann, for example, claims for his L), then it commits one to entities being logically and ontologically self-contained, and the presence of objective necessary connections rooted in Being itself implies that there is no such self-containedness. Heidegger, in 1912, published a report on “Recent Research in Logic.”114 In it he stated that “...it seems to me that above all it must be noted that symbolic logic never gets beyond mathematics and to the core of the logical problem.”115 He makes the same point in his dissertation: if 112

Bergmann, “Logical Positivism,” pp. 2-3. See the important essay of Albert Borgmann, “Heidegger and Symbolic Logic,” to which I am indebted. Borgmann’s essay appears in M. Murray, ed., Heidegger and Modern Philosophy, New Haven: Yale University Press, (1978), pp. 33-54. 114 “Neuere Forschungen über Logik” Literarische Rundschau für das katholische Deutscheland, 13, (1912), cols. 465-72,517-24, 567-70. Noted in Borgmann, “Heidegger and Symbolic Logic,” p. 5. 115 ”Neuere Forsuchungen,” col. 570, in Borgmann, p. 6. 113

245 logic were to be philosophically useful it would have to, as it were, impose on the world a separation of things that would prevent one from going deeper into the real connections of things. “...[I]ts formal nature [that of symbolic logic] prevents it from gaining access to the living problems of the meaning of propositions, of its structure, and of its cognitive significance.”116 The role of logic in doing philosophy gained in importance, especially as it was taken up by positivists such as Hans Hahn (Bergmann’s teacher) and Carnap. Heidegger was specifically the chosen object of Carnap’s attempt in his essay on “The Overcoming of Metaphysics through the Logical Analysis of Language”117 to show the philosophical relevance of symbolic logic and to show at the same time the emptiness of so much of what was then passing, and still passes, as metaphysics.118 Heidegger commented on this sort of claim in Being and Time: “In symbolic logic, propositions are dissolved into a system of ‘mapping and interconnecting’; they become the object of a calculus, not of an ontological interpretation.”119 Here again we find the assertion by Heidegger of a claim without argument. In this case it is the claim that there is something to reality that cannot be captured by symbolic logic, and therefore the further claim that the development of a formal language such as Bergmann’s L is at best irrelevant to ontology, and at worst something a commitment to which prevents access to deeper metaphysical realities. Heidegger replied more extensively to Carnap in lectures on metaphysics first delivered in 1935-36. There is an attempt here [in symbolic logic] at calculating the system of propositional connections by means of mathematical methods; hence this kind of logic is called “symbolic logic.” It sets itself a possible and valid task. However, what symbolic logic furnishes is anything but a logic, 116

Die Lehre von Urteil im Psychologismus. Ein kritisch-positiver Beitrag zur Logik, Freiberg, (1914), p. 97, n. 3; in Borgmann, p. 6. 117 R. Carnap, “Überwindung der Metaphysik durch logische Analyse der Sprache,” Erkenntnis, 2, (1932), pp. 219-241. Translated by A. Pap as “The Elimination of Metaphysics through Logical Analysis of Language,” in A. J. Ayer, ed., Logical Positivism, New York: The Free Press, (1959), pp. 60-81. 118 R. Carnap, “The Overcoming of Metaphysics through the Logical Analysis of Language,” trans. A. Pap, in A. J. Ayer, ed., Logical Positivism, Glencoe: The Free Press, (1959), pp. 60-81; first published in Erkenntnis, 2, (1931), pp. 219-41. 119 Heidegger, Sein und Zeit, Tübingen, (1960). References are to Being and Time, trans. J. Macquarrie and E. Robinson, New York: (1962). The passage cited occurs at p. 202, noted in Borgmann, p. 7.

246 i.e., a contemplation of the λόγος. Mathematical logic is not even a logic of mathematics in the sense that it determines or could at all determine the nature of mathematical thinking and mathematical truth. Rather, symbolic logic is itself a type of mathematics applied to sentences and sentential forms. Every mathematical and symbolic logic places itself outside whatever realm of logic because for its very own purposes it must posit the λόγος, as a mere connection of concepts which is basically inadequate. The presumption of symbolic logic of constituting the scientific logic of all sciences collapses as soon as the conditional and unre120 flective nature of it basic premise becomes apparent.

One can see what is Heidegger’s concern from this passage. Before commenting on this, however, some further remarks on Heidegger’s placing of symbolic logic are in order. Heidegger holds that symbolic logic is the “consistent degeneration” of logic, where logic is in turn the product of the thinking of the Greeks where their basic concern was with entities rather than being as such, a concern that originates from Being but in turn conceals it. The concern with entities rather than Being has not diminished. It continues to be realized in science, where everything is to be calculated. Modern symbolic logic is an outgrowth of this concern, and is part of the attempt of modern science at rendering everything calculable and in particular at reducing thought itself to endless and seemingly productive calculation. But this is to reduce thought to something that conceals what is really significant about thought, which is, of course, its capacity to proceed beyond entities to the Being which is their source.121 And to the Being which is also the source of this very attempt of science and symbolic logic to render everything calculable, to the Being that is the source of the modern way of thinking that conceals that source, to the Being that is the source of that which conceals itself from itself. Heidegger’s thought that symbolic logic has the power to dominate and to conceal is clearly expressed in the lecture course, What Is Called Thinking? Already symbolic logic is widely taken (particularly in Anglo-Saxon countries)122 as the only possible form of rigorous philosophy because its 120

Heidegger, What Is a Thing?, trans. W. B. Barton, Jr., and V. Deutsch, Chicago, (1967), p. 156. 121 Heidegger, The End of Philosophy, trans. J. Stambaugh, New York: (1973), p. 80. 122 Fortunately, at least to those who admire Heidegger, there remain the German-

247 results and procedures immediately yield something definite toward the construction of the technological world. Hence in America and elsewhere, symbolic logic, as the proper philosophy of the future, begins to assume its reign over the spirit. Through the coupling of symbolic logic with modern psychology and with psychoanalysis and with sociology, 123 the concern of future philosophy becomes perfect.

The pernicious nature of symbolic logic is made clear. At least what Heidegger takes to be its pernicious effect. But he also insists that this effect is not the result of mere perfidy on the part of human beings; there is a deeper cause, or, perhaps, higher. This concerning...is by no means the machination of men. Rather, these disciplines are under the destination of a power for which the Greek words ποίησις [poetry] and τέχυη [technology] perhaps remain the fitting names provided that they name for us, those who think, what makes one 124 think.

Symbolic logic, in spite of its misrepresentation of the logical structure of the world, is in the process of becoming the dominant form of all discourse.125 But it acquires this status only through the work, as it were, of a power, that is, of an active Being, which imposes that structure and so forms the world that that is the way the world is126—an active Being the speaking countries from which symbolic logic had been exorcised by the Nazis. They eliminated the real threat that logic posed to true thought, that is, the thought of Heidegger. Heidegger does not remark on this achievement of the Nazis. 123 Heidegger, What Is Thinking?, trans. F. J. Wieck and J. Glen Gray, New York: (1968), pp. 21-22. 124 Ibid., p. 22. 125 Ibid., p. 238. 126 One is reminded of Nietzsche’s comment on how “reason,” that is, the reason of the philosophers treats “the problem of error and of illusory appearances.” The concepts of the philosophers’ “reason” tell us that the world of ordinary experience is one of illusion; the concepts of “reason” tell us that our ordinary judgments about the ordinary world are inevitably flawed: we are “mired in error, drawn necessarily into error, precisely to the extent that the prejudice of reason forces us to make use of unity, identity, permanent, substance, cause, objectification....” In fact, he proposes, “...we enter into a crudely fetishistic mindset when we call into consciousness the basic presuppositions of the metaphysics of language—in the vernacular: the presuppositions of reason.” This “reason” of the philosophers is in fact anthropomorphic; it discovers causes that are outside the world of ordinary experience and that are understood in terms of unanalyzable activity. The “reason” of the philosophers: “…sees doer and deeds all over: it

248 working of which is grasped only by “us, those who think,” Heidegger and Company one presumes—where those who think are in fact the privileged or heroic few that are chosen by Being as thinkers, since the active Being is for those few who think that which “makes one think.” The idea of Russell and the early positivists, and, following them, Bergmann, was that one could construct a symbolic language that would show perspicuously the logical or rather the ontological structure of the world. The problem, according to Heidegger, is that it assumes only a “mere connection” of concepts and therefore only a world of logically or ontologically independent entities. It does this unreflectively, according to Heidegger. That philosophers who, like Bergmann, adopt as their philosophical tool a language that is that of symbolic logic take the world to consist of ontologically independent entities is quite correct: it is basic to Hume and Russell and the early positivists and to Bergmann that the entities in the world are logically and ontologically self-contained. Heidegger’s further claim that they adopt this position unreflectively is, however, wrong. It is done after reflection, reflection on what is in the world as given to us in ordinary experience. It is not simply assumed that the ideal language has this form, the form that is found in symbolic logic: it is ARGUED that it has this form. Heidegger does not credit such reflection on the part of Russell or Bergmann as genuine thought. Reflection, as he (Heidegger) intends it, takes one beyond the world of ordinary experience, which can be represented by a language in which there are “mere connections” among conbelieves that will has causal efficacy: it believes in the ‘I’, in the I as being, in the I as substance, and it projects this belief in the I-substance onto all things—this is how it creates the concept of ‘thing’ in the first place….Being is imagined into all things— pushed under everything—as a cause....” In this way the “reason” of the philosophers leads to the traditional metaphysics (and religion). But later, “...in a world more enlightened in thousands of ways, philosophers, to their great surprise, became conscious of a certainty, a subjective assurance in the way the categories of reason were applied: they concluded that these categories could not have come from the empirical world,—in fact, the entirety of the empirical world stood opposed to them.” “So where did they come from?” he concludes rhetorically. The answers of the philosophers are one thing; the reality is that they are illusions of language: “‘Reason’ in language: oh what a deceitful old woman this is!” (Nietzsche, Twilight of the Idols, Chapter on “‘Reason’ in Philosophy”, § 5, pp. 169-170).

249 cepts, to a level where there are connections that are more than “mere,” where the separability of things we experience disappears and the real connections among things are discovered. But the old problem remains: reflection in Heidegger’s sense seems to be, like Plato’s “reminiscence,” a way invented by the philosopher to give access to the supposed, and equally invented, necessary connections among things. Certainly, we are not told how to identify these special sorts of knowing. It is of course true that Heidegger is not so easily trapped by this sort of positivistic argument as was Plato or Aristotle. For these, and especially the latter, experiencing necessary connections had to be something that was a common occurrence, and the inability to show us how to identify them could be an embarrassment. But for Heidegger, while such necessary connections are there in the world, they are not there easily to be discovered. For the Being that founds those necessary connections at the same time so acts that also among the effects it founds is the way in which the world is known, and this way of knowing is one in which the necessary structures are hidden. The Being of the entities in the world, including ourselves, including ourselves as knowers, makes the world in which we live consist of apparently independent entities and makes our experience of that world the experience of a world of independent entities. If we take our knowing as a way in which Being knows the world it founds, then in this knowing of the world Being disguises itself, its real structure, from itself. So ordinary people and even ordinary philosophers do not and cannot perceive the necessary connections that underlie the apparent separateness of things. It is reserved to the heroic few, the genuine philosophers, usually not Anglo-Saxons and certainly not positivists, to experience those reflective modes of thought in which the real structure of the world is revealed. We must be grateful to Being for giving us a Heidegger who reveals that Being to us. At least, we must be grateful to Heidegger for revealing to us that he is a person who can reflect to a depth sufficient to grasp the real structure of the world. Anyway, it is said to be the truth that he reveals. In fact, he reveals, so he says, the real truth of things. To see what this involves we must recall another feature of the Aristotelian metaphysics, the coincidence of truth and being.

250 Now, in his essay on the Essence of Reasons127 (1969), Heidegger clearly takes for granted, here as elsewhere, that something like the Aristotelian account of reasons or causes is correct. To take this for granted simply begs the question against Hume and the positivists, about the capacity of reason to divine causes that go beyond the empirical, beyond the regularity view of Hume. Be that as it may, this is Heidegger’s starting point. And with that in mind—that dogma in mind—he attempts to delineate what is wrong with a logic that limits itself, like symbolic logic, to “mere connections” among concepts. He argues that there is more than “mere connection” to a proposition when that proposition is genuinely considered. Specifically, he argues that there is on the one hand the being of predication, that of “mere connection,” but that there is, on the other hand, beyond, or rather behind, it as the reason for the being of predication, a pre-predicative notion of truth. ...as the possible “subject” of a predicative definition, being must already be manifest both prior to and for our predications. Predication, to become possible, must be able to establish itself in the sort of manifesting which does not have a predicative character. Propositional truth is rooted in a more primordial truth (unconcealedness); it is rooted in the predica128 tive manifestness of being which we call ontical truth.

This, it seems clear, reflects the idea of the traditional substance philosophy of one like Aristotle that behind the properties present in the substance there is the Form or Nature or Essence the activity of which makes it the case that the various properties are present in it.129 What is predicated of 127

Heidegger, The Essence of Reasons, trans. Terrence Malick, a bilingual ed. incorporating the German text of Vom Wesen des Grundes, Evanston: Northwestern University Press, (1969). Vom Wesen des Grundes, Frankfurt am Main: V. Klostermann, (1949). 128 Heidegger, The Essence of Reasons, pp. 19-20. 129 J. Fritsche, Historical Destiny and National Socialism in Heidegger’s Being and Time, Berkeley: University of California Press, (1999), makes a similar connection to Plato and Aristotle, or, rather, to the substance philosophy as Aristotle’s emendation to Plato’s natural philosophy of the Phaedo; see p. 298. For discussion of this PlatonicAristotelian metaphysics, see R. G. Turnbull. “Aristotle’s Debt to the ‘Natural Philosophy’ of the Phaedo,” Philosophical Quarterly, 8, (1958), pp. 131-143; and F. Wilson, The Logic and Methodology of Science in Early Modern Thought, Chapter One. Also G. Vlastos, “Reasons and Causes in the Phaedo,” Philosophical Review, 78, (1960), pp. 291-325.

251 the substance is Nature naturata; and that about the substance that lies behind and produces those properties is Nature naturans, to use a terminology made familiar by Spinoza. And for those in the Aristotelian tradition, like Spinoza and like Aristotle himself, it is Nature naturans that constitutes the truth of the substance as it appears outwardly in the properties that are present in it. It is this Nature naturans that is the more primordial truth in which propositional truth, truths about the properties of things, finds its ground. Now, one of the important points that Heidegger goes on to make in this study is that Dasein projects itself in the context of a world. He raises the issue of “how the ‘relationship’ of Dasein to the world should be defined, i.e., how Being-in-the-world as the primordially unified constitution of Dasein should be expressed conceptually....”130 Two points are relevant here. The first is that Heidegger makes the understanding of Dasein’s being in the world a matter of grasping the unity of the two; conceptually; in that case, then, the reason that understands this relationship is a reason that grasps the unity of things, that is, the kind of reason that Hume and the positivists deny that we have but which Heidegger insists is necessary if we are to recognize ourselves as located in a more ultimate reality and objective meaning than is provided by the meaningless and unconnected world of everyday experience. The second point is that truth of this relationship is to be found in the being of Dasein. What is this world in which Dasein finds itself located? It is the world as a totality or whole, which Gadamer claimed it is the task of reason to grasp—real reason, according to these philosophers, not merely scientific reason. Heidegger tells us that “The world reveals itself to Dasein as the actual totality of what exists ‘for the sake of’ Dasein, but this means for the sake of a being that is equiprimordial with Dasein...”131 If Dasein creates out of its Being the world in which it finds itself and the limits of which it defines, then it is equally the case that the world is there in the primordial unity: the world too creates out of its primordial Being the Dasein that is in it. In their primordial roots, Dasein and its world—the world—cannot be separated. One is no doubt reminded of Hegel and the 130 131

Heidegger, The Essence of Reasons, p. 47. Heidegger, The Essence of Reasons, p. 101.

252 absolute Idealists. Interestingly enough, one is also reminded of Wittgenstein: “The limits of my language are the limits of my world.”132 and “The world is my world: this is manifest in the fact that the limits of language (of that language which alone I understand) mean the limits of my world,”133 and “Here it can be seen that solipsism, when its implications are followed out strictly, coincides with pure realism.”134 We should perhaps never forget that, while the Tractatus has logical atomism as its metaphysical core, as Bergmann sees quite correctly, it also has parts that point to a Being outside the world as described in the logical atomistic metaphysics, a Being that is at once the self that experiences that world as a whole but also is that world as a whole; it is a Being for which “solipsism coincides with pure realism.”135 But there is more to Heidegger’s thought about Dasein’s being-inthe-world than can be found in the Tractatus. This is the fact that the world creates Dasein’s being-in-the-world out of the primordial Being that grounds both. This is the creative activity of the world. As Heidegger puts it, “World never ‘is’: it ‘worlds’.”136 In this phrase—“the world worlds”—one is reminded once again of the similar expression that appears in the substance tradition and that has become familiar through Spinoza: one way of referring to the process of Nature by which Nature makes itself is “Nature naturans.” This is the primordial Being that produces out of itself the world in which we live, and this process is that in which “Nature natures.”137 It is not the case that Nature primarily is, that it is primarily Nature naturata, but that it is primarily or primordially an activity, a becoming, Nature naturans. The point is that this makes it clear that for Heidegger, the Being of Dasein has within it the Being of the world, and that the Being of the world has within it the Being of Dasein. Dasein thus finds itself located in a world that has a reality be132

Tractatus, 5.6 Ibid., 5.62. 134 Ibid., 5.64. 135 George Santayana saw this sort of thing clearly; see his Egotism and German Philosophy, London: J. M. Dent, (1916). 136 Heidegger, The Essence of Reasons, p. 103. 137 Cf. the interesting essay by F. Schäfer: “Heidegger’s Language: Metalogical Forms of Thought and Grammatical Specialties,” in J. J. Kockelmans, ed., On Heidegger and Language, Evanston: Northwestern University Press, (1972), pp. 281-301. 133

253 yond that of the ordinary world, the world of things as they are ordinarily experienced, the world of things that are, as Heidegger says, “ready-tohand.” There is, then, a reason that takes reason beyond the confines of the ordinary world of things that are ready-to-hand, to a more ultimate reality in which it can find a meaning not had by the ready-to-hand taken to be such. This reason is the reason of reflection, of metaphysics. It is this reason that becomes aware of those objective connections among things the existence of which renders shallow any ontology, like that of Bergmann, that is based on a logic, symbolic logic, that can only record “mere connections” among things. Canrap would object that such notions as that of a world that “worlds” violate the syntactical form of the logic of language, where that logic is declared to be the symbolic logic of Principia Mathematica.138 What he says in this respect is no doubt true. But one must go beyond that if one is to dismiss Heidegger. One must go after the substance metaphysics that Heidegger embraces. For, on this metaphysics there is more to the world than what is given in sense and can be represented in a logic like symbolic logic that aims only to record the order of sense. Beyond or behind the world of separable sensible appearances is the world of objective necessary connections, which give the real truth of things and which pro138

R. Carnap, “The Overcoming of Metaphysics through the Logical Analysis of Language.” The actual example that Carnap uses is Heidegger’s claim (in his lecture “What is Metaphysics?”, originally from 1929) that “Nothing noths.” What Heidegger says in the original German in the lecture “Was ist Metaphysik?” (in his Wegmarken, Frankfurt am Main: Klosterman, (1967), pp. 1-20 is “Das Nichts selbst nichtet” (p. 11). Carnap’s translator, A. Pap, translates this from Carnap’s original as “The Nothing itself nothings.” See the translation in Ayer, ed., Logical Positivism, p. 69. The English translator of Heidegger’s lecture renders it as “The nothing itself nihilates.” See M. Heidegger, “What Is Metaphysics?” in his Basic Writings, 2nd edition, trans. with Introductions by D. F. Krell. New York: Harper Collins, (1993), pp. 89-110, at p. 103. The latter is not quite right, missing the connection between the subject and the verb; Pap’s translation is better. I still prefer “Nothing noths.” This rendering originates, I believe, with A. J. Ayer. However one translates it, Carnap’s point is surely correct: it is hard to seen how one can think it syntactically correct; in Carnap’s terminology, it is indeed a pseudo-statement, which is not to say that it is impossible to give some sense to Heidegger’s talk about Nothing. It does take some work, however; and the result is that we discover that Heidegger’s point is hardly original, and certainly not correct. See Wilson, Socrates, Lucretius, Camus—Two Philosophical Traditions on Death, Chapter Two.

254 vide the foundation that justifies the use of symbolic logic. If one proposes, as Heidegger does, a logic that as it were lies behind symbolic logic, then it is no good simply to assert, as Carnap does, that symbolic logic really is self-sufficient. That is simply to pit Carnap’s dogma against Heidegger’s. To restrict oneself to symbolic logic one must go beyond Carnap and challenge Heidegger’s dogma. This is what Hume did and this is what Bergmann does when he insists on the Principle of Acquaintance: one argues that one is simply not acquainted with the objective necessary connections that Aristotelians from Aristotle to Heidegger have insisted are somehow there in the world. It is the argument from acquaintance that eliminates Heidegger, not the violation of logical syntax, that is, the logical syntax as specified by the symbolic logic of Principia Mathematica. That is why it is important to insist with Bergmann not only that the syntax of an ontologically perspicuous language L be that of symbolic logic but that this proposed L be interpreted in conformity with the Principle of Acquaintance. One can easily dismiss Heidegger’s pretentious metaphysics with a statement that “That’s nonsense!” but it is important to get right what makes it nonsense, and that is simply because it tries to introduce entities with which one is not acquainted: one can insist on “that’s nonsense!,” as Carnap does, but only if one adds, as Bergmann does, “because we aren’t acquainted with the entities you are claiming are somehow there!” If one rests content with Carnap’s reply and does not give grounds for excluding the transcendent entities, one is open to the reply that since the entities are there, the syntax of the language needs to be modified so that they too can be represented. One needs to exclude the entities, and only the appeal to PA will do that. In the case of Heidegger that appeal is easy. It is he who claims the existence of extra-ordinary entities and who claims that there are extraordinary ways of knowing them. Extraordinary claims require extraordinary arguments. These are not to be found. In their absence, Heidegger’s claims can be dismissed. We can perhaps see how Heidegger gets himself into some of his obscure positions. Suppose one accepts the Principle of Localization. Suppose, therefore, that one is committed to an ontology of tropes. That leaves one with the problem of sameness. Suppose one then proposes to solve this problem

255 as Plato did, through the positing of a realm of Forms. Now, the forms are a priori, relative to the world of ordinary experience. If one is to have objective necessary connections, then these too will have to be a priori, ontologically prior to the world as known in sensible experience. These objective necessary connections will be connections among the forms. So far as sensible experience of the world is concerned, all we can know is the regularity (a) All F are G = (x)(Fx e Gx) where F and G are Forms under which individual tropes, this-F, that-G, etc., fall. In our sensible experience of the world, all we experience are the tropes. These are separable, ontologically independent of one another. That is why (a) is a mere regularity. But there is a connection among the forms: (b) R(F, G) where this connection guarantees the truth of (a). (b) is a truth about the forms; it is therefore known a priori. Since (b) is such that it guarantees the truth of (a), the latter is also known a priori to be true. (a) therefore is more than a mere regularity; it is a necessary truth. That, at least, is how the story goes. There are, however, complications. If (b) is to guarantee the truth of (a), then we also need (c) R(F, G) e (x)(Fx e Gx) to be a necessary truth. No one, however, has proposed a logic in which (c) is a necessary truth. Bergmann has pointed out this fact: there is no guarantee that connections among the forms guarantee similar connections among the individuals that fall under the forms.139 The way out of this problem is often this. It is assumed that the connection (b) constitutes a pattern that guides a causal force that brings it about that the individuals falling under the forms appear in conformity to the regularity (a). This is the Aristotelian pattern. The structure (b) is an essential truth about those forms. So, (b) represents an essence. As an essence it is the causally efficacious nature of a substance. As an active essence, (b) is Nature naturans. It brings about in the world of sensible appearances properties that exhibit the pattern (a). This is Nature naturata. 139

Bergmann makes this point in his essay, “The Ontology of Edmund Husserl,” in his Logic and Reality, pp. 193-224, at pp. 204-205.

256 The essence as an active power so acts causally that the individuals that fall under the forms have the structure required by the essential truth (b); thus, as an active power, it guarantees that the substance of which it is the essence appears to sense with its properties ordered in the pattern (a). Those regularities among properties in the world of sense that are necessary have their truth guaranteed by the active essence of the substance of which those properties are the sensible appearances.140 This is the pattern found in Heidegger. Thus, Heidegger speaks of things appearing to us under the “destination of a power,”141 a power that lies beyond the world as it appears. This power is, ultimately, the power of Being, or, rather, the power, which is Being, that is, to emphasize, the active power that Being in itself is. It is the active power that is the World as it “worlds”—Nature naturans creating through its activity Nature naturata. It is Being that presents itself as beings, which beings are located in a necessary ontological structure, where the necessity of the conformity of the beings to this structure is guaranteed by Being as the power presenting itself as those beings, those things in the ordinary world of sensible appearances. The original sin is tropes, or, if you wish, the Principle of Localization. This creates the problem of sameness for which the Forms are the solution. But this in turn creates the problem of why the world as we experience it conforms to the structure of the Forms. Making the Forms active powers that create the world of appearances in conformity to the essential structure of those Forms solves this problem. This non-Humean causation is also a sin—much as Putnam would have it otherwise. But it is not the original sin. The original sin, as we have suggested, is nominalism. This is the ultimate source of Heidegger’s profoundly pretentious metaphysics. The positivist’s PA will exclude the entities that that metaphysics purports to introduce. But if Bergmann is correct, as we have argued that he is, that same PA can also dispose of the nominalism which is the ultimate source that generates the Heideggerian metaphysics. 140

John Stuart Mill had already made this point; see his System of Logic, eighth edition, ed. J. Robson, 2 vols., Toronto: University of Toronto Press, (1973), Book II, chapter ii. See also F. Wilson, The Logic and Methodology of Science in Early Modern Thought, Chapter two. 141 Heidegger, What Is Thinking?, p. 22.

257 So much for Heidegger: his metaphysics is one of those that Bergmann’s phenomenological method declares incapable of reconstruction. Return to the ontology of tropes. Platonism, including Heidegger’s version of Platonism, provides one solution created by this ontology. Bergmann’s method excludes this solution. Bergmann therefore comes to reject the ontology of tropes, Quine’s “fact ontology.” But the defenders of tropes have proposed another solution to the problem created by this sort of nominalism. This solution is a sort of linguistic nominalism. It is exemplified best, perhaps, by Wilfrid Sellars. This, the Sellarsian solution, proposes to solve the problem of sameness created by the trope ontology by making all sameness linguistic: two tropes are the same because they are called the same, because that is the linguistic convention that we have adopted. This-red and that-red, the red in Socrates and the red in Phaedo are both reds simply because we have chosen the convention that they are both called by the term ‘red.’ As one speaks nowadays, with Quine and Davidson, we have so chosen our conventions that the predicate ‘red’ is satisfied by both this and that.142 The difficulty with this position is, of course, dialectical; it really doesn’t solve the problem. But more deeply, there is the simple difficulty that it denies the obvious phenomenological fact that it is given to us in experience that things are the same. Contrast two red things—our two red shirts, our two red spots, Phaedo and Socrates—and a green thing; call the latter Simmias. The two red things are presented as having the same (shade of) color. Having let the term ‘red’ refer to the color of the one, and, setting the rule of interpretation that the same term will be applied to things we experience which are experienced as indistinguishable in respect of color, it is not arbitrary that we must apply that same term to the other red thing and, of course, not apply it to the green thing. It is a fact about the things presented—a presented fact about the things presented—that Phaedo and Socrates are indistinguishable with respect to color and are therefore the same in that respect, while both are distinguishable in that respect from Simmias, from which they both are therefore different. We can of course define x is gooch = Df x = Phaedo or x = Simmias 142

Cf. W. Sellars, “Naming and Saying,” in his Science, Perception and Reality.

258 and then hold that both Phaedo and Simmias satisfy the predicate ‘gooch’ and are therefore the same in that respect. Similarly, we can have x is red = Df x is Phaedo or x is Socrates and then hold that both Phaedo and Socrates satisfy the predicate ‘red’ and are therefore the same in that respect. This won’t secure linguistic nominalism, however. For, it is still true of these presented things that Phaedo and Socrates are presented as indistinguishable in respect of color while both can be distinguished from Simmias. Defining a class by listing its members, that is, defining it by extension, is one thing. And it is a matter of arbitrary convention which things we pick to be in the class we are choosing. It is quite another thing to specify a class by specifying that the members of the class be those things that are presented as indistinguishable (in a certain respect) from one another: this is not a matter of convention, not a matter of arbitrary choice, it is a matter of the way in which these things, as a matter of presented fact, are, that is, what they are, there in the world, prior to any decisions we might make about the conventions of the language we use to represent them. Bergmann’s method, which requires one to start with phenomenology, with the world as it is presented to us, has the immediate consequence that linguistic nominalism is untenable. Heidegger is close to this sort of linguistic nominalism. For Heidegger, Being as such discloses itself to us in things structured in conformity with the forms F and G structured as in (b) by R. The things that it discloses itself as are disclosed as F’s and G’s, and are so disclosed that they fit pattern (a). Being discloses itself as things that are the same and different according to the forms that are within it as Nature naturans, and in an order also deriving from the structure of those forms, that is, from the structure (b), and which exhibits itself in itself as Nature narurata, as the pattern (a). For Heidegger, sameness and difference and the order of the things that are the same and different all derives from Being as an active power. For the linguistic nominalist, such as Wilfrid Sellars or Quine, sameness and difference and the order of the things that are the same and different all derives from the linguistic community that establishes the conventions from which those patterns derive. Being—linguistic community—Heidegger—Sellars—they are of a piece: for both, the sameness and difference of things presented to us and the order of those things is not in

259 those things themselves but is established by, and grounded in, a sort of entity, that is, Being as such or the linguistic community, that is external to the world that we suppose to be given to us in ordinary sensible experience. Both are reckoned wrong and are dismissed through Bergmann’s positivistic use of the Principle of Acquaintance, through, in other words his phenomenological method. This method ensures that our ontology is of, and is securely tied to, the world as we experience it, not a world whose nature we get to know only by going outside of it to a realm of beings that transcend it. Take particularity or, more generally, the logical form of the world. It was Wittgenstein of the Tractatus who first saw that this logical form is a feature of the world, objectively there in the things of the world that is presented by our ontologically perspicuous language L. Wittgenstein notes that 3.221: Objects can only be named. Signs are their representatives. I can only speak about them: I cannot put them into words. Propositions can only say how things are, not what they are.

But objects have a logical form: individual constants name particulars, predicate constants name universals; and when one sort of name is concatenated in the appropriate way with another, as in ‘Fa’, then we have a proposition, something that represents a state of affairs by picturing it, in this example, the state of affairs of the individual a having the property F. The “appropriate way” to concatenate names is given by the formation rules of the language. The formation rule says that one concatenates signs with the a-shape and signs with the F-shape. This formation rule always results in a picture of a state of affairs of an individual exemplifying a property because signs of the a-shape always represent particulars and signs of the F-shape always represent properties. That a sign has the ashape shows itself in the sign; it is not something that is said, that is, said in a proposition. Similarly for signs showing and not saying that they are of the F-shape. Signs with the a-shape always name particulars; the shape in the sign thus represents the logical form in the object. It does not say that the object that it names has that logical form, it does not picture it, yet the logical form of the thing is there objectively in the object and the form of the name reflects this logical form. And so we find Wittgenstein saying:

260 4.121: Propositions cannot represent logical form: it is mirrored in them. What finds its reflection in language, language cannot represent. What expresses itself in language, we cannot express by means of language. Propositions show the logical form of reality. They display it.

and that: 4.1212: What can be shown, cannot be said.

He does succeed in saying that: 4.1221: An internal property of a fact can also be called a feature of that fact...

and that: 4.123: A property is internal if it is unthinkable that its object should not possess it.

So, particularity is a logical property, it is internal to certain objects, objectively there in them, and the objects are such that, if they do have this property or form, they have it necessarily. But none of this can be said, though it does show itself in language, at least in the ontologically perspicuous language L. And: 143

7: Whereof one cannot speak, thereof one must be silent.

So, here we have a world with a certain ontological structure. This structure is there, objectively there, in objects and in states of affairs. We know it is there because it shows itself in language. But in the end we cannot say anything about it: it is ineffable. This is true in particular of particularity. It is a virtue of Bergmann’s later moves in ontology that he established, as we have seen, that one can, Wittgenstein notwithstanding, speak about the logical form of the world, and in particular of the example we have chosen in order to illustrate the point, the logical property of particu143

Or, in the more pedestrian prose of the later translation, “What we cannot speak about we must pass over in silence.”

261 larity. Here, then, is one response to Wittgenstein’s claim that logical form is ineffable: find out how after all to say something about it. Another response is possible. It is this. If logical form is ineffable, then nothing can be said about it, and if nothing can be said about it, then it is itself nothing. Wittgenstein said that the logical form is something objective in the world. But he also showed that it is nothing. It therefore cannot be there, objectively, in the world. This response will come easily to one who finds metaphysics to be on the whole something that is objectionable. To be sure, one must distinguish, on the one hand, what can be called “bad” metaphysics, that is, metaphysics that tries to introduce entities that violate PA, and, on the other hand, “good” metaphysics that tries to say, though in a confused way, what cannot be said, at least not in the material mode of speech, simple talk within L, rather than talk about L. Bergmann made that distinction. So, on the one hand, he excluded, via PA, the transcendent entities of Plato and Aristotle and Heidegger and all the rest, and, on the other hand, attempted to explicate what remained so that good sense could be found in it. For such explication, talk about L was essential, that is, L interpreted in conformity to PA. And so Bergmann recognized that logical form was important and is there in the world as we experience it, and found a way to talk about that form, about, for example, particularity. But if one was not clear on this distinction between good and bad metaphysics, one might be tempted to throw the good out with the bad; after all, what good could there be in something that turns out to be ineffable? This was the response of those positivists who were more dogmatic than Bergmann, and less philosophically perceptive. It was the response of Carnap—Carnap who could not abide the mere association by Bergmann of the word ‘metaphysics’ with the term ‘positivist’: Bergmann was trying to do the impossible, he was trying at once to be a positivist and a metaphysician. But Carnap notwithstanding, it is possible to be both, as we have been arguing. However, Carnap’s response to Wittgenstein on logical form is one that could be expected: one way of taking the Tractatus is that, since talk about logical form is nonsense and that it is best therefore to be silent about it, what it is could be nothing more than nothing; if it is something, it is something that is not there in the world. This is not quite what the Tractatus says, but this work is sufficiently oracular and mysterious that one

262 could not be blamed for so reading it. Certainly, that is how Carnap read it: logical form might well be there, but there is nothing objective about it, it is a mere artefact of the language we use to represent the world. There are, to be sure, all the grammatical categories in one’s language L, the categories of subject term and predicate term in particular. But it has just been concluded that these represent nothing, that is, nothing objectively there in the world. What, then, can these be? The conclusion seems to be, for some at least, that it is a matter of linguistic convention. That at least was Carnap’s conclusion. Carnap took it that the meaning of the logical forms of the descriptive constants derived from representing some objective logical and ontological form but from the linguistic and logical roles that they had in one’s language, the rules that determined how they related to other terms, that is, the formation rules and the logical rules of inference. Meaning was not a matter of representing but of context. This is, of course, to resort to the idea of an “implicit definition.” The idea is that the axioms of a theory define the terms they contain—not, of course, explicitly define but rather “implicitly define.” One can find the idea in Hilbert who, commenting upon his axiomatization of geometry, wherein ‘point’, ‘between,’ etc., all occur as primitive descriptive constants, claimed that “The axioms of [order] define the idea expressed by the word ‘between’.”144 The essential point against this notion was made long ago in 1903 by Frege: “...one can never expect basic propositions and theorems to determine the reference of a word or symbol.”145 Syntax alone will not determine a fit to the world, it will not give a semantics. Moreover, the notion implies an unacceptable holism with regard to meaning; again, as Frege put it, “Only by stating all the axioms, which, according to Hilbert, belong, for example, to the definition of a point, does the word ‘point’ receive a sense. Accordingly, only through the totality of axioms in which the word ‘point’ occurs, does each of them receive a full sense. It is impossible to separate the axioms so as to regard some of them as holding and others as not holding, because in doing so we would also alter the sense of the ones which we wanted to 144

D. Hilbert, The Foundations of Geometry, Chicago: Open Court, (1902), p. 5; italics added. 145 G. Frege, “On the Foundations of Geometry,” English translation, Philosophical Review, 69, (1960), pp. 3-17, at p. 5.

263 count as holding good.”146 But, in spite of these criticisms, Carnap relied on this notion as the way to give meaning to the concepts of logical form. Carnap adopted what he called a “Principle of Tolerance”: use language of any form you wish so long as it works.147 That means that you can choose any universe of discourse that you please and fit your language to it. You thus get to choose what counts as a particular and what counts as a property and so on. The universe could be atoms, it could be numbers, it could be sense data, it could be material objects; any one of those sorts could count as particular. In this way “is a particular” could be “is an atom” or “is a number” or etc. In this sense, one can chose what meaning one gives to particularity and to the other elements of logical form. In this sense, the meaning of particularity is simply a matter of choice and of linguistic convention. And to this extent, Carnap’s response to Wittgenstein is correct: it is a matter of convention because it is simply a matter of choice. But after all, one chooses what “works.” Here one might have “works for the logical analysis of physical theory” or “works for the analysis of number theory” or “works for the analysis of the language of psychology.” All these are fine. There is one, however, that Carnap will not allow: this is “works for the solution of metaphysical problems.” It is precisely this one that Bergmann insisted be included in the list. It was this insistence that got him into trouble: he could not have that and be a positivist too.148 So, as Bergmann slipped into metaphysics, Carnap slipped away 146

Frege, “On the Foundations of Geometry,” p. 8. Carnap, Logical Syntax of Language, pp. 51ff. Carnap makes clear things are to be decided by laying down conventions: his “Principle of Tolerance” is “It is not our business to set up prohibitions, but to arrive at conventions” (p. 51). 148 Later Carnap was to distinguish “external questions” and “internal questions.” See his “Empiricism, Semantics and Ontology,” in his Meaning and Necessity, Chicago: University of Chicago Press, (1956), pp. 205-229. Thus, the philosophical issue of realism about material objects becomes the issue of whether or not to adopt as one’s language a language in which the primitive terms refer to material objects and their properties and relations. “Internal questions,” e.g., is this a tree?, occur as it were within the thing language, and “are to be answered by empirical investigations. Results of these observations are to be evaluated according to certain rules as confirming or disconfirming evidence for possible answers.” Philosophy is the task of clarifying these rules of evidence. “This evaluation [whether and to what extent the evidence confirms or disconfirms] is usually carried out, of course, as a matter of habit rather then a de147

264 from it.149 Bergmann found particularity and found out how to talk about it, and Carnap took another idea from Wittgenstein’s Tractatus and fell into publishing tomes of combinatorial arithmetic that he disguised as the necessary truths of logical probability. Others, like Wilfrid Sellars, made a response much like Carnap’s but more nuanced and more philosophical. Sellars notes that Carnap takes the patterns of logic to be given by what he (Carnap) refers to as L-rules. The logical rules take for granted those concepts, such as particularity, that constitute the logical forms of things. But what these pick out is a matter of convention. For logic, then, the L-rules alone are relevant to the meanings of the concepts such as particularity that give the logical form of objects. As one says, it was concluded that the concepts of logical form are “implicitly defined” by the L rules of the language. The meaning of these concepts is given, not by reference, but by the connections of these concepts to other similar concepts. Sellars distinguishes world-word transitions, word-word transitions, and word-world transitions.150 The first are rules for sentences to enter the liberate rational procedure. But it is possible, in a rational reconstruction, to lay down explicit rules for the evaluation. This is one of the main tasks of a pure, as distinguished from a psychological, epistemology.” But the issue of realism as opposed to phenomenalism is an “external question” and has no cognitive content. It becomes the issue of whether to adopt a language in which the primitive constants refer to material objects or, alternatively, to adopt a language in which the primitive constants refer to phenomena in the sense of momentary sensible objects and in which material objects are re-constructed as it were as patterns of such sensible objects. (Note that the sensible phenomena need not be taken as intrinsically mental, and can be construed as existing unsensed; in that sense, the phenomenalist can also be a realist, in another sense of ‘realism’). For Carnap the issue realism vs. phenomenalism becomes the issue of a thing language vs. a phenomenalist language. This not Bergmann’s way of speaking, but he would not disagree. The point is that for Bergmann the issue is a philosophical one and is to be debated in respect to making sense of the traditional dialectic. But for Carnap, such an “external question” is not subject to rational dispute at all, but only to purely pragmatic considerations of convenience. See “Empiricism, Semantics and Ontology,” § 2. 149 Carnap was to extend the notion of implicit definition from logical form to include non-logical concepts; these were his notions of “reduction sentence” (in his “Testability and Meaning”) and A-truth (in his “Meaning Postulates”); see F. Wilson, “The Notion of Logical Necessity in Carnap’s Later Philosophy,” § 3.5. 150 See W. Sellars, “Reflections on Language Games,” in his Science, Perception and Reality, pp. 321-358.

265 language game, rules of designation and rules for empirical evidence. The second are rules by which one sentence is inferred from another. The third are speech acts—of assertion, of denial, or supposing, of ordering, and so on. What Carnap was arguing about concepts of logical form, that is, for concepts like that of particularity, was that what count as world-word connections are irrelevant to their meaning; their reference is irrelevant to their cognitive content. Their cognitive content is given, rather, by the wordword connections that define their use, the L-rules in other words. Carnap also allowed P-rules in his language L; these are rules for physical inferences justified by their conforming to regularities found in nature.151 Carnap took for granted that these were unlike L-rules, in that word-world connections are relevant to the cognitive content of the concepts that occur in them (though he did waffle on this152). The descriptive terms, that is, the non-logical terms, have their cognitive content given in part by the L-rules. The P-rules give the word-word connections for these descriptive concepts. But he takes for granted that for these concepts their cognitive content is given by the designation rules that determine their reference. Sameness is objective, in the world, and concepts refer as they do because of these objective relations of sameness and difference. Sameness is not simply a matter of linguistic convention. Nor do P-rules obtain simply as matters of convention. It is because the things referred to are as they are that the P-rules are as they are. It was Sellars who drew the conclusion that just as world-word connections are irrelevant to the cognitive content of the logical concepts such as particularity, so for the descriptive vocabulary of L.153 For these concepts too, the cognitive content is given by the P-rules in which they occur. Their meaning in the cognitively significant sense of ‘meaning’ is not a 151

Carnap, The Logical Syntax of Language, pp. 315ff. See note 44, above. 153 See W. Sellars, “Meaning and Inference.” This paper accurately defines with unusual clarity many themes that are central to Sellars’ philosophy as he later developed it. He seldom states with such clarity his position that the cognitive content of a descriptive concept is not given by a designation rule but by the connections, logical and lawful, in which it stands to other concepts, i.e., that its cognitive content is given by its word-word connections and by nothing else. Others hold the same view, e.g., Rorty and Putnam and other defenders of “internal realism,” but few are as clear about it as is Sellars in this essay. It is unfortunate that this work is largely neglected. 152

266 matter of reference, not a matter of the designation rules, but is rather given by the P-rules, the word-connections that, as the theoretical propositions of science, connect these concepts to one another. What counts for cognitively relevant meaning are the P-rules, which give in effect the “implicit definitions” for these descriptive constants; what does not count for cognitive content are rules of designation, which justify their entrance into language use, either in word-word transitions as justified by the P-rules (but also the L-rules) or in word-world or language exit transitions in speech acts such as assertion. Sellars suggested—and just suggested, since he gives no argument to support his suggestion—that what goes for logical concepts like particularity goes parallel for the descriptive terms, like, for example, the descriptive content of the concept or term ‘red’: the meaning, in the cognitively relevant sense, of the descriptive terms is given not by their reference but by the P-rules, or, in other words, by the statements of law and of scientific theory in which they occur. Sameness and difference of things is now determined not objectively by the things referred to— objective reference is no longer relevant to cognitive content—but by the decisions of the linguistic community about how to reckon as same and different the entities in the world that introduce or re-introduce words or concepts into our language game. And so, relations of exclusion among colors become necessary truths, part of the meaning of color concepts. Thus, for example, the incompatibility of red and green, the truth that the presence of red in something excludes the presence of green, (x)[Re(x) e ~ Gr(x)] becomes part of the meaning of ‘red’ ( = ‘Re’) and ‘green’ ( = ‘Gr’), and true a priori, true ex vi terminorum. In Plato, the a priori necessity of such sentences is grounded in the Forms, in Heidegger it is grounded in the Being that structures the world, in Sellars it is grounded in the conventions determined by the linguistic community. In Heidegger sameness and difference and structure are created in the world by Being itself as it discloses itself to us and to itself in the world of ordinary things and of which we have or ordinary sensible experience. In Sellars, it is the linguistic community that creates the sameness and difference and structure that we seem to find in the world as it appears to us. In either case, it is a form of idealism in which there is an active power, essentially mental, which imposes

267 structure of the world where it is objectively not to be found. But if we adopt Bergmann’s phenomenological method, then we have an immediate response: sameness and difference are objectively there in the world as we experience it, as it is presented or given to us. This is the reply to the nominalism and the Principle of Localization, which are the real roots of the idealism found in both Heidegger and Sellars. As for the idealism of Heidegger and Sellars that makes color incompatibilities and other laws of nature true a priori, it follows for them that one cannot identify something as red without also recognizing facts about it, such as the fact that it is incompatible with green. But this too is false: What is given to us in experience is, as Bergmann insists, “self-contained.”154 In the case of Sellars, the slide into idealism and the conventionality of sameness and difference begins with Carnap’s move to de-mystify logical form as described in the Tractatus by making that logical form— particularity and the other formal or logical properties of things— conventional. But it is this that generates the move towards linguistic nominalism. The inference is clear: if the logical or formal properties are conventional and imposed on the world rather than found there, if they are, as Carnap suggests, conventional in defining this sort of sameness and difference rather than those categories being objective, then one can hardly avoid committing oneself, as Sellars recognizes, to accepting the same for the material properties of things, that is, to accepting that all sameness and difference, material as well as formal, are conventional, and therefore committing oneself to adopting the linguistic idealism consequent upon that acceptance. Bergmann’s phenomenological method provides the reply to the nominalism that generates the idealism. It also stops the slide into linguistic conventionalism that one finds in Sellars. That slide begins with Carnap’s move to make conventional the formal properties of things. One avoids the slide by refusing to follow Carnap in this response to the mystification of logical form, the declaration of its ineffability, that is found in the Tractatus. Bergmann’s phenomenological method here again saves us from the philosophical suicide that one finds in the positions of Heidegger and Selllars. One insists with the Tractatus, and against Carnap and Sel154

Bergmann, “Logical Positivism,” p. 8.

268 lars, that logical form, particularity for example, is objectively there in the world. This alone prevents the slide. Bergmann so insists. But that is not quite enough. One must also insist, as Bergmann also does, with Carnap and against the Tractatus, that there is nothing so ineffable about particularity and other features of logical form in the world that prevents us from speaking about them and with a clear conscience admitting them into our ontology. And let me emphasize, it is the Bergmann who is both a positivist and an ontologist—one who is both an anti-metaphysician and a metaphysician who is a realist—it is Bergmann as such a one, as a positivist and as an ontologist, who provides us with the philosophical tools to use to resist the obscurantist metaphysics of Heidegger and the nihilism of Sellars.155 Let me close by remarking once again that the metaphysics of the sort we find in Heidegger are not merely “ridiculous,” as Hume said of errors in philosophy, but also “dangerous,” as Hume said of errors in religion. The attitudes such as those we have noted—that “metaphysics lays the foundation of an age”—are not only arrogant and pretentious but also, let us emphasize with Hume, dangerous. They make a claim to power. This claim is based on the assumption that the philosopher knows these metaphysical truths. This knowledge is gained by “reflection” which takes “courage”: in such reflection we “question as deeply as possible” our presuppositions in order to locate “the exact place of our own aims.” Correctly locating the “place [in Being] of our own aims,” the philosopher is therefore able to provide for our well being by fitting us for patterns that conform correctly with the “foundation of [our] age.”156 Note three things. First, there is the clear claim of the philosopher— Heidegger, one guesses, or, passing on the mantle, maybe his student Gadamer—to the role and power of a philosopher-king. Second, the claim to such knowledge is spurious: it does not survive the positivist critique.157 155

And let us once again remind ourselves that we should not forget the linguistic idealism which one finds in Richard Rorty or the “internal realism” of such a one as Hilary Putnam. 156 Heidegger, “The Age of the World View,” p. 341. The context of these quotes has been given already, above. 157 Recall Nietzsche’s remark that with regard to philosophers, that is, about metaphysicians (he cites Kant) who declare this world less than real and dependent upon the reality of a deeper world beyond it, what is so dismaying about them is not “how often

269 And therefore, third, it is only the illusion of knowledge that moves the philosopher who makes these claims: in other words, the Philosopher-King will be imposing on us what are merely the products of his or her own fantasy life.158 and how easily they fall into error and go astray,” nor how easily such error is detected by the positivist critique, but rather that these spurious philosophers “...display altogether insufficient honesty while making a mighty and virtuous noise as soon as the problem of truthfulness i.e., even touched on....They pose as having discovered and attained their real objective through the self-evolution of a cold, pure, and unperturbed dialectic...: while what happens at bottom is that a prejudice, a notion, an ‘inspiration’ generally a desire of the heart sifted and made abstract, is defended by them with reasons sought after the event... (Beyond Good and Evil, § 5 ) 158 And it is dangerous. Consider one of Heidegger’s speeches to students delivered while he was Rector of the University of Freiburg, a position which he achieved after the Nazi seizure of power: “May you ceaselessly grow in the courage to sacrifice yourselves for the salvation of our nation’s essential being and the increase of its innermost strength in its polity. Let not your being be ruled by doctrine or ‘ideas’. The Führer himself and he alone is the German reality, present and future, and its law. Study to know: from now on all things demand decision, and all action responsibility. Heil Hitler!” Quoted in H. Ott, Martin Heidegger: A Political Life, trans. Alan Blunden, New York: Harper Collins, (1993), p. 32. The speech can be found in the Freiburger Universitätsreden, vol 13 (Ott, p. 391). Ott notes (p. 391) that “This lecture was one of a series of public lectures given in the winter semester of 1933-34 under the title ‘The role of intellectual life in the National Socialist state’.” The students are exhorted to give up “doctrine”, i.e., the church, and “‘ideas’,” i.e., the study of Plato—and the rest of the Western tradition—forget academic freedom (one of the themes of the Rectorial address) and do what Being, now unhidden, has made manifest, that a new order is now demanded and one must struggle as it commands. Hugo Ott reasonably comments on the remarks just quoted: “The verb ‘to be’, italicized by Heidegger for emphasis, contains within it the message of Being. Here the philosopher-rector—not just any academic figure who happens to be the leader and rector of the provincial university of Freiburg, but a true leader of German science and scholarship, which can only mean philosophy, Heidegger’s philosophy— has succeeded in reducing his thought by a monumental feat of compression and compaction, to a single valid formula. Here is the summons of Being in the forest clearing, the alethia: the locus of truth. In the troubled times before Adolph Hitler, Being itself—and no longer just the entities of being in the world—was called into question for the first time. He who was asking such questions was following in the footsteps of Heraclitus, ‘the obscure’ and yet ‘the bright’, inasmuch as he illuminates the primordial essence of truth in its hidden beginnings.” (Ott, op. cit. pp. 32-33). Ott’s is a detailed study that thoroughly demolishes the notion that Heidegger, when he joined the Nazi Party and extolled Hitler as the agent through which Being was doing its work, was a politically naive intellectual caught up against his will in a movement he did not

270 understand. The address that Heidegger gave upon becoming Rector is only slightly less political in its overt content than is this speech to the students. But there is the same call to heroic struggle, following the insights into the path of Being, as discerned through listening to its call. The new Rector notes that Geist and Volk are held to codetermine our Dasein; and there is also an intimate tie between the Volk and the state (Staat). Heidegger of course assumes that he himself has the required insight into Being, the required “engaged understanding,” and so when he proposes that the university be a place of “spiritual legislation,” what he is proposing is that he himself will provide this leadership. He will lead the leader. It is true that the term “Führer” does not occur in the Rectorial address, but the concept is there, and the reference, implicit though it be, is clear. See M. Heidegger, “The Self-Assertion of the German University,” trans. with an intro. by Karsten Harries, with a Preface by Hermann Heidegger, Review of Metaphysics, 38, (1985), pp. 467-480. Hermann Heidegger offers the usual apology, and manages to miss what is obvious, the extolling of the new order. The translation just mentioned of the address is followed by a translation of Heidegger’s own apologetic essay, “The Rectorate 1933/34: Facts and Thoughts,” ibid., pp. 480-502. There is a second apology in an interview with Der Spiegel, which can be found as “Only a God Can Save Us,” trams. M. P. Alter and J. D. Caputo, Philosophy Today, (Winter, 1976), pp. 267-284. J. Fritsche, Historical Destiny and National Socialism in Heidegger’s Being and Time,” argues persuasively that Heidegger’s Being and Time is largely an argument for National Socialism, in particular, § 74. It is interesting to note how the Rectorial Address advocates an ethics of heroic struggle, while the apologies explain away what he was doing by an appeal to an ethics of submission: the new Rector was for the most part doing simply what the new order, partly as interpreted by himself, demanded. The contradiction between the moral stances adopted as appropriate cannot be missed. Neither can one miss the hypocrisy of the apologies. Nor, further, can one miss, in both the address and the apologies, the arrogance of this philosopher: throughout he assumes he has insight into the worldhistorical development of Being and therefore of beings. There is little good to say about Heidegger. To be sure, his thought is in the end empty; it does not survive the positivist critique. But it found its audience. Hannah Arendt makes clear that in the ’20’s and ’30’s Heidegger, through his very obscurity, which secured an apparent profundity for his thought, found in the universities, among the students, a ready audience for his views. The feeling of depth which Heidegger was able to create was in itself important, sufficiently important to obviate the need for intellectual analysis. Carnap’s demand for syntactical clarity was ignored. The obscurity of Heidegger’s contorted German was preferred. (Heidegger suggested that philosophy could be done only in German and Greek. So much for Hume and Russell.) The clarity of the positivists could shed little light where the felt need for a profundity from beyond this world prevented the admission of anything that could so illuminate Heidegger’s thought that its emptiness could be recognized. (See Arendt’s essay, “Martin Heidegger at Eighty,” in

271 What could be more dangerous than that! One is consoled only by the fact that one knows that these arrogant philosophers who claim to know the “foundation of their age” rarely achieve power. Even when they do, perhaps as Rector of their university, it rarely amounts to much: professors are seldom world-historical figures, real politicians are usually better at the game than academics.159 Even so, when propagated the metaphysics becomes ideology that casts into contempt the ordinary world and the knowlM. Murray, Heidegger and Modern Philosophy, pp. 293-303.) Nonetheless, it is still hard to understand why Heidegger became famous, even revered. Why is Heidegger revered and Russell denigrated? Maybe it is a crap shoot, after all. In any case, positivism flourished in the Vienna Circle and the Austrian sunlight where in Germany it was lost in the obscurity of the mists that were everywhere present in faculties of philosophy, at least until 1936, with the assassination of Schlick. By then the flight was on. Einstein had gone to Princeton (but the mathematician who secured him a position earlier at Prague, George Pick, was a victim of the camps of the Nazis). Among philosophers, the existentialists of course stayed. But by 1936, Wittgenstein was long gone, back to Cambridge, and Feigl and Carnap were already in America. Neurath was to get out shortly thereafter, fleeing first to Amsterdam and then, with the invasion of the Netherlands, escaping in a small boat to England. Bergmann did not get out until 1938. In saying that positivism and the Vienna Circle flourished until 1936, might one also be permitted to mention in this context Carnap’s publication in America in the same year of “Testability and Meaning,” where he repudiated the use of the positivist’s Principle of Acquaintance. 159 Heidegger objected when the politicians, for whom he would be their philosopherking, did not take his philosophy seriously, but was all a matter of politics, and they out-politicked him in his exercise of what he viewed as the prerogatives of the Rectorial post he had conspired to take over. For the politics of Heidegger’s becoming Rector, see Ott, Martin Heidegger: A Political Life. For a careful analysis of Heidegger’s Rectorial address, see C. Bambach, Heidegger’s Roots, Ithaca: Cornell University Press, (2003), Chapter Two. However, see also Graeme Nicholson, “The Politics of Heidegger’s Rector’s Address,” Man and World, 20, (1987), pp. 171-187. Bambach brings out well Heidegger’s petulance at not being taken seriously by the powers that then were. But as Ott, Martin Heidegger: A Political Life, emphasizes, Heidegger never repudiated or apologized for what he had done and said, or ever recognized that he had done or said anything that was wrong. It won’t do, as Nicholson does, and as others do, to point out that, like many others, Heidegger was submitting to the “spirit of the age” (to use John Stuart Mill’s phrase): when it comes right down to it, Heidegger, like King John, was not a good man. One can only speculate: would Heidegger’s taking seriously the positivist critique have helped?

272 edge we gain of it through experience—and the democratic institutions through which, we have learned, experience can best be transformed into action for the general good. What we need, we are in effect told, is not experience and certainly not democratic debate but reflection; this alone reveals the truth of our ordinary world, not by looking at that world, but by going outside it to its profounder “foundations” in the realm of Being. This is reality being challenged by the fantasy the ideology mistakes for the structure of Being that “founds the age.” The point is that, even if we find that this fantasy is rarely imposed on us, its propagation undermines legitimate politics and reasonable efforts based on our ordinary (but fallible) human experience of the world to “place our own aims” in that world and, one hopes, to satisfy as far as we reasonably can those human aims. Nietzsche put it clearly. This sort of metaphysics denigrates our ordinary world and our experience of it. The world of the philosophers, according to the philosophers, is “the womb of being,” it is “the intransitory and hidden god,...the ‘thing in itself’—that is where the cause [of our ordinary world] must lie and nowhere else!”; compared to this world, the world of ordinary experience, our metaphysician characterizes as “this transitory, seductive, deceptive mean little world,” full of “confusion of desire and illusion.”160 Our metaphysicians have us indulge “our fundamental tendency…to assert the falsest judgments (to which synthetic judgments a priori belong) are the most indispensable to us”; they would have us “measuring reality against the purely invented world of the unconditional and self-identical....”161 The world beyond the ordinary is the realm of truth, the realm of the ordinary is the realm of untruth—truth vs. falsehood. But of course, the realm of the ordinary is where we live, it is the realm of life. The metaphysicians would have us embrace the truth, would have us “renounce false judgments”; that, however, “would be to renounce life, would be to deny life.” That means that if we wish to live, live as ordinary human beings, “we must recognize untruth as a condition of life.”162 Or rather, if we are to take the ordinary world for the reality that it is, that here there is genuine truth and not there, in some fantasy world beyond our own, then we must renounce the metaphysicians and their world of pure 160

Nietzsche, Beyond Good and Evil, § 2. Ibid., § 4. 162 Ibid., § 5. 161

273 truth and being itself—that is the falsehood that must be renounced. ...perhaps we and the logicians will one day accustom ourselves to getting along without that little ‘it’ [in ‘it thinks’] (which is what the honest 163 old ‘I’ has evaporated into).

What is needed is a philosophy that accepts what the metaphysician denies, what is needed are persons “who represent by word and deed the unbelief in the lordly task and lordliness of philosophy,”164 those, in other words, who accept the principle—the positivist principle—that the life we live is to be lived in the world of ordinary experience. Hume, and following him, Nietzsche and the positivists argued that there can be no a priori knowledge of fact. They also argued that there can be no a priori politics.165 Heidegger denies both—and proposes his own thought as providing that a priori knowledge and founding that a priori politics.166 Dangerous indeed. We must be vigilant in continuing to pursue the positivist programme. 163

Ibid., § 19. Ibid., § 204. 165 Details aside, this is the basic sense of the “emotivist” view of ethics—which is the central thought in the ethical theories of Hume, of Nietzsche, and of the positivists. 166 Compare the following, clearly political judgment: “... if fateful Dasein, as Beingin-the-world, exists essentially in Being-with-Others, its historicizing is a cohistoricizing and is determinative for it as Destiny. This is how we designate the historicizing of the community, of a people. Destiny is not something that puts itself together out of individual fates, any more than Being-with-one-another can be conceived as the occurring together of several Subjects. Our fates have already been guided in advance, in our Being with one another in world and in our resoluteness for definite possibilities. Only in communicating and in struggling does the power of destiny become free. Dasein’s fateful destiny in and with its ‘generation’ goes to make up the full authentic historicizing of Dasein.” M. Heidegger, Being and Tine, trans. J. Macquarrie and E. Robinson, New York: Harper and Row, (1962) § 74, p. 436. The point is clear: the individual person finds his or her full being not in a community of individuals coming together by agreement (e.g., in the marketplace, or in a liberal democracy [read: Gesellschaft, or: Weimar republic]), but as a people (read: Volk, Gemeinschaft). That is where human beings find their real destiny. And this is Heidegger’s a priori politics, a call to action based on nothing more than “philosophical” reflection on what is supposed to be the real reality of human beings. Clearly, this is Heidegger’s manifesto for the politics of the right—not liberal politics, not merely conservative politics, but radically right politics—which in 1928 and the years following can mean only one thing. 164

274 Bergmann understood this. He recognizes the pernicious evil in a society being permeated by pretentious and pretending metaphysics such as that of Heidegger, and he warns against the cultural impact of such thinking even where it is not dominant—it can at its least dangerous be “bothersome,”167 but in the immediate post-war world he warned us that we must be wary of such “tunes” as those that “sound very sweet to the T. S. Elliots and to the Pied Pipers from Chicago.”168 What he would have thought of the later, more sombre, tunes from the Dalhousie seminars of George Grant and James Doull, let alone the Chicago seminar of Leo Strauss, I do not know. But for the latter, at least, see Shadia Drury, Leo Strauss and the American Right,169 to uncover how such stuff and such tunes are once again not just bothersome but dangerous. Dead though Bergmann’s way of doing philosophy seems sometimes to be, if one reads what often passes now for profound philosophical thought, one will realize that one still needs, the world still needs, the sort of positivism defended by Bergmann. But it also needs an ontology of the sort defended by Bergmann, one formed by his phenomenological method. They both—such a positivism and such an ontology—stand for commonsense and the world of ordinary experience against the claims and fallacies of the pretentious. They both show that a concern for this world as we ordinarily experience it is genuinely more profound than thought that searches for truth by attempting to fly to another world. For his positivism and for his ontology—for both of these, Bergmann truly deserves to be celebrated.

167

G. Bergmann, Review of H. Kelsen, General Theory of Law and State, in Ethics, 57, (1947), pp. 213-15. 168 See G. Bergmann, Review of José Ortega y Gasset, Mission of the University, in The Mississippi Valley Historical Review, 32, (1945), pp. 266-67. 169 Shadia Drury, Leo Strauss and the American Right, New York: St. Martin’s Press, (1997).

Reminiscences

Reminiscence of Gustav Bergmann CARL LEIDEN University of Texas at Austin

A

lthough I didn’t know it at the time, my ultimate link with Gustav Bergmann began in the mid-30’s when Harlan Herrick and I became best friends in Boone, Iowa. Beset by illness throughout his life, Harlan was dedicated to high intellectual achievement. We stimulated one another but I have always thought I was the real gainer by our friendship. I not only learned chess and what I know of music from Harlan, but our daily discussions, arguments and competitive commitments led to insights and knowledge that could not have been learned in school. And it was Harlan who introduced me to Gustav Bergmann. Harlan was valedictorian in the class of 1940. We both put in a year in Junior College in Boone. In the fall of 1941 Harlan went to the University of Iowa and I to Iowa State University (then College). Harlan was to do his degree in mathematics and philosophy—hence his inevitable encounter with Bergmann—and I in mathematics. Then was intervened in my case and illness in Harlan’s so that we basically received our degrees at the same time in 1945. In 1946 I had an MPA from Wayne State University and Harlan an MA from Iowa. He did that degree with Bergmann; his thesis was on some of the propositions of Whitehead and Russell’s Principia Mathematica. In the summer of 1946 I arrived in Iowa City to begin a Ph.D. in political science; Harlan had married and was about to go to Yale to work on a Ph.D. in mathematics. In talking to me in the short time that we saw each other in August 1946, Harlan said, “You just have to find a way to meet Gustav Bergmann.” Harlan’s later employment with IBM in New York led to his membership in a team that created the computer language FORTRAN. He attributed his ability to do this to his work with Bergmann in Principia Mathematica. By one of those rare coincidences, he worked at IBM with Stefan Mengelberg, my best friend at Iowa State University in 1944-45 and, like Gustav

278 Bergmann, a refugee from Nazi oppression. By my good fortune in making friends with genius, I had also in Mengelberg a friend who was truly multi-faceted—a truly perfect—mind. A mathematician and a legal scholar—as was Bergmann—Mengelberg was also the assistant conductor of the New York Philharmonic Orchestra under Leonard Bernstein. How lucky I was to have known these men and to be able to call them friends. After Harlan left Iowa City, I made sure that I would meet Bergmann by simply walking over to the building that housed philosophy and knocking on Bergmann’s office door. And that began it all. I registered for the first semester of his Philosophy of Science seminar and followed it up with a second semester and then a summer session on the Philosophy of the Social Sciences. Then to top it off a readings course under him in which I read and discussed with him Hans Kelsen’s Reine Rechslehre. A few years later I was a visitor at the University of California at Berkeley. Hans Kelsen was on the faculty then, but when I was there, he was on leave and I never met him. By chance, Bergmann lived in the same part of Iowa City in which I lived. Indeed, we were only a block apart. Often we would wait for the bus together. Or on some occasions we would return from downtown Iowa City on the same bus. Waiting with Bergmann always meant lively conversation and I came to cherish these opportunities as much as I did the formal encounters with him at the University. It was not long before I conceived of a dissertation in which some of Bergmann’s ideas might be applied to the slippery and soft slopes of political science. Bergmann encouraged me in this and helped block out a rough adumbration of what a dissertation might be like. I had, of course, to have a chair out of the Political Science Department and I chose Kirk Porter, the then department chair. Porter was uneasy with what he termed the unusual arrangement, but he did not oppose it. It was only some months later when I had completed several chapters of the proposed dissertation that I detected his basic unwillingness to go along with the idea. He said, “Why can’t you be like the others and write on Iowa?” I caught the drift instantly. I hurried to consult Bergmann and he said merely that it did not surprise him. But he urged me to go along with Porter, get the degree and get on with life. That was May 1948. I chose a topic on political redistricting in Iowa, did the research and, working over

279 the summer, defended in December; by early 1949 I had my degree in hand and was in my first teaching position. But I had not written the dissertation I wanted, nor did I have Bergmann on my committee as I desired. Bergmann was a natural seminar professor. It is hard for me to visualize him in a lecture situation. I grew accustomed to his continually circling the seminar table, at which six or seven of us would be seated, and talking or answering questions or guiding discussion. I myself never had his talents in this, finding the lecture hall more to my tastes, but I have often regretted not being able to lead a group of young people, as he did, in their wrestling with ideas and thoughts. After a removal of about 70 years, I perhaps can be forgiven in not remembering in detail the contents of his seminar. But one area does stand out. Bergmann did devote several weeks to an analysis of Darwin’s theories of evolution and I gained my first and real understanding of it from his exposition. He was proud of his association with the Vienna Circle. I had not heard of it at the time. As I recall, the University of Chicago had published some booklets on the unity of science from the point of view of the Vienna Circle and we used several of them as texts, so to speak. Bergmann then, at least, was anxious to clear the shrubbery from around scientific efforts and speculation. This meant the abandonment of much metaphysics. And of course religion had no place in the structure. Whatever his later views, his ideas then have clung to me ever since. He sometimes spoke of Kurt Gödel, the logician, whose 1931 paper on the incompleteness of arithmetic-like structures (e.g., Principia Mathematica) upset the apple cart of modern mathematics. The two were classmates I believe, and friends. Bergmann at least explained what the 1931 paper was all about. What Bergmann did for me, and probably for others too, was to plant ideas that in many cases matured only years later. It would be 20 years before I used a long plane trip to read and study Gödel’s 1931 paper. In fact, I have spent much of my life in a soft science, clinging to ideas that Bergmann first gave me in the late 1940’s. As a seminar teacher, Bergmann was always courteous and patient. He gently refused to be drawn into discussions of religion or ideology. This did not mean he was ideology-free but only that his classes were as

280 purified of these things as he could make them. His class on logic and scientific method, which I did not take, was an undergraduate effort that enamored many an undergraduate. There is a medical doctor, John Parks of Muscatine, Iowa, and a mathematician before he went into medicine, as an example. He took that class very early on and was so impressed by Bergmann’s expositions of it that he considered it the most important intellectual experience he ever had. He has steadfastly extolled the virtue of Bergmann’s work and has made many efforts to rejuvenate interest in it. As a teacher myself in later years, I could only appreciate the frustrations of being a teacher in the late 1940’s. Once I had to participate in a program created by my department. It conflicted with Bergmann’s seminar. I asked him whether he would excuse me for the purpose. He said that he would gladly do so but then went on to grumble about those who never asked but simply stayed away. I was reminded of this many years later when Steven Weinberg, Nobel Laureate in Physics, taught an introductory physics course to non-physics majors in my University of Texas. As one of his students told me, one day Weinberg found two-thirds of his class missing; I had been grumbling about all the absentees I had but then remarked that those who passed up Weinberg on physics to sleep a little longer or otherwise waste their time could not be forgiven. It was foolish to pass up Bergmann too. I always came away from his classes with something worth saving. But students don’t always know their own best interests. What kind of grader was Bergmann? I have no memory of tests but of course there may have been some. We did write papers however, and individually we were sometimes asked to report to the seminar on some topic or to lead the discussion. In short, there was ample opportunity for Bergmann to evaluate us. Like many professors, Bergmann fancied himself a chess player. I was then at the peak of my chess playing skills, probably somewhat shy of the master level. I played a little chess on some Sunday afternoons at the University Union with Kirk Porter and I recall seeing Bergmann playing there as well. It was not long before he invited me to his home to play chess. He did not like to lose—who does?—and would often emit a string of imprecations when he found himself in a bad situation. I don’t recall ever losing to either Porter or Bergmann. As a result they wanted to play

281 me less and less; finally they did not play at all. Chess is a game of skill that requires constant practice and neither had any time for practice. But while it lasted it was fun. When I went to Bergmann’s house I met his wife Leola. She was a very attractive woman of perhaps 35 or so, of Norwegian background and holder of a Ph.D. in American Studies. Undoubtedly she was an inspiration to Bergmann, and perhaps buried in his prose or symbols there could be found her ideas as well. In any case she was and remains a charming person. Each gave the other a great deal and both were very fortunate indeed. But I was fortunate in finding at the University of Iowa between 1946 and late 1948 one of the truly inspirational teachers. I learned from him many things of great value; today they permeate my mind so thoroughly that I can no longer separate them out. Today Harlan and Stefan and Gustav are no longer among us. We are gathered here today to honor Gustav Bergmann. He left Austria under duress but America and Iowa have been richer for his presence. And most ironically, given my interest and knowledge of the hard sciences and mathematics, my later career veered widely from expectations. In publications and in teaching I became a specialist in Middle Eastern politics along with violence and revolution. It was a far cry from philosophy of science as Bergmann thought of it. Yet whatever I did I found ideas and inspiration from the mind of Gustav Bergmann. I do so still.

Reminiscence of Gustav Bergmann (1987) REINHARDT GROSSMANN Indiana University

I

made the acquaintance of Gustav Bergmann in the fall of 1952 in Iowa City, Iowa. This happened thirty-five years ago, but I still remember our first meeting. I studied at that time at the Pädagogische Hochschule in Berlin and had won a Fulbright scholarship in psychology. I was sent to the University of Iowa which was famous for its experimental psychology. I reported the next morning to the Psychology Department and was greeted the American way, by a young assistant professor by the name of McAllister: “Well, you are the student from Germany. We have arranged everything for you. But you must first meet Professor Bergmann; he, too, is from Germany (!) and he has his office right around the corner in the same building.” He took me in tow and, after a few steps, we arrived at the open door, as is the custom in the States, of Bergmann’s office. Bergmann was seated at his desk, reading a book, paper and pencil at hand. We were introduced, McAllister disappeared, and I sat down. I was the first young German Bergmann had met since the war, and it was obvious that he was very reserved, but at the same time curious what the Nazis had achieved with their indoctrination. We talked for half an hour in English. Bergmann asked me about the occupation of my father, where I went to school, what I wanted to study, and why I was interested in philosophy. When I told him that I had been already in high school interested in the philosophy of the Vienna Circle, and that I had read some articles by Feigl, Hempel, and Carnap, he visibly relaxed. He had convinced himself that I was not a disguised Nazi and that I had good taste in philosophy. Years later, he confessed to me that our conversation had convinced him “that the decent Germany had not vanished.” I believe that he never seriously doubted this anyway, even though he had good reason to. As so many others, Bergmann was forced to leave

284 Austria in 1938. In the summer of that year, he had applied for an immigration permit for the United States, and a good friend of his, Dr. Walther Mayer, had supported his application. He had lived since 1934 in the States and was a professor at Princeton. Dr. Mayer states in his affidavit that his weekly salary is $85, that he has a bank account of $1435, and that he lives in a four-room apartment. The Israelitische Kulturgemeinde of Vienna certified on August 9th, 1938, that Bergmann was a “pure Jew,” and the military office in Vienna, therefore, had no objections to his immigration. Bergmann went from Vienna to The Hague, where he was met by Otto Neurath. Neurath gave Bergmann a relatively large sum of money for his voyage to the States and the first few weeks in New York City. As Bergmann describes it in a letter to me, Neurath said with his booming voice: “This is not a gift; this is your honorarium for a memoir you are to send me on your impressions of the Vienna Circle.” Bergmann followed Neurath’s order. Among his posthumous papers is a twelve-page description of his impressions of the Vienna Circle, which he wanted to be published after his death…. [See the bibliography of Bergmann’s publications, at the end of this book, for the locations of both the English and German versions of this memoir.] In New York City, the Ph.D. of mathematics and law earned his living as an actuary and bookkeeper with the company of S.H. and Lee Wolfe. Bergmann also made inquiries about how many of his Austrian semesters would count toward a degree in bookkeeping, in case he would decide to enter the university. The answer was that his previous study entitled him to four years of high school and two years of liberal arts college. Fortunately, Bergmann did not become a bookkeeper. In March 1939, he was employed by the University of Iowa, in Iowa City, where he lived until his death a few months ago. He got his position at the university through Herbert Feigl. Feigl was an old friend of Bergmann’s from the days of the Vienna Circle. He had arrived in the States at the beginning of the thirties and had taught since then at the University of Iowa. He put Bergmann in touch with the psychologist Kurt Lewin who also taught at the University of Iowa. Lewin was looking for a collaborator with a special combination of interests; namely, someone who was not only interested in psychology, but who also

285 had mathematical training. Lewin wanted to apply topology to problems in theoretical psychology. Bergmann was perfect for this role. Thousands of intellectuals arrived at that time in the United States and were looking for jobs at the universities and colleges. The Rockefeller Foundation, overwhelmed with requests for help, set aside a certain amount of money, which was administered by a “Committee in Aid of Displaced Foreign Scholars.” The condition for help was simple: if a university was willing to employ a refugee, he received $1800 for the first year from the Foundation; after that it was sink or swim. The University of Iowa, with the recommendation of Lewin, employed Bergmann under this condition. On March 15th, 1939, Bergmann became a research assistant in the Child Welfare Department of the University of Iowa. At first, Bergmann worked for Lewin at Harvard, where Lewin was during the spring semester; and then he followed him to Berkeley, where Lewin taught during the summer. Finally, in the fall, Bergmann arrived at Iowa City and taught there for the rest of his life until his retirement. In June, however, he visited Iowa City for three days, on the way from Harvard to Berkeley, and met on that occasion a young lady who worked on her Ph.D. and who was employed in the Child Welfare Department. Her name was Leola Nelson. Four years later, in 1943, Bergmann married Leola. During the years after Bergmann’s arrival, the small Department of Philosophy at the University of Iowa became a department of national status. Feigl soon left for the University of Minnesota, but Wilfrid Sellars had arrived at Iowa a year before Bergmann and, in addition, Everett Hall was a member of the Philosophy Department. For a short time, then, there existed in the middle of the endless fields of yellow corn a small group of philosophers who, partly through their numerous students, greatly influenced the philosophical life of the United States. One is tempted to call them the “Iowa Circle.” I saw Bergmann the last time in April 1986 in Iowa City. As always, he received me and my wife with open arms and an infectious smile. I realized that his memory and power of concentration were impaired, even though he could be as sharp and quick as ever. A few months later, it was discovered that he had Alzheimer’s disease. Our friendship lasted for more than thirty-four years. Just as Neurath, Mayer, and Feigl once helped the

286 refugee to make his home in the United States, so Bergmann later on helped me to settle down in the States and become a philosopher. In 1962, Bergmann taught for one year in Sweden. He wrote a letter to one of his former students and his good friend Edwin Allaire in which he talks about a meeting with a German colleague: “We have a visitor [in Lund], a “good German,” stout anti-Nazi from way back… I was the first Jew she personally met, it was a tremendous problem for her…clear, is it not, it is much harder for a decent German to be a German, than for a surviving Jew…so, when I saw who she was, I of course met her with outstretched hand and open heart…” Ten years had passed since Bergmann’s first, very cautious, meeting with a young German and this very gracious greeting of a German by a surviving Jew.

Reminiscences of Bergmann’s Last Student L. NATHAN OAKLANDER University of Michigan-Flint

I

am very happy to have this opportunity to reminisce about my experiences with Gustav Bergmann. Due to a series of events that I will recount before I conclude this talk, I did not go to GB’s memorial service the year after he died. For that reason, having the occasion on GB’s 100th birthday to talk about what he meant to me as a teacher, mentor, scholar and, dare I say, father figure, is extremely welcome. In February 1964, at the age of 18, I arrived in Iowa City from the Bronx, via the Village. I received my BA with honors in philosophy in 1967, spent one year at Ohio State (because Herbert Hochberg, Alan Hausman and Robert Turnbull were there), returned to Iowa from 1968 through 1972, and then started my teaching career at the University of Michigan-Flint. At the end of my first year at UM-Flint, I returned to Iowa to complete my dissertation, and it got done in no small measure due to Bergmann’s determination to help me get it done. Of that more later. I want to start at the beginning of my education at Iowa and talk not only about Bergmann, but also about some of my experiences with other teachers while I was at Iowa. I came to Iowa knowing that I wanted to be a philosophy major. I assumed that philosophy and religion went together and for that reason I should take courses in religion as well as philosophy. With that mindset I went up to the Field House to register. I walked up to the philosophy table and after saying that I wanted to major in philosophy, I asked the faculty member there, somewhat hesitantly: “I guess I ought to take a course in religion?” Laird Addis was the faculty member and he responded: “Why would you want to do that?” He explained to me why there is no reason to take courses in religion if you are majoring in philosophy. Upon reflection I now view that experience as my first lesson at Iowa in learning to be wrong, and in how to develop a thick skin. For a student at Iowa in the 60’s and early 70’s, those were, I found, valuable lessons indeed.

288 My first experience with Bergmann as a teacher was in the second semester of my sophomore year. I was a student in his “baby” (undergraduate) philosophy of science course. It was a great course, and there was one particular experience that stands out in my mind even though it occurred over 40 years ago. He had completed his discussion of scientific determinism when he started to get excited. He started to raise his voice and then he literally yelled that there is absolutely no incompatibility between freedom and determinism. I had never before experienced a professor screaming in the classroom, and it was inspiring to see someone so passionate about an abstract philosophical idea, and it was also somewhat frightening. I certainly would not want him yelling at me like that! Yet during that semester, I also saw a kinder and gentler side. Iowa City had an arts movie theatre—the Iowa theatre—where foreign films were frequently shown. I was standing in line waiting to see an Ingmar Bergman movie, and I was on crutches, due to a foot injury. GB and Leola were also in line, and he must have recognized me as one of his students for he inquired with some concern about my injury, and mentioned it to Leola. His interest meant a lot to me then, and I still recall it now. Perhaps this is a good place to mention a story about Bergmann that I recently heard from James Van Cleve, who did his undergraduate work at Iowa roughly at the same time as I did. He had taken GB’s undergraduate philosophy of science course and in his senior year he registered for GB’s more advanced course, History and Systems of Psychology. Van Cleve recounted how on the first day Bergmann asked if there were any undergraduates in the class. Those who raised their hands were asked to stand up. Once they stood up he said to them, “OK you can leave now.” I did not have any direct contact or communication with GB for the remainder of my undergraduate education, but his presence was palpable. Through my courses with Laird Addis, Phillip Cummins, Robert Baker (with whom I did my honors thesis on G.E. Moore and the problem of other minds) and Edwin Allaire (whose seminar on perception I sat in on during the summer when I graduated), I became keenly interested in ontology in general and Bergmann’s ontology in particular. I remember after the end of my sophomore year going to California for the summer packing

289 one philosophy book; namely, Essays in Ontology.1 I also remember, in my senior year, there being quite a bit of animosity between Allaire and Bergmann as Allaire came to reject the basic tenets of Bergmann’s philosophy. The policy at Iowa at the time was to have undergraduate philosophy majors who wanted to pursue graduate work in ontology spend at least their first year of graduate school away from Iowa. I was one such student. With Hochberg and Hausman at Ohio State, I thought I would still feel at home, philosophically speaking, in Columbus. With Hochberg I studied contemporary philosophy, Russell, and universals and particulars, and with Hausman, Descartes. When they left so did I. During my first semester back at Iowa I took Bergmann’s seminar on mathematical logic, devoted to the Liar Paradox. It was the second semester of a two-semester course on mathematical logic and in every class I was completely blown away. Thank heavens for my good friend and fellow graduate student Silvano Miracchi whose notes saved all of us. That semester I also took a course, taught by Addis, on the philosophy of time, in which I had to give a report on a selection from Richard Gale’s anthology The Philosophy of Time by C.D. Broad titled, “Ostensible Temporality.”2 My keen interest in the philosophy of time has continued virtually unabated since then. There are two beliefs that Bergmann asserted that come to my mind here: He often said that we each stand on someone else’s shoulders and that great philosophers always go back to the same fundamental problems. I certainly do not view myself as a great philosopher, but I have continually returned to issues surrounding the philosophy of time (and I am indebted to my former teachers). Early on, my second or third graduate semester, GB gave a seminar on C.D. Broad’s The Mind and its Place in Nature.3 This was a particularly important course for me. It was my first opportunity to present a paper before GB, and I worked very hard to do a good job, and make a good impression. The room in which the seminar was held had an oval table 1

Essays in Ontology, (ed.), Edwin B. Allaire. The Hague: Martinus Nijhoff, (1963). C.D. Broad, “Ostensible Temporality,” in The Philosophy of Time (ed.), Richard Gale. Garden City: Anchor Books, (1967). 3 C.D. Broad’s The Mind and its Place in Nature. Eighth impression; London: Routledge & Kegan Paul, Ltd., (1968). 2

290 with the students and GB sitting around it. I began to give my presentation and I could see that GB was pleased. He seemed to be agreeing with what I was saying and was nodding and smiling as I read on. Then, much to my amazement, he got up on the table and lay down on it, resting his elbow on the table with his hand holding up his head rather like a pin-up girl. He looked at me with such (philosophical) delight (if there is such a thing) that I probably started to blush. When I was finished he looked at me and said that there was a dissertation on Broad. (Leave no stone unturned, I thought, reflecting on the dissertations of several of his former students.) Since I was still a year or two away from my comprehensive examinations I certainly wasn’t thinking about my dissertation, but at that moment it was clear to me that I would be doing my dissertation with GB! Due to my interest in the philosophy of time and the influence of McTaggart on Broad, I asked GB if we could meet to discuss McTaggart’s The Nature of Existence.4 He agreed, and we decided to meet on Saturday mornings for a couple of hours. I had an experience during one of those Saturdays that is memorable, but first let me set the stage for it. Our meetings generally took the following form. Starting from the beginning of McTaggart’s book, I would come in with something to say about what I had read and then pick out a passage that was unclear to me, or one that I thought was particularly rich and ripe for analysis. Bergmann listened to what I had to say, and he would read the passage I had picked out, and then the fun would begin. GB liked to write as he thought through arguments. So he would construct arguments by scribbling them on a piece of paper, and it was left to me to reconstruct them after our session. He would be going on and I would blurt out a question or make a comment—needless to say, it was not very easy to get a word in edgewise—and that would generally excite him more. I in turn fed off his energy and was able to follow him just enough to keep the dialogue going. Sometimes he got mad and yelled at me, but he quickly made it clear that he was not really angry at me and, in any case, my skin had thickened enough to easily endure it. At other times I knew that he was pleased with our talks. One occasion particularly stands out in that regard. 4

John Ellis McTaggart, The Nature of Existence (ed.), C.D. Broad. Cambridge: Cambridge University Press, (1927).

291 I do not remember what the topic was, but we were both having a good time. He was exhibiting his brilliance and I with a question or comment led him to even greater heights. And so it went, back and forth until finally GB jumped up in his chair slammed his fist on the table, bent his neck forward and put his face next to mine and said, as only he could (or would!), “Nate, we fuck so well together!!!” What does one say to that? I confess I do not remember what I said at that point, but I will never forget what he said. Another time, after an ongoing interchange he said to me “enough foreplay.” Working on my dissertation with GB was also quite an experience. I left Iowa in the fall of 1972 to begin my professional career at the University of Michigan-Flint where I still teach. Naturally, I missed Iowa and saw how difficult it was for my colleagues to complete their dissertations while they were teaching, so I decided to go back to Iowa during the summer after my first year of teaching, and GB strongly supported that decision. At the time my wife Linda was pursuing a Masters degree in Library Science (after having completed her MA in Philosophy at Iowa five years earlier), so I went to Iowa City by myself. When I arrived I had one chapter—the first, on propositions and time—in hand. I gave it to GB and he read it quickly and was very pleased. He said to me that if the other chapters were like this one then there would be no difficulty in completing the dissertation. Unfortunately, things didn’t go as smoothly for the remaining chapters, which were yet to be written. I gave GB my second chapter, which he gave back to me with copious comments and said we needed to talk about the material. For the remainder of the summer working with him went more or less as follows. I would go to his office and we would discuss the chapter in Broad that I was going to be writing on. We would go through the arguments and, as I mentioned earlier, he wrote as he thought. I would then start working on the notes he wrote, my recollections of the discussion, and the text. I then rewrote the second chapter and gave it to him. He would return it to me usually the next day with very extensive comments. If he thought that I still didn’t get it right, then he would call me back into his office and we would go through the text once more. I had heard stories about professors (not at Iowa, of course!), who held on to chapters for weeks on end, or were on sabbatical and could not

292 read chapters until they returned, and so on. My experience was quite the opposite. I had the impression that GB was in his office waiting for me to give him something to read. And then as soon as I gave him something to read, he would read it. I remember one occasion, working on the next to the last chapter on Introspection. It was a complicated chapter and I was having a hard time with it. We had discussed the material and I wrote up the chapter. I turned it in and he did not like it. So we talked again, and again he didn’t like it. So I rewrote it yet again. I gave him the chapter and waited in an office down the hall while he read it. An hour or so later and he came from his office to mine and stood at the door and handed it to me and shook his head, alas, side to side and not up and down. I will never forget how I felt, but somehow got I through it. The last chapter was also noteworthy. As usual, we met to discuss the structure and content of the chapter and as he talked he wrote down notes of his arguments. I returned home and began to write the chapter and study his notes when I noticed an error in his reasoning. For the first and only time I called GB at home, and explained to him the mistake I had taken him to make. He listened attentively. He accepted my criticism and was proud of me for offering it. When I was finished he said to me, “Nate, you have become a philosopher.” He was so excited that the next day, when he saw me in the hallway while he was talking to Panayot Butchvarov, he called me over and recounted to Butchvarov our conversation the day before, and how I was now a philosopher. With Bergmann’s determination to get my dissertation finished, my determination to finish it and Linda’s patience to type and retype it (Linda was making mistakes and had to retype several times as she was watching the Watergate hearings at the same time), I completed the dissertation by the end of the summer. In October I returned to Iowa for my oral defense. These are difficult memories since I think of myself as a strong-willed person but Bergmann is such a domineering person. At my defense I was hardly able to get a word in edgewise. Whenever a criticism was directed toward me he would answer before I could. I recall that one time during the orals I got so angry that I interrupted him and said that I wanted to answer the question. I also recall one occasion during that summer when GB invited Linda and me to his home for dinner. I think it was then that I first had an

293 opportunity to meet and get to know Leola. Linda and I must have been rather shy on that occasion since shortly after that I recall GB asking me if Linda and I ever talked! Once I returned to Michigan I corresponded with GB on a fairly regular basis. I would share with him my good fortune at Michigan professionally and he would promptly reply with pleasure. I would send him my papers for comment and I would ask for his advice on other professional matters. He was always very supportive. In 1976, I sent him a paper that I was going to submit to the APA, and wanted his opinion of it before I did so. He wrote: Dear Nate, Thank you for sending me your papers. Since your main purpose, I take it, was, quite properly, to make sure that there is nothing “untoward” in the one you want to submit to [the] APA, I read it immediately. There is nothing wrong with it. Nothing at all. So go ahead [and submit it]. And, I am delighted to say, it is very nicely written. Very nicely indeed. There has been a “quantum jump” in your expository skill, which bodes well for the future. Best wishes to both of you. GB

In 1977 I published “Particulars, Positional Qualities and Individuation,” in Philosophy of Science.5 In that paper I argued against Bergmann’s and Hausman’s claim that bare particulars provide a deeper ontological ground of individuation than positional qualities. I sent him a copy of the paper and he wrote back: Dear Nate: Thank you for sending me your essay. Pleased as I am that it has been accepted, and pleasantly written as it is, it is yet such that—surprise!—I shall reply. Since it would no doubt be to our mutual advantage if your piece and mine appeared in the same issue, I would like to submit mine now, at the same time informing the editor that you have permitted me to do so. Will you give me this permission? Please let me know. Best wishes. GB 5

L. Nathan Oaklander, “Particulars, Positional Qualities and Individuation,” Philosophy of Science, 44 (1977), pp. 478-490.

294 He further showed me respect by giving the following title to his reply, “Some Comments on Professor Oaklander’s ‘Particulars, Positional Qualities and Individuation.’”6 Our correspondence continued and in 1978 I sent him my review of The Ontological Turn: Studies in the Philosophy of Gustav Bergmann subsequently published in Philosophia in 1981.7 GB was pleased with it, and I welcomed the opportunity to write it. His comments about the review were as follows: 6-22-78 Dear Nate, Thank you for sending me a copy of the Philosophia piece. I read it with growing pleasure and appreciation. I think it is very well done and, more importantly, done just the way this sort of thing, i.e., longish reviews, ought to be done, but not always are. Thus, I am confident, it will help you and help Iowa; nor will it hurt any of those you discuss. With best wishes to both of you, GB

The most significant reminiscence of mine with Bergmann, one that had a profound effect on me, had to do with his reaction to a paper I coauthored with Miracchi. It probably occurred in 1979 or 1980. The seeds were planted earlier. My good friend from graduate school, Silvano Mirrachi, wrote his dissertation with Addis on negative facts. Out of his dissertation came a paper he submitted to Philosophy of Science and although it was not accepted for publication, the editors were willing to reconsider. Silvano dropped out of philosophy to go to law school. He asked me if I would like to take over the paper and thus co-author it. I decided to do so, delving into the research on negative facts and reading classical papers as well as all of the Iowa ontologists’ papers on the topic. As I developed the 6

Gustav Bergmann, “Some Comments on Professor Oaklander’s ‘Particulars, Positional Qualities and Individuation’,” in Philosophy of Science, 44 (1977), pp. 491-493. 7 L. Nathan Oaklander, “Review of The Ontological Turn: Studies in the Philosophy of Gustav Bergmann,” (eds.), E.D. Klemke and M.S. Gram, in Philosophia, 9 (1981), pp. 445-453.

295 paper I came to realize that the essence of it was that the idea of an ideal language that is to serve as both a formal representation of a natural language and a metaphysical representation of the world broke down in the case of negative sentences. Simply because negation was ineliminable in a logically adequate language, it did not follow that negation was a feature of reality. I recall having a long conversation with Allaire at a central division APA in Cincinnati in the late 70’s where I tried to get him to clarify his disagreement with Bergmann, and explain the main point of his break with GB as reflected in his Philosophical Studies article “Relations: Recreational Remarks.”8 I learned much from those conversations and incorporated some of them into my paper. It is not my purpose here to defend the thesis that there are no negative facts, but to point out how it influenced my future relationship with GB. In 1980 I published, “Russell, Negative Facts and Ontology,” (with Miracchi), in Philosophy of Science.9 I do not remember if I sent GB a typescript or a reprint of the article, and I no longer have the letter he wrote in response to my sending it to him, but I do remember its tenor. He wrote me that we no longer had anything to talk about or, less stridently, that he had nothing to say to me. My distinct sense was that since I was strongly going against him he was thereby terminating our relationship. GB died in 1987 and a year later there was a memorial service in his honor. Of course I was notified and planned on attending, and clearly everyone thought that I would attend, but when the time came I decided not to go. I somehow deceived myself into believing that I was not going because I could not get away for work and family reasons, but I came to realize it was my anger at being rejected that made it impossible for me to attend. Unfortunately, it took me almost 25 years to come to terms with my relationship with GB, and only fairly recently have I come to see that GB was not rejecting me at all. As the years passed, my anger towards GB dissipated. In 2001 I coauthored an article defending bare particulars against an argument of Mi-

8

Edwin B. Allaire, “Relations: Recreational Remarks,” in Philosophical Studies, 78 (1978), pp. 81-89. 9 L. Nathan Oaklander and Silvano Miracchi, “Russell, Negative Facts and Ontology,” in Philosophy of Science, 47 (1980), pp. 434-455.

296 chael Loux’s10 and in 2004 I dedicated my book The Ontology of Time11 to the memory of Gustav Bergmann. I wanted to deliver a copy of it directly to Leola, but it didn’t appear until the end of July and I was not able to get to Iowa City before the end of the summer so I sent her a copy. I decided to read a paper at the Central States Philosophical Association meetings in Iowa City that year so that I could see Leola. For me it was also a way of making amends with GB. Linda and I went to her apartment for dinner. As soon as we came in and sat down Leola showed me that my book was front and center right next to the three volumes of Bergmann’s writings reprinted by Ontos Verlag and edited by Erwin Tegtmeier. We spent the next few hours talking mostly about GB. I kept on asking Leola questions about GB and she was delighted to answer them. She took us into her study where I saw a couple of pictures of GB. It was the first time I had “seen” him in over 30 years and it meant so much for me to do so. The visit helped me achieve a kind of closure in my relationship with GB, but more was to come. Before I explain why, I want to recount even now, almost twenty years after he died, the impact GB has had on me. As many of you may recall, in 1992, shortly after the posthumous publication of Bergmann’s last book, New Foundations of Ontology,12 there was a conference at Indiana University to discuss his work. I attended this conference and met one of the world’s foremost Bergmann scholars, Erwin Tegtmeier, there. In fall 2001, while I was at Cambridge I heard from Tegtmeier again when he asked me if I would contribute an article on time for Metaphysica.13 We corresponded that fall and then again last year regarding our both attending the Wittgenstein symposium in August of 2005. We met at the symposium and spent a lot of time together talking ontology. His colleague, Rafael Hüntelmann, editor at Ontos Verlag, offered to publish my next book. I did not give that much thought since I was not working on a book, but then 10

L. Nathan Oaklander and Alicia Rothstein, "Loux on Particulars: Concrete or Bare," in The Modern Schoolman, LXXVIII (2000), pp. 91-95. 11 L. Nathan Oaklander, The Ontology of Time. Amherst: Prometheus Books (2004). 12 Gustav Bergmann, New Foundations of Ontology, William Heald (ed.), (Madison: University of Wisconsin Press, (1992). 13 See L. Nathan Oaklander, “McTaggart’s Paradox Defended,” in Metaphysica: International Journal of Ontology and Metaphysics, 3 (2002): 11-25. Reprinted in L. Nathan Oaklander, The Ontology of Time. Amherst: Prometheus Books, (2004).

297 a couple of months later Tegtmeier asked me if my dissertation was my first book, Temporal Becoming and Temporal Relations,14 and I said “no.” It then occurred to me that I had never published anything from my dissertation or even attempted to. I told Tegtmeier that my dissertation was on The Ontology of C.D. Broad’s The Mind and Its Place in Nature, and it occurred to me that the editor of Ontos Verlag might be interested in it. I decided to send it to Tegtmeier and see what he thought. As it happens Tegtmeier is an editor of the Philosophical Analysis series at Ontos Verlag. He read the manuscript and thought highly of it, but I had reservations. GB had a heavy hand in the writing of my dissertation and so I was conflicted. I decided to publish it, as C.D. Broad’s Ontology of Mind,15 in part as a tribute to GB and also because I have come to make peace with him and this was a further sign of that for me. Interestingly, my philosophical work has started to come full circle. I recently wrote an invited entry on Negative Facts for the Routledge online encyclopedia of philosophy.16 I have been invited to write a paper on temporal states of affairs for Metaphysica, and my work in the philosophy of time has led me to the question of the order of temporal relations. In my own way, therefore, I think of myself as continuing the legacy of GB. I am very grateful to GB and my many other teachers at Iowa. GB was very fatherly toward me. His letters reflect that. He was also very supportive and encouraging. I now think that I misunderstood the meaning of his claim “that we have nothing more to talk about.” In the preparation of this paper I went back to my correspondence and found several short letters from GB. He always expressed support and fatherly concern. Certainly, if I was going to reject his method of philosophy then there would be nothing for us to talk about philosophically, but at a personal human level GB was always there for me. In 1982, two years after my negative facts paper he wrote how pleased he was with my promotion and the com14

L. Nathan Oaklander, Temporal Becoming and Temporal Relations: A Defense of a Russellian Theory of Time. Lantham: University Press of America, (1984). 15 L. Nathan Oaklander, C.D. Broad’s Ontology of Mind in the Philosophical Analysis series vol. 12. Frankfurt, Paris, Lancaster, New Brunswick: Ontos Verlag (2006). 16 L. Nathan Oaklander, “Negative Facts,” in Routledge Encyclopedia of Philosophy online www.rep.routledge.com (2005).

298 pletion of my first book. It was the last letter I received from him, or at least the last that I still possess. I regret that I did not attend the memorial service one year after his death, but I view this talk as my way of making amends. I am grateful for having the opportunity to reflect on some of my experiences with GB and to pay homage to him as a teacher, scholar and philosopher par excellence.

Afterword Read Bergmann!

I

know an ardent follower of Hilary Putnam who readily changed from so-called “internal realism” (a variant of Kantian idealism, corresponding roughly to what Kant himself misleadingly called “empirical realism”) to some kind of direct realism when Putnam did so. I asked him whether a turn to direct realism was not a reason to read Gustav Bergmann, who is the foremost direct realist of the 20th century. He reported that ten years ago Putnam advised him to study Bergmann’s Realism: A Critique of Brentano and Meinong, but that he gave up after a few days. Also, one can find no traces of Bergmann’s philosophy in Putnam’s writings. I suspect that he wanted his disciple to read the difficult book for him. Like a traditional philosopher, Bergmann has to be studied for years. It seems that no leading mainstream analytic philosopher did learn from Bergmann. Analytic philosophers are used to technical, mathematical complications and difficulties but not to complications and difficulties of thinking, especially of metaphysical thinking. As to metaphysics (including the metaphysics of knowledge) most analytical philosophers are not very far from popular philosophical views. What makes traditional philosophical texts difficult reading, what makes Bergmann’s texts difficult to read? Mainly, that these texts represent comprehensive and fundamental theories of the whole world, even though only with respect to its categorial structure. Metaphysical theories are rather complicated and they demand a rethinking of our most fundamental concepts. That fundamental rethinking is a hard task and requires many years of intense training. To understand Bergmann, one has to have completed such a training. But that is the same as with any science or any special subject. It needs a long training in physics to understand an advanced book on, say, quantum mechanics and a long training in psychology to understand an advanced book on the theories of learning.

300 Now, Bergmann’s papers and books in ontology mostly do not have an introductory character. However, his ontological books are very detailed and articulate. Thus they are right for advanced learning. They allow the reader to experience metaphysical thinking at work. There are very few contemporary texts that contain or are based on a comprehensive metaphysical theory. Mostly, there are only rudiments of such a theory. That is the case even with analytic philosophers who see themselves as ontologists such as David Lewis. The comprehensive theories applied in mainstream analytic philosophy are mathematical theories such as set theory and model theory, partly taken as ontologies, though they are mostly not relevant to the basic problems of ontology. Naturally, it also takes a long time to learn these theories, but analytic philosophers had to learn them anyway and thus they know them already. Moreover, most analytic philosophers are accustomed to relatively simple ontological assumptions that are not embedded in any comprehensive ontological theory. Analytical philosophers are not accustomed to complexity and sophistication in ontological matters. Analytical philosophy is traditionally anti-metaphysical. Therefore it lacks metaphysical knowledge and skill even after it has rehabilitated metaphysics. If a certain subject has been derided and neglected for a long time, naturally the skills it requires are no longer available. Because the horizon of contemporary analytic philosophy is with respect to philosophy rather narrow (the relation to other sciences is different), there is little chance to acquire metaphysical skills. The heroes of 20th-century analytical philosophy—Bertrand Russell, G. E. Moore, and Ludwig Wittgenstein—were not good examples, though all three started as metaphysicians. But they did not hold out. They soon abandoned metaphysics in favor of conceptual or linguistic and behavioral analysis. At least they did not develop their metaphysics further. Thus with respect to metaphysics, the three heroes furnish discouragement and not much metaphysical sophistication. Other analytical metaphysicians, such as Wilfrid Sellars, Roderick Chisholm, and P. F. Strawson were too strongly influenced by Kant to be good models for metaphysics. Unfortunately, analytic metaphysicians who don’t replace metaphysics by transcendental or conceptual analysis typically have to offer only rudiments of a metaphysics.

301 Only one analytical metaphysician stands out. Although he had been a member of the decidedly anti-metaphysical Vienna Circle, he started a thorough and brilliant reconstruction of metaphysics, stayed the course, and reached higher and higher levels of sophistication. I would claim that he was the only complete and the best metaphysician of the 20th century. Not only is there enough articulate and comprehensive metaphysics in Bergmann for him to serve as a model; he can play this role also because he has to offer first-class metaphysics, which is astonishing because there was so little competition in the analytic philosophy of his time. It seems that analytical philosophy, even mainstream analytical philosophy, has arrived today at the conclusion (at least at the feeling) that ontology is important and necessary. My advice and demand is to learn the metaphysical craftsmanship and not to dabble in the field of metaphysics as is customary in analytical philosophy. There is no better way today to achieve real metaphysical mastery than to read Bergman, to read him closely, to study him, to study particularly his two great ontological books, Realism: A Critique of Brentano and Meinong and New Foundations for Ontology. Naturally, there are not only impediments of proficiency and style that prevent the analytical philosopher to make use of Bergmann’s philosophy but also impediments of content. Several of the ontological doctrines and entities Bergmann advocates are unfashionable in analytical philosophy such as his ontological grounding of logic, his universal realism, his acknowledgment of facts including potential facts, and his advocacy of bare particulars. It is true that Bergmann broke with some of the orthodoxies of the empiricist tradition. But that is also true of mainstream analytical philosophy, which adapted partly to American pragmatism involving Kantian and even Hegelian ideas, while freely adopting Aristotelian elements such as dispositions and essences. The logical positivist Bergmann complained again and again against the deviations from the empiricist tradition. However, his main points were that the deviation and reintroduction of traditional conceptions was not compelling, that the respective problems could be solved in an empiricist way, and that the old empiricist arguments against those Aristotelian or Kantian or Hegelian conceptions had not been taken care of. That indicates that the views accepted in mainstream analytical philosophy are

302 not necessarily those supported by the best arguments. Hence, there is no good reason to reject ontological views just because they disagree with the current views in analytical philosophy. The swimmers in mainstream analytical philosophy need to be wakened from their “dogmatic slumber” (to use Kant’s phrase). Bergmann’s books and papers are just right for that role. They make for challenging reading because there is hardly any basic tenet of mainstream analytical philosophy that Bergmann does not reject with good reasons. Erwin Tegtmeier

Bibliography of the Writings of Gustav Bergmann 1928 Zur Axiomatik der Elementargeometrie. Weiner Akademie Der Wissenschaften, 26, (1928): 1-3.

1929 Über eine mit den Hypertorsen verwandte Flachenklasse. Monatshefte für Mathematik und Physik, 36, (1929): 259-268. Zur Axiomatik der Elementargeometrie. Monatshefte für Mathematik und Physik, 36, (1929): 269-284.

1930 Ebenen und Bewegungsgruppen in Riemannschen Räumen. Jahresbericht der Deutschen Mathematiker Vereiningung, 39, (1930): 54-55. (with E. Lukács) Ebenen und Bewegungsgruppen in Riemannschen Räumen. Monatshefte für Mathematik und Physik, 37, (1930): 303-324. (with E. Lukács)

1931 Zwei Bemerkungen zur abstracten und kombinatorischen Topologie. Monatshefte für Mathematik und Physik, 38, (1931): 245-256.

1932 Zur algebraisch-axiomatischen Begründung der Topologie. schrift, 35, (1932): 502-511.

Mathematische Zeit-

1935 Zur analytischen Theorie literarischer Wertmasstäbe. Imago, 21, (1935): 498-504.

1940 On Physicalistic Models of Non-Physical Terms. Philosophy of Science, 7, (1940): 151-158. On Some Methodological Problems of Psychology. Philosophy of Science, 7, (1940): 205-219.

304 The Subject Matter of Psychology. Philosophy of Science, 7, (1940): 415-433.

1941 The Logic of Probability. American Journal of Physics, 9, (1941): 263-272. Operationism and Theory in Psychology. Psychological Review, 48, (1941): 1-14. (with K. Spence)

1942 An Empiricist Schema of the Psychophysical Problem. Philosophy of Science, 9, (1942): 72-91. Syntactical Analysis of the Class Calculus. Philosophy of Science, 9, (1942): 227-232. Remarks Concerning the Epistemology of Scientific Empiricism. Philosophy of Science, 9, (1942): 283-293. The Indexical and Presentative Function of Signs. Philosophy of Science, 9, (1942): 372-374.

1943 Notes on Identity. Philosophy of Science, 10, (1943): 163-166. Psychoanalysis and Experimental Psychology: A Review from the Standpoint of Scientific Empiricism. Mind, 52, (1943): 122-140. Outline of an Empiricist Philosophy of Physics. American Journal of Physics, 11, (1943): 248-258 and 335-342.

1944 The Logic of Psychophysical Measurement. Psychological Review, 51, (1944): 1-24. (with K. Spence) An Empiricist’s System of the Sciences. Scientific Monthly, 59, (1944): 140-148. Holism, Historicism, and Emergence. Philosophy of Science, 11, (1944): 209-221. Pure Semantics, Sentences, and Propositions. Mind, 53, (1944): 238-257. Notes on Identity. Philosophy of Science, 11, (1944): 123-124.

1945 The Formalism in Kelsen’s Pure Theory of Law. Ethics, 55, (1945): 110-130. (with L. Zerby)

305 A Positivistic Metaphysics of Consciousness. Mind, 54, (1945): 193-226. Frequencies, Probabilities, and Positivism. Philosophy and Phenomenological Research, 6, (1945): 26-44. Review of Ebenstein’s The Pure Theory of Law. Ethics, 56, (1945): 71-72. Review of Ortega y Gasset’s The Mission of the University. Journal of American History (formerly Mississippi Valley Historical Review), 32, (1945): 266-267.

1946 Remarks on Realism. Philosophy of Science, 13, (1946): 261-273. (also in The Metaphysics of Logical Positivism, 1954) Some Comments on Carnap’s Logic of Induction. Philosophy of Science, 13, (1946): 71-78. Review of Cohen’s The Faith of a Liberal. Iowa Law Review, 32, (1946): 178-181.

1947 The Logic of Quanta. American Journal of Physics, 15, (1947): 397-408 and 497-508. Russell on Particulars. Philosophical Review, 56, (1947): 59-72. (also in The Metaphysics of Logical Positivism, 1954) Undefined Descriptive Predicates. Philosophy and Phenomenological Research, 8, (1947): 55-82. Sense Data, Linguistic Conventions, and Existence. Philosophy of Science, 14, (1947): 152-163. (also in The Metaphysics of Logical Positivism, 1954) Review of Kelsen’s General Theory of Law and State. Ethics, 57, (1947): 213-215.

1948 Conditions for an Extensional Elementaristic Language. Analysis, 8, (1948): 44-47. (also in The Metaphysics of Logical Positivism, 1954) Concerning Carnap’s Definition of “Extensional” and “Intensional”. (1948): 494-495.

Mind, 57,

Contextual Definitions in Nonextensional Languages. Journal of Symbolic Logic, 13, (1948): 140. Descriptions in Nonextensional Contexts. Philosophy of Science, 15, (1948): 353-355. Review of Köhler’s Gestalt Psychology. Psychological Bulletin, 45, (1948): 351-355.

306 1949 The Finite Representations of S5. Methodos, 1, (1949): 217-219. On Non-Perceptual Intuition. Philosophy and Phenomenological Research, 10, (1949): 263-264. (also in The Metaphysics of Logical Positivism, 1954) Professor Ayer’s Analysis of Knowing. Analysis, 9, (1949): 98-106. (also in The Metaphysics of Logical Positivism, 1954) A Syntactical Characterization of S5. Journal of Symbolic Logic, 14, (1949): 173-174. Two Criteria for an Ideal Language. Philosophy of Science, 16, (1949): 71-74. Review of Russell’s Human Knowledge: Its Scope and Limits. Physics Today, 2, (1949): 27-28.

1950 Logical Positivism. A History of Philosophical Systems, edited by Vergilius Ferm, Philosophical Library, (1950): 471-482. (also in The Metaphysics of Logical Positivism, 1954) Semantics. A History of Philosophical Systems, edited by Vergilius Ferm, Philosophical Library, (1950): 483-492. (also in The Metaphysics of Logical Positivism, 1954) A Study of Simple Learning Under Irrelevant Motivational-Reward Conditions. Journal of Experimental Psychology, 40, (1950): 539-551. (with K. Spence and R. Lippitt) A Note on Ontology. Philosophical Studies, 1, (1950): 89-92. (also in The Metaphysics of Logical Positivism, 1954) Review of Born’s Natural Philosophy of Cause and Chance. Philosophy of Science, 17, (1950): 196-199.

1951 Concerning the Definition of Classes. Mind, 60, (1951): 95-96. Ideology. Ethics, 61, (1951): 205-218. (also in The Metaphysics of Logical Positivism, 1954) The Logic of Psychological Concepts. Philosophy of Science, 18, (1951): 92-110. Comments on Professor Hempel’s “The Concept of Cognitive Significance”. Daedalus (formerly Proceedings of the American Academy of Arts and Sciences), 80, (1951): 78-86. (also in The Metaphysics of Logical Positivism, 1954)

307 Logical Atomism, Elementarism, and the Analysis of Value. Philosophical Studies, 2, (1951): 85-92. (also in The Metaphysics of Logical Positivism, 1954) Comments on Storer’s Definition of “Soluble”. Analysis, 12, (1951): 44-48. Review of Schilpp’s (editor) Albert Einstein: Philosopher-Scientist. Philosophical Review, 60, (1951): 269-274. Review of Kraft’s Der Weiner Kreis and Juhos’s Die Erkenntnis und Ihre Leistung. Philosophy and Phenomenological Research, 12, (1951): 139-143. Review of Lasswell’s and Kaplan’s Power and Society: A Framework for Political Inquiry. Ethics, 62, (1951): 64-65.

1952 Two Types of Linguistic Philosophy. The Review of Metaphysics, 5, (1952): 417-438. (also in The Metaphysics of Logical Positivism, 1954) The Problem of Relations in Classical Psychology. Philosophical Quarterly, (1952): 140-152. (also in The Metaphysics of Logical Positivism, 1954) A Criterion of Necessity. The Review of Metaphysics, 6, (1952): 128-129. Two Cornerstones of Empiricism. Synthese, 8, (1952): 435-452. (also in The Metaphysics of Logical Positivism, 1954) Multiplicative Closures. Portugaliae Mathematica, 11, (1952): 169-172. Review of Brunswick’s The Conceptual Framework of Psychology. Psychological Bulletin, 49, (1952): 654-656.

1953 Theoretical Psychology. Annual Review of Psychology, 4, (1953): 435-458. The Identity of Indiscernibles and the Formalist Definition of “Identity”. Mind, 62, (1953): 75-79. (also in The Metaphysics of Logical Positivism, 1954) Logical Positivism, Language, and the Reconstruction of Metaphysics. Revista Critica di Storia Della Filosofia, 8, (1953): 453-481. (also in The Metaphysics of Logical Positivism, 1954) Review of Wisdom’s Other Minds. Philosophy and Phenomenological Research, 14, (1953): 112-114.

308 1954 Particularity and the New Nominalism. Methodos, 6, (1954): 131-147. (also in Meaning and Existence, 1959) Sense and Nonsense in Operationism. Scientific Monthly, 79, (1954): 210-214. (reprinted in The Validation of Scientific Theories, edited by P. Frank, The Beacon Press, (1957): 41-52) Some Remarks on the Ontology of Ockham. Philosophical Review, 63, (1954): 560571. (also in Meaning and Existence, 1959) Review of Goetz’s Die Entstehung der Ordnung. Philosophy and Phenomenological Research, 15, (1954): 287-288. Bodies, Minds, and Acts. (first published in The Metaphysics of Logical Positivism, 1954) The Metaphysics of Logical Positivism, Longmans, Green and Co., (1954). (reprinted by The University of Wisconsin Press, 1967)

1955 Professor Quine on Analyticity. Mind, 64, (1955): 254-258. (also in Meaning and Existence, 1959) Reduction. Current Trends in Psychology and the Behavioral Sciences, University of Pittsburgh Press, (1955): 59-81. Intentionality. Archivio di Filosofia, Semantica, (1955): 177-216. (also in Meaning and Existence, 1959) Dispositional Properties and Dispositions. Philosophical Studies, 6, (1955): 77-80. Dell’ Ideologia. Occidente, 11, (1955): 519-535. (translation of “Ideology,” 1951) Review of Frenkel-Brunswik’s Psychoanalysis and the Unity of Science. Journal of Philosophy, 52, (1955): 692-695. Review of Nagel’s Sovereign Reason and Other Studies in the Philosophy of Science. Philosophy and Phenomenological Research, 16, (1955): 128-130. Review of Stegmüller’s Metaphysik, Wissenschaft, Skepsis. Philosophical Review, 64, (1955): 655-667.

309 1956 The Contribution of John B. Watson. Psychological Review, 63, (1956): 265-276. (reprinted in Theories of the Mind, edited by J. Scher, The Free Press, (1962): 674-688) Russell’s Examination of Leibniz Examined. Philosophy of Science, 23, (1956): 175203. (also in Meaning and Existence, 1959) The Representations of S5. Journal of Symbolic Logic, 21, (1956): 257-260. The Logic of Measurement. Proceedings of the Sixth Hydraulics Conference, State University of Iowa Publications, (1956): 19-33. Some Remarks on the Philosophy of Malebranche. The Review of Metaphysics, 10, (1956): 207-226. (also in Meaning and Existence, 1959) Review of Blaha’s Logische Wirklichkeitsstruktur und Personaler Seinsgrund. losophical Review, 65, (1956): 265-267.

Phi-

1957 Philosophy of Science. The University of Wisconsin Press, (1957). (reprinted by Greenwood Press, 1977) Concepts. Philosophical Studies, 8, (1957): 19-27; with Herbert Hochberg. (also in Meaning and Existence, 1959) Elementarism. Philosophy and Phenomenological Research, 18, (1957): 107-114. (also in Meaning and Existence, 1959) The Revolt Against Logical Atomism. Philosophical Quarterly, 7, (1957): 323-339, and 8, (1958): 1-13. (also in Meaning and Existence, 1959)

1958 Some Reflections on Time. Archivio di Filosofia, Il Tempo, (1958): 49-82. (also in Meaning and Existence, 1959) Alcune Riflessioni sul Tempo. Archivio di Filosofia, Il Tempo, (1958): 83-113. (translation of “Some Reflections on Time,” 1958) Frege’s Hidden Nominalism. Philosophical Review, 67, (1958): 437-459. (also in Meaning and Existence, 1959) Analyticity. Theoria, 24, (1958): 71-93. (also in Meaning and Existence, 1959) Individuals. Philosophical Studies, 9, (1958): 78-85. (also in Meaning and Existence, 1959)

310 Sameness, Meaning, and Identity. International Congress of Philosophy: Proceedings, 4, (1958): 19-27. (also in Meaning and Existence, 1959) Il Contributo di John B. Watson. Rivista di Psicologia, 52, (1958): 311-325. (translation of “The Contribution of John B. Watson,” 1956) Review of Kaila’s Terminalkausalität in der Atomdynamik. Philosophical Review, 67, (1958): 424-426.

1959 Meaning and Existence. The University of Wisconsin Press, (1959).

1960 Ineffability, Ontology, and Method. Philosophical Review, 69, (1960): 18-40. (also in Logic and Reality, 1964) Duration and the Specious Present. Philosophy of Science, 27, (1960): 39-47. (also in Logic and Reality, 1964) Acts. The Indian Journal of Philosophy, 2, (1960): 1-30 and 96-117. (also in Logic and Reality, 1964) Dell' Atto. Rivista di Filosofia, 51, (1960): 3-51. (translation of “Acts,” 1960) The Philosophical Significance of Modal Logic. Mind, 69, (1960): 466-485. Strawson's Ontology. Journal of Philosophy, 57, (1960): 601-622. (also in Logic and Reality, 1964) The Ontology of Edmund Husserl. Methodos, 12, (1960): 359-392. (also in Logic and Reality, 1964) Review of Korner’s (editor) Observation and Interpretation: A Symposium of Philosophers and Physicists. Philosophical Review, 69, (1960): 267-270.

1961 Physics and Ontology. Philosophy of Science, 28, (1961): 1-14. (also in Logic and Reality, 1964) La Gloria e la Miseria di Ludwig Wittgenstein. Rivista di Filosofia, 52, (1961): 387406. (also as: The Glory and the Misery of Ludwig Wittgenstein, in Logic and Reality, 1964) Filosofia de La Ciencia. Editorial Tecnos (Madrid), 1961. (translation of Philosophy of Science, 1957)

311 1962 Generality and Existence. Theoria, 28, (1962): 1-26. (also in Logic and Reality, 1964) Meaning and Ontology. Inquiry, 5, (1962): 116-142. (also as two essays: (1) Meaning (2) Inclusion, Exemplification, and Inherence in G. E. Moore, in Logic and Reality, 1964) Purpose, Function, Scientific Explanation. Acta Sociologica, 5, (1962): 225-238.

1963 Alternative Ontologiche. Giornale Critico dell Filosofia Italiana, 17, (1963): 337405. (also as: Ontological Alternatives, in Logic and Reality, 1964) Stenius on the Tractatus: A Special Review. Theoria, 29, (1963): 176-204. (also in Logic and Reality, 1964)

1964 Synthetic A Priori. (first published in Logic and Reality, 1964) Realistic Postscript. (first published in Logic and Reality, 1964) Logic and Reality. The University of Wisconsin Press, 1964.

1967 Realism: A Critique of Brentano and Meinong. The University of Wisconsin Press, 1967.

1968 Diversity. American Philosophical Association: Proceedings and Addresses, 41, (1968): 21-34.

1977 Some Comments on Professor Oaklander’s “Particulars, Positional Qualities, and Individuation.” Philosophy of Science, 44, (1977): 491-493.

1978 Esbozo de un Inventario Ontologico. Teorema, 8, (1978): 93-105. (translation of “Sketch of an Ontological Inventory,” 1979)

312 1979 Sketch of an Ontological Inventory. JBSP: The Journal of the British Society for Phenomenology, 10, (1979): 3-8.

1981 Notes on Ontology. Noûs, 15, (1981): 131-154. Notes on the Ontology of Minds. Midwest Studies in Philosophy, 6, (1981): 189-213.

1988 Erinnerungen an den Wiener Kreis: Brief an Otto Neurath. Vertriebene Vernunft II: Emigration und Exil Österreichischer Wissenschaft, edited by F. Stadler, Jugend und Volk, (1988): 171-180.

1992 New Foundations of Ontology. The University of Wisconsin Press, 1992. (edited by William Heald)

1993 Memories of the Vienna Circle: Letter to Otto Neurath: 1938. (translated and edited by William Heald and Guenter Zoeller) Scientific Philosophy: Origins and Developments, edited by F. Stadler, Kluwer Academic Publishers, (1993): 193-208. (translation of the item of 1988)

2003 Collected Works, Vol. I: Selected Papers I. Ontos Verlag, (2003). (edited by Erwin Tegtmeier) Collected Works, Vol. II: Selected Papers II. Ontos Verlag, (2003). (edited by Erwin Tegtmeier)

2004 Collected Works, Vol. III: Realism: A Critique of Brentano and Meinong. Ontos Verlag, (2004). (edited by Erwin Tegtmeier)

This bibliography of the writings of Gustav Bergmann was compiled by me and is, to the best of my knowledge, accurate and complete except that it does not include all of the anthologizations of some of his essays. Laird Addis

PhilosophischeAnalyse PhilosophicalAnalysis 1 Herbert Hochberg Russell, Moore and Wittgenstein The Revival of Realism

8 Rafael Hüntelmann Existenz und Modalität Eine Studie zur Analytischen Modalontologie

2

9 Andreas Bächli / Klaus Petrus Monism

ISBN 3-937202-00-5 334 pp., Hardcover € 94,00

Heinrich Ganthaler Das Recht auf Leben in der Medizin Eine moralphilosophische Untersuchung

ISBN 3-937202-01-3 167 pp., Hardcover € 58,00

3 Ludger Jansen Tun und Können Ein systematischer Kommentar zu Aristoteles’ Theorie der Vermögen im neunten Buch der „Metaphysik“ ISBN 3-937202-02-1 302 pp., Hardcover € 70,00

4 Manuel Bremer Der Sinn des Lebens Ein Beitrag zur Analytischen Religionsphilosophie ISBN 3-937202-03-X 134 pp., Hardcover € 58,00

5 Georg Peter Analytische Ästhetik Eine Untersuchung zu Nelson Goodman und zur literarischen Parodie ISBN 3-937202-04-8, 332 pp. Hardcover € 94,00

6 Wolfram Hinzen / Hans Rott Belief and Meaning Essays at the Interface ISBN 3-937202-05-6 250 pp., Hardcover € 58,00

7 Hans Günther Ruß Empirisches Wissen und Moralkonstruktion Eine Untersuchung zur Möglichkeit von Brückenprinzipien in der Natur- und Bioethik ISBN 3-937202-06-4 208 pp., Hardcover € 58,00

ISBN 3-937202-07-2 189 pp., Hardcover € 58,00

ISBN 3-937202-19-6 340 pp., Hardcover € 70,00

10 Maria Elisabeth Reicher Referenz, Quantifikation und ontologische Festlegung ISBN 3-937202-39-0 ca. 300 pp., Hardcover € 89,00

11 Herbert Hochberg / Kevin Mulligan Relations and Predicates ISBN 3-937202-51-X 250 pp., Hardcover € 74,00

12 L. Nathan Oaklander C. D. Broad's Ontology of Mind ISBN 3-937202-97-8 105 pp., Hardcover € 39,00

13 Uwe Meixner The Theory of Ontic Modalities ISBN 3-938793-11-2 374 pages, Hardcover,€ 79,00

14 Donald W. Mertz Realist Instance Ontology and its Logic ISBN 3-938793-33-3 252 pp., Hardcover, EUR 79,00

15 N. Psarros / K. Schulte-Ostermann (Eds.) Facets of Sociality ISBN 3-938793-39-2 370 pp., Hardcover, EUR 98,00

16 Markus Schrenk The Metaphysics of Ceteris Paribus Laws ISBN 13: 978-3-938793-42-8 192pp, Hardcover, EUR 79,00

EditedBy • HerbertHochberg • RafaelHüntelmann ChristianKanzian • RichardSchantz • ErwinTegtmeier

PhilosophischeAnalyse PhilosophicalAnalysis 17 Nicholas Rescher Interpreting Philosophy The Elements of Philosophical Hermeneutics ISBN 978-3-938793-44-2 190pp., Hardcover € 89,00

18 Jean-Maurice Monnoyer(Ed.) Metaphysics and Truthmakers ISBN 978-3-938793-32-9 337 pp., Hardcover € 98,00

19 Fred Wilson Acquaintance, Ontology, and Knowledge Collected Essays in Ontology ISBN 978-3-938793-58-9 XX, 726., Hardcover, EUR 159,00

EditedBy • HerbertHochberg • RafaelHüntelmann ChristianKanzian • RichardSchantz • ErwinTegtmeier