Metaphysics and Ontology Without Myths [1 ed.] 9781443868273, 9781443863919

Citation preview

Metaphysics and Ontology Without Myths

Metaphysics and Ontology Without Myths

Edited by

Fabio Bacchini, Stefano Caputo and Massimo Dell'Utri

Metaphysics and Ontology Without Myths, Edited by Fabio Bacchini, Stefano Caputo and Massimo Dell'Utri This book first published 2014 Cambridge Scholars Publishing 12 Back Chapman Street, Newcastle upon Tyne, NE6 2XX, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2014 by Fabio Bacchini, Stefano Caputo, Massimo Dell'Utri and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-6391-2, ISBN (13): 978-1-4438-6391-9

TABLE OF CONTENTS

Introductory Note ...................................................................................... vii Fabio Bacchini, Stefano Caputo and Massimo Dell’Utri Chapter One ................................................................................................. 1 The Myth of the Distinction between Ontology and Metaphysics Andrea Bottani Chapter Two .............................................................................................. 16 Realism in the Desert Achille Varzi Chapter Three ............................................................................................ 32 Realism and Truth from a Quinean Point of View Antonio Rainone Chapter Four .............................................................................................. 48 Why Antirealists Can’t Explain Success Mario Alai Chapter Five .............................................................................................. 67 The Content Challenge to Nominalism about Mathematics Matteo Plebani Chapter Six ................................................................................................ 83 Is There Such a Thing as Metaphysical Knowledge? Claudine Tiercelin Chapter Seven.......................................................................................... 101 The Identity of Indiscernibles and the Principle of No Co-location Roberto Casati and Giuliano Torrengo Chapter Eight ........................................................................................... 110 New Realism, Documentality and the Emergence of Normativity Maurizio Ferraris

vi

Table of Contents

Chapter Nine............................................................................................ 125 Why Society Is Not Made of Documents Richard Davies Chapter Ten ............................................................................................. 139 How to Incorporate (with) Documents Petar Bojaniü Chapter Eleven ........................................................................................ 153 To Be Realist With History: What Really Are Historical Documents? Stefano Vaselli Abstracts .................................................................................................. 170 Contributors ............................................................................................. 175 Index of Names........................................................................................ 178



INTRODUCTORY NOTE

Metaphysics and ontology feature among the traditional and fundamental concerns of philosophers. Gaining a picture of the world—whether this may correspond precisely to its physical appearance, or whether there may be much more beyond—and the kind of objects that do exist out there, is for most philosophers (past and present) a preliminary aim upon which other theoretical activities depend. In fact, it seems that sound conclusions on topics relevant to ethics, aesthetics, psychology, common and scientific knowledge etc. can be achieved only after one has been given a picture of that sort. This brief introduction is hardly the appropriate place to summarise the main results obtained over centuries of close philosophical scrutiny, since a great deal of water has flowed under the metaphysical and ontological bridges. What we would like to stress though is that from time to time the tribunal of history has managed to put its finger on some flawed conclusions. But “flawed” in which sense? And who is to tell? To address the second question first, it goes without saying that a judgment of the plausibility of a metaphysical and ontological picture is inescapably linked to the particular historical and philosophical context in which the judgment itself has been raised, so that in principle we can obtain as many different judgments as there are contexts of this kind. Yet, relativistic worries aside, what one wants is an objective verdict on this and other matters—common sense requires it, and common sense presupposes that knowledge (metaphysical knowledge in our case) is not only possible but capable of making progress. And this is actually the case, to put sceptical worries aside too. It is hence from the point of view of the best metaphysical stance we are able from time to time to adopt (the one which inherits piecemeal the outcomes of a historically long discussion, separating the wheat from the chaff) that we can tell whether or not a given metaphysical and ontological picture is flawed. But, again, in which sense? Metaphysical knowledge is a complex thing. However, the difficulty to say what it amounts to notwithstanding, the idea according to which both scientific knowledge and common sense have to play a role in it, even if not a necessarily crucial one, is by now quite widespread. One can therefore consider something to be a flaw in a metaphysical and ontological picture if



viii

Introductory Note

it clashes with both scientific knowledge and common sense. To take a time-worn example, who is willing nowadays to embrace wholeheartedly the abstract world of ideas and forms Plato envisaged all those centuries ago? Who would now subscribe to his own sharp distinction between appearance and reality, where the metaphysical and ontological privilege is conferred only to that which exists in the abstract world of perfect unchanging models of the things one can come across in the sphere of appearance (to the great detriment of the latter)? The picture he gave us is nothing but a myth—an account which is too far away from what common sense and science could accept, too detached from the usual ways of conducting a rational discussion. Pictures of this kind appear to be supported by nothing but dogmas, i.e. uncompromising principles taken as true without any previous critical analysis. And Plato has no shortage of company. Think of essentialism in biology as regards the notion of species. There was a time in which scholars used to refer to species as eternal entities endowed with essential features. This conviction was to be unmasked as a myth by biological research, which pointed out how evolution—involving phenomena such as splitting, budding, fusing, etc.—makes the question of species a real problem for biology as well as philosophy. Granted, cases like these which make room for talk of advances in metaphysics are indeed few. Most of the time what we encounter are different parties mutually opposing one another—implicitly accusing each other of offering “mythical” views, to continue with our metaphor. Take for example any metaphysical and ontological picture according to which there exists a fixed predefined structure of reality, a sort of bedrock made up of “things in themselves” and susceptible to a single overall and complete description. To some this may not seem to be what physics tells us we can have—physics, they maintain, tends to favour a view according to which reality has no predefined structure and is describable in many equivalent correct but incompatible ways without making any concession to relativism. But, is this true? Again, depending on the general philosophical stance we embrace, we may feel inclined to see (at least part of) a given picture as unjustified and think that it pertains less to metaphysics than myth. Thus far we have not given a definition of either metaphysics or ontology. Of course, for reasons of space we cannot embark on a painstaking analysis of metaphysics, ontology and their relationship—you can find more on this in the following pages. But, interestingly enough, in an attempt to define them we may discover that there is room for talk of myth in this connection too. Take for instance the following definition:



Metaphysics and Ontology Without Myths

ix

“Ontology seeks to say what there is, whereas metaphysics to say what it is—what its fundamental nature is”. A statement along these lines is at the centre of the debate among philosophers of different bents, but it is far from uncontroversial that that statement amounts to a good definition, or that it represents a good basis on which to try one. Moreover, it could be somewhat startling to realize that there is little agreement on whether metaphysics is prior to ontology or vice versa, or that the very distinction between the two can be called into question. In cases like these one might say that if somebody keeps standing by his or her point in spite of the best counter-arguments available at the moment, then again what we get is something that pertains less to metaphysics than myth. This and other issues revolving around metaphysics and ontology are tackled in the essays in this volume, which aim to approach a secular debate in fresh and original ways. Questions such as the possibility of a clear-cut distinction between metaphysics and ontology (Andrea Bottani), the legacy of one of the great fathers of contemporary ontology, Willard Van Orman Quine (Achille Varzi and Antonio Rainone), how to account for the success of novel predictions in science (Mario Alai), the nature of mathematical objects and the viability of nominalism (Matteo Plebani), the possibility of genuine metaphysical knowledge (Claudine Tiercelin), the principle of identity of indiscernibles and the role of intuitions (Roberto Casati and Giuliano Torrengo), the character of and the puzzles posed by social reality (Maurizio Ferraris, Richard Davies, Petar Bojaniü, Stefano Vaselli) are here addressed lucidly and probed deeply, providing the necessary tools for clearing the field of unpalatable metaphysical and ontological items. Sassari (Italy), March 3, 2014 Fabio Bacchini Stefano Caputo Massimo Dell’Utri



CHAPTER ONE THE MYTH OF THE DISTINCTION BETWEEN ONTOLOGY AND METAPHYSICS ANDREA BOTTANI

1. Ontology and metaphysics It is a widely shared idea that no systematic account of the most fundamental aspects of reality can be complete unless it includes both an account of what things there are (an sit) and an account of their fundamental nature (quid sit). What Aristotle originally called “first philosophy”, whatever it may be, can be nothing but a complex of these two components. Here are some ways of stating the idea: One very important part of metaphysics has to do with what there is, with what exists. This part of metaphysics is called ontology. Ontology, that is, is that part of metaphysics that deals with metaphysical sentences of the form “An X exists”, and “There are Y’s” (van Inwagen 1998, 16). According to a certain, familiar way of dividing up the business of philosophy, […] ontology is concerned with the question of what there is (a task that is often identified with that of drafting a “complete inventory” of the universe) whereas metaphysics is concerned with the question of what it is (i.e., with the task of specifying the “ultimate nature” of the items included in the inventory) (Varzi 2009, 1). We might wonder whether there could be anything to metaphysics other than ontology, and indeed ontology does seem to be a large part of the metaphysical enterprise. But metaphysics is concerned not just with what is, but also with the way that it is. Objects do not merely exist: they have certain features. […] Simply to list the things that are […] does not capture the way they are (Le Poidevin 2009, xix).

Those who distinguish in this way the account of what there is from the account of what it is, ordinarily conceive of the two accounts as

2

Chapter One

somehow independent, in the twofold sense that neither entails the other and that it is possible to give at least one of them without having already given the other. But they disagree as to which has priority over the other. Some think that the former comes first and the latter later (one can only start saying what there is, only later one can wonder what it is), others think that it is the latter that “comes first” and the former later (one can only start asking in abstracto what things might be, only later one can wonder which of them there really are). Those who accept the distinction also diverge over terminology, for the word “metaphysics” has sometimes been used to name the theory of what there is (Ingarden 1964, who calls “ontology” the theory of what it is), sometimes the theory of what it is (Varzi 2009, who calls “ontology” the theory of what there is—see above) and sometimes a general account that includes both (van Inwagen 1998, Le Poidevin 2009, see above). I shall say nothing of this lexical muddle here, except that I shall conform to Varzi’s usage and call “ontology” the theory of what there is and “metaphysics” the theory of what it is.1 The distinction between ontology and metaphysics is closely related to Quine’s distinction between ontology and ideology of a language or theory in general (Quine 1951, 1983). According to Quine, the ontology of a theory is the set of things that the theory assumes as existing, while its ideology is the set of properties that can be expressed by its predicates. So one might simply treat metaphysics as the ideology of first philosophy. But metaphysics is ordinarily conceived of as a theory of what things are (viz. of their ultimate nature), and in ideology there is much more than this (viz. a theory of how they are). If one is willing to distinguish what something is from how it is, metaphysics is rather a proper part of the ideology of first philosophy. In what follows, I shall argue that the distinction between ontology and metaphysics is flawed. It is just a myth and should be given up. As I shall show, this necessarily follows from the following three assumptions: 1) ontology is a systematic taxonomy of what fundamentally exists, not a disordered list of existential generalizations; 2) no apparent disagreement about the nature of the entities belonging to some taxonomical category counts as a real disagreement about the ultimate nature of something, provided that there is total agreement about the nature of what belongs to the highest categories, those that restrict no other category; 3) every disagreement about the nature of what belongs to the highest categories is eo ipso a disagreement about what there is. Since none of 1)-3) is obviously true, I shall separately argue for each of them (not in this order). The rejection of the distinction of ontology and metaphysics has some bearing on the very idea of ontology. One important consequence is that it

The Myth of the Distinction between Ontology and Metaphysics

3

is only in a very weak sense that ontology says what there is. Ontology reorganizes ontological commitments that come from common sense and sciences. It cannot start without an external preliminary input of independently established existential commitments.

2. What there is, what it is and how it is There are natural reasons to be prima facie suspicious of the distinction. One is that it can make no sense to distinguish within physics (or biology, or set theory) a discourse aimed at saying what physical (or biological or set-theoretical) objects exist from another discourse aimed at saying what they are. Generally, assumptions of both types occur fairly mixed in the same discourse, and very often they are embedded in the meaning of the same sentences. Why on earth should first philosophy be regarded as an exception? But this line of attack misinterprets the distinction. In Quine’s approach, ideology and ontology are not conceived of as two distinct discourses, theories or accounts, but rather as two different aspects of one and the same discourse, theory or account. Ontology and metaphysics should be charitably interpreted in the same way, as different kinds of information mixed up in the same discourse, theory or account—like different colour threads in the same cloth. But the two kinds of information must be logically independent, however mixed up in the same discourse, otherwise the distinction would collapse. Can information of one kind remain unchanged even when information of the other kind evolves? Those who distinguish ontology and metaphysics ordinarily think that people can agree on what there is even when they disagree on its ultimate nature (see again Varzi 2009, Le Poidevin 2009, quoted above). If so, one can stick to a fixed theory of what exists while changing one’s ideas on its ultimate nature. And one can meaningfully assert that something exists while ignoring what it is that she is asserting the existence of. But this is strange: how could the assertion that entities of a certain kind exist have any meaning in the absence of any information about what these entities are? The worry can perhaps be kept under control by embracing a direct theory of reference. If Kripke 1980 is right, one can succeed in referring to tigers even if one knows nothing about tigers (this is the case if someone who has never seen a tiger and has never heard anybody naming tigers hears for the first time an utterance of “in India there are tigers” and asks “what are tigers?”). If one is able to refer to tigers, one can utter

4

Chapter One

meaningfully the sentence “tigers exist”. Therefore, a fortiori, the bare idea that tigers exist entails or presupposes nothing about what tigers are. This might seem to miss the point. If one knows nothing at all about tigers, indeed, one can hardly express or understand something informative by saying or hearing that tigers exists, even if she succeeds in saying or hearing that. And ontology, whatever it may be, has to convey information about what there is. Therefore, nothing ontologically informative can be said by uttering the sentence “tigers exist” unless one knows at least something about tigers. But one might know something about tigers, it may be suggested, without knowing what they are. If so, one can tell something ontologically informative by saying that tigers exist even if one is ignorant of what are tigers, provided one knows at least something about how tigers look. We started with a twofold distinction between what there is and what it is. Now we have a threefold distinction between what there is, how it is and what it is, the idea being that we must distinguish between how something is and what it is in order to legitimate the distinction between what there is and what it is. But the distinction between what something is and how it is can itself be put in question. The notion of what something is is traditionally associated with the idea that objects have a nature, and what they are consists in the nature they have. Very often, natures are treated as essences and conceived of as common to all things of the same kind (for example, to all tigers). But many properties of objects do not depend on their nature and information about them is merely about how objects are, not about what they are. So, “tigers are animals” says what are tigers, but “tigers are striped” merely says how they are. It is a widely shared idea that natures can be explained in terms of conditions of identity specifying what it is for objects of some sort to be the same (a trite example concerning sets treats them as identical just in case they have the same members, another concerning material objects treats them as identical just in case they have the same parts).2 Anti-essentialist philosophers deny that common nouns (either sortal terms like “animal” or mass nouns like “gold”) have any special role in specifying the nature of what exists. They think that there are plenty of objects that do not fall within the extension of any common nouns—the material contents of certain spacetime regions can easily be, for example, partly gold and partly water, and so neither (entirely) gold nor (entirely) water. And they may think that even uniformly golden contents of spacetime regions are not essentially gold. But the emphasis on identity criteria may still be retained, in accordance with the Quinean dictum “no entity without identity”. Material objects of any sort can be thought to be

The Myth of the Distinction between Ontology and Metaphysics

5

the same just in case the spacetime regions they occupy are the same. So we may tentatively make the following assumption: to tell what are the members of some category F (for example horses, or even material objects) amounts to telling what it is for members of F to be the same. So, the question becomes: can one say that the Fs exist without entailing or presupposing what it is for things that are F to be the same? As a matter of fact, this is often the case. For example, philosophers and even ordinary people are perfectly able to refer to persons and say that persons exist while wondering what it is to be the same person; and they can agree that persons exist while radically disagreeing on what it is to be the same person. Therefore, it is perfectly possible to say that something exists without saying what it is. But then, it might be argued, it is possible to do ontology without doing metaphysics. For, if there are sentences saying that some things exist without saying or presupposing what they are, then there are lists of such sentences. And what might ontology be if not a list of that kind? Therefore, one may conclude, ontology is one thing, metaphysics quite another.

3. Ontology and taxonomy The reason why I think that this conclusion should be resisted is precisely that I do not believe that ontology can be simply treated as a list of existential generalizations. Consider Borges 1952, where “a certain Chinese Encyclopedia” is described in which it is written that animals are divided into: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

those that belong to the Emperor, embalmed ones, those that are trained, suckling pigs, mermaids, fabulous ones, stray dogs, those included in the present classification, those that tremble as if they were mad, innumerable ones, those drawn with a very fine camelhair brush, others, those that have just broken a flower vase, those that from a long way off look like flies.

6

Chapter One

This “chaotic enumeration” is nothing but a list of existential generalizations saying that animals of each of the fourteen listed varieties exist. But this is not an ontology, not even a “regional ontology” for the zoological realm. Whatever it may be, ontology is a systematic taxonomy of what exists (either a systematic taxonomy of whatever exists or of what exists in a specific region of the real, such as the zoological realm). A systematic taxonomy must satisfy certain formal constraints, among which completeness (nothing should be left outside all categories), disjointness (nothing should be put in different categories unless one category is included in the other, as the category of horses is included in the category of animals) and uniformity (no ontological category is admissible unless the entities belonging to it have the same conditions of identity). Could a systematic classification of what exists fail to specify what the entities belonging to the envisaged categories are? If an ontological taxonomy could incorporate one or more categories of entities for which no conditions of identity are given, it should in principle be possible to disagree about the ultimate nature of the entities belonging to those categories (their conditions of identity) while leaving unchanged the system of categories itself, viz. ontology. For some categories of entities—for example, numbers—this is commonly held to be the case. Suppose two philosophers agree that numbers exist, believe that there is an ontological category to which all numbers belong, and never disagree about the existence of any categories of things. Certainly, they accept the same ontology, but this does not seem to prevent them from disagreeing about what numbers are. Philosopher A may see numbers as abstract individuals encoding the properties that follow from certain axioms (Zalta 1999), while philosopher B sees them as cumulative sets, conceived of either à la Zermelo or à la von Neumann. Therefore, one might conclude, an ontology committed to numbers seems to be fairly compatible with divergent theories of their ultimate nature (the point can be found stated in Varzi 2009). There are reasons to resist this conclusion, however, for it is far from clear that A and B really disagree about the nature of something. Consider that, by hypothesis, A and B completely agree about what categories of things there are. So they agree, in particular, that there are both abstract entities and sets, and agree that numbers are nothing over and above entities of these categories. If this is so, however, they can disagree about the ultimate nature of something only if they disagree about the ultimate nature of abstract entities or of sets (or both). Can they disagree, say, about the nature of sets? Suppose that A says that sets are entities of a kind F and B says that sets are entities of a kind G. Again, it would seem natural to

The Myth of the Distinction between Ontology and Metaphysics

7

conclude that A and B disagree about the nature of sets. But remember that A and B never disagree about the existence of some category of things, so they agree that both Fs and Gs exist and agree that sets are nothing over and above the Fs and the Gs. Therefore, disagreeing about the nature of something is for them tantamount as disagreeing about the nature of either Fs or Gs (or both). And we can ask now about Fs and Gs the same question we asked before about sets: can A and B disagree about their nature? In that direction, we can ascend up to the highest categories (those that restrict no other category), without being able to decide whether A and B really disagree or not about the nature of something. The moral is that, as long as there is total agreement on what categories of things exist, no apparent disagreement about the nature of the entities belonging to some category can count as a real disagreement about the nature of something, provided that there is total agreement about the nature of the entities belonging to the highest taxonomical level. Commitments to the ultimate nature of what exists can only be localized at the highest taxonomical level.

4. Semantical vs. metaphysical divergence One may be tempted to resist this conclusion by arguing along the following lines. Suppose that A and B agree that F and G are the highest categories (so that everything is either an F or a G), and agree about the ultimate nature of both Fs and Gs, but disagree as to which highest category numbers belong (it may be that A believes that numbers are abstract entities and abstract entities are Fs, while B believes that numbers are sets and sets are Gs). If any commitment to the ultimate nature of what exists were localized at the highest taxonomical level, as I said, then A and B would agree about the ultimate nature of everything. And, of course, A and B never disagree about the existence of anything, for by hypothesis they share one and the same ontology. So, what do they disagree about, if they disagree at all? It would be natural to answer: they disagree about numbers, about what they are, about their ultimate nature! If there were nothing about whose nature they disagree, in no way would A and B be disagreeing, but this is absurd. I see that there is a disagreement here, and I see that the disagreement has to do with numbers. It has to do with numbers, however, in no more than two ways. In the first place, A and B disagree about the reference of numerical expressions like “1”, “2” and so on (A believes that they refer to abstract entities satisfying certain axioms, while B believes that they refer to sets) and about the extension of the predicate “number” (A believes that

8

Chapter One

the extension of “number” is a set of abstract entities, while B believes that it is a set of sets). To the extent that the disagreement is about the reference of some words, it is semantic and has nothing metaphysical to it. It does not concern the ultimate nature of things, but merely the relations between words and things. The disagreement on ordinary objects between an orthodox perdurantism à la Lewis and a stage view à la Sider (see Lewis 1986, Sider 2001) can be regarded as a similar case. Lewis and Sider agree that both worms and stages exist, and substantially agree about their nature. But they disagree about ordinary objects (for example, cats). According to Lewis, cats are spatiotemporal worms while according to Sider they are instantaneous stages. Again, the disagreement seems to be semantic rather than metaphysical. It concerns the extension of ordinary predicates like “cat” and the reference of ordinary names like “Felix”—that is, how the words of our everyday language can best be mapped onto the fixed items, conceived of in the same way, of a shared ontology. So, in the first place, the disagreement between A and B is semantic. But it would be strange if it were only semantic, just as it would be strange if the disagreement between Lewis and Sider about ordinary objects were only semantic. For, provided we succeed in referring to something by uttering such words as “3” or “4” and such words as “Felix” or “Bobby”, we can ask what these entities are and disagree as to the correct way to answer that question. And this disagreement seems to be metaphysical rather than semantic. Is it? I think it is not semantic, but neither is it metaphysical. Suppose two people (say, A and B) are at some common point in a labyrinth and completely agree about the overall structure of the labyrinth. They have, say, a map and both believe that every part of the map represents a real part of the labyrinth. And both believe that the map fits fairly well the real structure of the labyrinth. In particular, they agree that the labyrinth is circular and can be divided, say, into three kinds of regions: the inner regions (up to 1/3 radius from the centre), the intermediate regions (up to 2/3 radius from the inner regions) and the external regions (up to the circumference from the intermediate regions). Therefore, so far as the labyrinth is concerned, A and B have the same ontology and the same metaphysics, the same theory of what there is and the same theory of what it is. But they are completely ignorant of where they are located inside the labyrinth. So they can look around themselves, refer to the path where they find themselves and disagree as to what sort of region it is. A says that it is an inner region while B says that it is an intermediate region. The disagreement is neither about the existence of some region of the labyrinth

The Myth of the Distinction between Ontology and Metaphysics

9

nor about its nature. The map clearly represents every existing region, as well as its (inner, intermediate or external) “nature”. And both A and B believe that the map is fairly correct. But A and B give different reconstructions of the points at which their perspectival experience of the labyrinth touches the non-perspectival overall image of it.3 So, the point remains. If A and B share a common ontology, and agree about the ultimate nature of what belongs to the highest categories, they cannot disagree about the ultimate nature of anything. They can indeed diverge as to how their perspectival knowledge of the world and their perspectival speech about it fits a shared non-perspectival overall image of the world, but they can only agree about the ultimate nature of everything (viz. they can only share one and the same metaphysics). Commitments to the ultimate nature of what exists can only be localized at the highest taxonomical level. If this is true, different people can accept the same ontology but different metaphysics in no more than one case: when they agree about what categories of things there are but disagree about the ultimate nature of what belongs to the highest categories. But it is far from clear that this can be the case. Suppose F and G are the highest categories of some ontological taxonomy. Can different people agree on what there is but diverge on what Fs and Gs are? Or equivalently, can one change one’s theory of what either Fs or Gs are while leaving unaltered one’s ideas about what there is? I strongly doubt it. One might change one’s theory of what Fs and Gs are in two different ways. First, one can discover that F and G have different extensions from what was previously thought (some Gs are nothing over and above Fs, or vice versa); second, one can leave F’s and G’s extensions unaltered while discovering that Gs (or Fs, or both) have criteria of identity different from what was previously thought. In both cases, it seems hardly believable that nothing has changed in the theory of what there is. On the one hand, to move entities from one category to another amounts to reducing entities of one category to entities of another. Where there were two entities, now there is just one. How might one’s ontological commitments have remained unchanged? On the other hand, if the conditions that Fs must satisfy in order to be identical change, it cannot be that the set of Fs remains the same. The moral to be drawn is that the theory of what there is pre-empts the theory of what it is. Unless it is a “chaotic enumeration” in Borges’ style, ontology cannot be separated from metaphysics.

10

Chapter One

5. Ontology as a “sphere of inquiry”? One possible reaction to this conclusion may be to bite the bullet and embrace the idea that ontology cannot be asked to be anything more than a “chaotic enumeration” in Borges’ style. Systematic ontology and metaphysics, it might be observed, are indeed two faces of the same coin. But pure ontology is chaotic and metaphysically innocent. It cannot be asked for an “inventory” or a “systematic classification” of what there is, but just for a possibly incomplete and possibly redundant list of ontological commitments that it seems reasonable to accept. From this point of view, Varzi has recently contrasted ontology as a theory with ontology as a sphere of inquiry. Conceived of as a theory, ontology is a systematic inventory of all that there is. Conceived of as a sphere of inquiry, ontology is concerned with what there is in a completely non-systematic perspective, its only task being the assertion or denial of a disordered mass of existential sentences, like those that typically figure in a chaotic enumeration. Inasmuch as two philosophers assert and deny exactly the same sentences of the list, they accept the same ontology, regardless of whether they embrace or not the same theory of the ultimate nature of what they are asserting or denying the existence of. Suppose, for example, that both “statues exist” and “lumps of matter exist” are contained in the list. If two philosophers assert both sentences but one thinks that statues are identical with lumps of matter while the other does not, their disagreement does not concern what there is, but only what it is, so there is nothing ontological to it, provided that ontology is conceived of as a “sphere of inquiry”. The idea of ontology as a sphere of inquiry is far from clear. The word “ontology”, as it is currently used, does not seem to denote anything like what is currently called “a sphere of inquiry”. Spheres of inquiry are normally conceived of as sets of questions, problems or issues rather than sets of assertions, while ontology, as it is currently conceived of, can only consist of assertions. More importantly, ontology is not only aimed at producing existential assertions, but also at supporting them by arguments. And very often arguments for the existence of some sort of things are grounded in some idea of what things of that sort are (for example: numbers exist because they are sets, and sets exist). Finally, the idea that an absolute boundary can be traced between an sit assertions and quid sit assertions might come out to be ultimately untenable. We said before that, if two philosophers A and B accept exactly the same list of existential generalizations, among which “statues exist” and “lumps of matter exist”, but one thinks that statues are lumps of matter

The Myth of the Distinction between Ontology and Metaphysics

11

while the other does not, their disagreement is purely metaphysical, in no way ontological. But suppose one adds to the list of existential sentences jointly accepted by A and B a sentence like “entities that occupy the same place as a statue at some time and pre-exist that statue exist”. If A and B disagree about the identity of statues with lumps of matter, they must also disagree about the truth of the added sentence, so they would no longer accept the same list of existential sentences. A metaphysical disagreement has now become ontological, even though it seems to have exactly the same content as before. The example can be generalized, the conclusion being that every metaphysical disagreement can be turned into an ontological disagreement merely by enlarging the chaotic list of existential generalizations that is under ontological analysis. In the same way, every ontological disagreement can be turned into a metaphysical disagreement simply by restricting the list of existential generalizations that is under ontological analysis. Regardless of whether the idea of ontology as a sphere of inquiry is acceptable or not, it does not succeed in making ontology independent from metaphysics.

6. Ontology and chaotic enumerations One may wonder what might be the role of chaotic enumerations in ontology, if any. Chaotic enumerations can be neither a part of ontology nor a product of ontological reflection. It is not the aim of ontology to randomly collect ontological commitments and put them in a disordered list. Neither can ontology aim at assigning a truth-value to the items of the chaotic list. There can be no sort of discourse in which one can decide whether it is true or not that there are true friends, blue azaleas, molecules, desires, and bacteria of the species Pseudomonas aeruginosa. In a sense, however, chaotic enumerations do indeed play a crucial, indispensable role for the rise of a systematic theory of what there is. One key point is that to do ontology is not to begin ex nihilo the game of saying what there is. We already play this game in ordinary language and science—indeed in anything we say. If we list in whatever order some of the things we are (pre-philosophically) committed to in any of the different domains that are for us of some interest, we have a chaotic enumeration. How can the game of saying what there is be carried over in ontology? Certainly not by adding new items to the chaotic list, which could only increase the chaos. Ontological reflection begins when one wonders how to turn a chaotic enumeration into a systematic theory of what there is, viz. how to bring order, coherence, unity, precision and completeness into a

12

Chapter One

confused congeries of disconnected existential commitments. Ontology essentially aims to systematize our pre-philosophical heterogeneous ontological commitments, not to add new “internal” commitments to the old “external” ones. But the only way one can turn a list of heterogeneous ontological commitments into a systematic taxonomy of what exists consists in saying what the entities mentioned in the list are, viz. what is their ultimate nature. So, to build a systematic ontology amounts to imposing a metaphysics on a chaotic list of ontological commitments. The outcome of this process is something that is at once a metaphysics and an ontology—both a systematic theory of what there is and a theory of its ultimate nature. One can find mixed together in a chaotic enumeration existential generalizations belonging to extremely heterogeneous discourses speaking of completely different domains of objects (quarks, rhododendrons, chairs, waves, colours, black holes, pains, fields, numbers, true friends and so on). The only way to unify all these ontological commitments in an ontological taxonomy is to ascend to the highest levels of generality. This is why the whole process is oriented to the individuation and characterization of the highest categories that all the other categories restrict (they may reduce to just one, for example, tropes). For any entity, indeed, the question “what is it?” is in the last analysis answered by reference to the highest category it belongs to. This shows that it is only in a very weak and derivative sense that ontology can be said to specify what there is. None of the ontological commitments one can find in ontology are entirely grounded in ontology itself, for in a sense ontology can only reorganize ontological commitments that come from outside (from the sciences and from common sense). By saying what are the entities to which we are committed outside ontology, one can turn a great number of prephilosophical commitments into a smallest number of wider commitments internal to ontology. An example is: 1) there are trees, houses and elementary particles; 2) trees, houses and elementary particles are bundles of universals; 3) there are bundles of universals. Only 2) and 3) are properly philosophical statements, 3) is grounded in 1) and 2), but 2) says nothing about what exists. This explains why the items of a chaotic enumeration usually disappear in an ontological taxonomy, and also explains why to leave no place for them in a taxonomy does not amount to deny their existence (for example, the existence of trees and houses). It may indeed be sometimes the case that some external ontological commitments are not subsumed under some wider commitments in an ontological taxonomy, but just given up (which

The Myth of the Distinction between Ontology and Metaphysics

13

is the case, for example, of the commitment to chairs and tables in van Inwagen’s theory of what there is).4 The decision to give up in that way some of the entities we are ordinarily committed to, however, cannot be grounded on empirical bases, but only on the impossibility of putting anything of that kind into any of the admissible categories (a claim that can only be justified by a metaphysical analysis of the ultimate nature of the entities of that kind). This cannot be the rule, however, for a systematic taxonomy according to which a tiny minority exists of what is thought to be existent by ordinary people and scientists would be hardly acceptable. An ontological taxonomy, indeed, is evaluable in terms of its efficacy in giving a structure, order and unity to a huge number of heterogeneous existential commitments widely acceptable in ordinary language and sciences. Here is the general idea in Lewis’ words: One comes to philosophy already endowed with a stock of opinions. It is not the business of philosophy either to undermine or to justify these preexisting opinions to any great extent but only to try to think of ways of expanding them into an orderly system. A metaphysician’s analysis of mind is an attempt of systematizing our opinions about mind. It succeeds to the extent that 1) it is systematic; and 2) respects those of our prephilosophical opinions to which we are firmly attached [...]. So it is throughout metaphysics and so it is with my doctrine about possible worlds (Lewis 1973, 88; see also Rescher 1997).

This idea of first philosophy as having primarily to do with systematization might seem to entail a descriptive conception of first philosophy, one according to which first philosophy aims to represent the world as we conceptualize it rather than as it is in itself. Here is why this is not the case. The more general is a category, the less its characterization depends on the particular nature of the entities we are actually committed to. So, the highest categories, those that restrict no other category, are purportedly such that anything possible belongs to them, regardless of its specific nature and kind. Therefore, first philosophy is not pre-empted by the analysis of our current ontological commitments. The chaotic set of the existential generalizations we pre-philosophically accept can be systematized in various and divergent ways, for it is compatible with a number of different theories of the ultimate nature of what we are committed to. The general point remains: since there is no way to do ontology without doing metaphysics and vice versa, “ontology” and “metaphysics” are different ways of naming the same discourse account or theory, just as “Hesperus” and “Phosphorus” are different ways of naming

14

Chapter One

the same planet. The distinction is flawed and should be given up. It is just a myth.

References Borges, J.L. 1952, “El idioma analitico de John Wilkins”, in Otras Inquisiciones, Buenos Aires: SUR Frege, G. 1884, Die Grundlagen der Arithmetik. Eine logisch mathematische Untersuchung über den Begriff der Zahl, Breslau: Koebner Ingarden, R. 1964, Der Streit um die Existenz der Welt, Tübingen: Niemeyer Inwagen, van, P. 1990, Material Beings, Ithaca (NY): Cornell University Press —. 1998, “The Nature of Metaphysics”, in Laurence S. and C. Macdonald (eds.) 1998, Contemporary Readings in the Foundations of Metaphysics, Oxford: Blackwell, 11 Kripke, S. 1980, Naming and Necessity, Cambridge (MA): Harvard University Press Le Poidevin, R. 2009, “What is Metaphysics?”, in Cameron R.P., R. Le Poidevin, A. Mc Gonigal and P. Simons (eds.) The Routledge Companion to Metaphysics, London: Routledge, xviii Lewis, D. 1973, Counterfactuals, Oxford: Blackwell —. 1986, On the Plurality of Worlds, Oxford: Blackwell Perry, J. 1979, “The Problem of the Essential Indexical”, Nous 13(1), 3 Rescher, N. 1997, Profitable Speculations. Essays on Current Philosophical Themes, New York: Rowman & Littlefield Quine W.V.O. 1951, “Ontology and Ideology”, Philosophical Studies 2 (1), 11 —. 1983, “Ontology and Ideology revisited”, The Journal of Philosophy 80 (9), 499 Sider, T. 2001, Four-Dimensionalism. An Ontology of Persistence and Time, Oxford: Clarendon Press Varzi, A. 2009, “On Doing Ontology without Metaphysics”, Paper presented at the Conference on Ontology, Raleigh (North Carolina), September 26 Zalta, E.N. 1999, “Natural Numbers and Natural Cardinals as Abstract Objects: A Partial Reconstruction of Frege’s Grundgesetze in Object Theory”, Journal of Philosophical Logic 28, 619

The Myth of the Distinction between Ontology and Metaphysics

15

Notes 1

In order to refer to a general account that includes both theories, I shall use the Aristotelian expression “First philosophy”. I do not believe that this is the best lexical choice in general. As far as I can see, however, it is the most suitable in the present context. 2 The very idea of a criterion of identity goes back to Frege 1884, § 62. 3 There is some analogy with Perry’s famous example of the “messy shopper” (Perry 1979). When the messy shopper discovers that it is himself who is the messy shopper, the discovery is not about the supermarket, its structure and what exists in it, but rather about the way in which a first person perspectival experience is related to a non-perspectival overall map of the supermarket. 4 See van Inwagen 1990.

CHAPTER TWO REALISM IN THE DESERT ACHILLE C. VARZI

1. The Desert The desert, of course, is Quine’s: the simple world of spatiotemporal sand he advocated in his early ontological writings, beginning with “On What There Is”.1 Quine spoke of it as a sober alternative to the McX’s and Wyman’s overpopulated universe, which is to say Meinong’s jungle,2 with all of its “rank luxuriance” of extraordinary creatures: unactualized possibilia, ideas in the mind, abstracta, things that do not quite exist in the good old sense of the term but that ought nonetheless to enjoy some form of being in order for what we say to be meaningful. Forget all that, said Quine. Perhaps a Fregean therapy of individual concepts might help us navigate through it all, but à quoi bon? We’d do much better simply to clear the slum and be done with it. Welcome to the desert. I like the desert. It may not be beautiful, but at least it doesn’t offend our aesthetic sense. It may not be comfortable, but at least it is safe. It is dry, clean, simple, quiet. It is as light as a place could be. And it is metaphysically extensional: in the desert, what you see is what you get. It is not, however, on the desert/jungle opposition that I wish to focus here. Everybody has thought about it at length and has already made up their mind one way or the other, and I have learned from experience that it is very hard to produce arguments against the jungle people. It is like arguing against dialetheists: you can’t win. You can only hope that your opponent will get tired of pretending to be serious about the content of their pronouncements. More importantly, I do not intend to dwell upon the contrast between the desert and the jungle because that contrast arises mainly—if not entirely—at the level of ontology, understood rather strictly as the doctrine of what there is. It’s about ontological commitments, and more specifically about Plato’s beard, Occam’s razor, and whether the blade of the razor is sharp enough to do a good shaving job. Those are all

Realism in the Desert

17

important issues, and for a long time the debate on realism has in fact focused on them. But there is a different side to the debate, one that pertains to issues in the domain of genuine metaphysics, as opposed to mere ontology, and that in my opinion bites even deeper. That’s the side I intend to focus on. I am appealing, here, to a certain distinction between ontology and metaphysics that I have tried to articulate and defend elsewhere.3 I am aware of its limits and I know it is controversial, but in the present context it is not crucial to endorse it wholeheartedly. I’m just using it to set the stage. Intuitively, the idea is that ontology is concerned with the question of what there is, i.e., what entities exist, whereas metaphysics seeks to explain, of those entities, what they are and how they are organized—their nature, how they relate to one another, what laws govern them, and so on. Thus, if you like C.D. Broad’s popular metaphor,4 ontology is concerned with the task of drawing up a complete “inventory” of the world, of specifying its content, whereas metaphysics is concerned with the structure of that inventory, its internal organization, its foundations if you like. Or again, in the terminology of the scholastic tradition, ontology is concerned with an sit questions, whereas metaphysics deals primarily with quid sit questions. This distinction is, I think, intuitive and helpful—and relatively uncontroversial. It becomes controversial if you match it (as I would) with the claim that ontology comes first, i.e., that it is in some sense prior to metaphysics: one must first of all figure out what things exist or might exist; then one can attend to the further question of what they are, specify their nature, speculate on those features that make each thing the thing it is.5 But this priority claim is not required for the distinction to make sense, and I don’t need to rely on it. I only want to say that, just as we are bound to face questions of realism at the ontological level, in the good old sense of the term, so we are facing questions of realism at the metaphysical level, in the sense I have just explained: realism about the structure of the world, not about its content. And in this connection the opposition is not between Quine and Meinong—between the desert and the jungle. It is between Quine and Aristotle, between the desert and the garden, so to speak—and I mean a natural garden, like the Garden of Eden, with its tidily organized varieties of flora and fauna neatly governed by natural laws that reflect the essence of things and the way they can be, or the way they must be. To the extent that you believe that the world is like a garden in this sense—that it comes structured into entities of various kinds and at various levels and that it is the task of philosophy, if not of science generally, to “bring to light” that structure—to that extent you are a realist. But if you think that the Edenic tree of the knowledge of good and

18

Chapter Two

evil is a fiction and that a great deal of the structure we are used to attribute to the world out there lies, on closer inspection, in our head, in our organizing practices, in the complex system of concepts and categories that underlie our representation of experience and our need to represent it that way—then, to that extent you are not a realist. I realize this is all highly metaphorical, but I trust the picture is clear and familiar enough. I also hope it is obvious that you don’t get to live in a metaphysical desert of this sort unless you already live in an ontological desert. That, however, is only a necessary condition, not a sufficient one. Your ontological desert could still have a lot of structure. For example, structure in terms of essences, laws of causation, unity, persistence through time, etc. Not so if you live in a metaphysical desert. So, in a way, when it comes to the metaphysical part of the story, Quine’s notion of a desert—at least, the notion of a desert I am interested in here—goes back to Hume, not to Occam.6 Causation is perhaps the best example. For Hume, it is a paradigm example of a metaphysical “fiction”, a concept that does not correspond to a genuine feature of reality.7 There is nothing, in reality, necessarily connecting what we call “cause” and what we call “effect”. Or rather, there is nothing we can observe in reality except for certain relations of succession, contiguity, and constant conjunction. Hence we cannot form any philosophically respectable concept of causation over and above that of a constant conjunction of like objects in like relations of succession and contiguity, pace our natural “propensity” to go for something bigger. Likewise, consider Hume on identity, i.e., diachronic identity.8 On the face of it, the thought that things persist through time underlies much of our everyday interaction with the world of ordinary experience. We readily suppose that an object may continue numerically the same, in spite of the fact that it may undergo several qualitative changes and for most of the times it is absent from the senses. Bananas ripen, houses deteriorate, people lose hair and gain weight. In this world of flux, persisting things are the only anchor we have, but the source of their persistence is a puzzle that has been with us since the early days of philosophy. What grounds our belief that the things around us (and ourselves, too) may survive from day to day in spite of the many changes that affect them? How can we say that they are the same things, if they are no longer the same? The answer, for Hume, is that we can’t. On the face of it, the identity relation can only apply to constant and unchangeable objects. It is only “the smooth passage of the imagination” along the ideas of resembling perceptions that makes us ascribe identity to variable or interrupted objects, it is our propensity to unite broken appearances of resembling perceptions that produces the

Realism in the Desert

19

“fiction” of a continued existence. Ditto for personal identity, where again Hume is quite explicit in using the language of “fiction”. And ditto for unity.9 Here, too, Hume famously argued that we have a propensity to attribute existence to multiplicities, such as a school of fish, when strictly speaking existence in itself belongs only to unity—a fish. Strictly speaking, unity is never applicable to a multiplicity except on account of the “unites” of which that multiplicity is composed. Thus, again, when we give way to our propensity to say more, strictly speaking we engage in a mental construction, a pretence, a “fiction”. Now, to all this Quine would add that even the notion of an object, a material object, which is the most basic notion that we have when it comes to the world out there, is the result of some operation of the mind. Even that involves an intellectual or ideological construction of some sort. And one way of putting this claim, which is more typical of the later Quine,10 is based on a consideration of how our experience of the world is shaped by our cognitive development. At the beginning there is just world (mass term). It’s not all alike: here is mama, there is cold, over there—noise. Soon we begin to distinguish and to recognize: more mama, more cold, more noise! Yet initially these things appear to be all of a type. Each is just a history of sporadic encounter, a mere portion of all there is. Only with time does this fluid totality in which we are immersed begin to take shape. With time, objects begin to objectify; they begin to move, fall, break, disappear and reappear. Sensations acquire more definite contours, fade out and come back, resemble one another in our memories. Noises vary depending on the things around us. We begin to act and to predict. We launch into giving names, using verbs, painting adjectives. Such marvellous unfolding is the subject of much inquiry by psychologists and biologists, and eventually by sociologists. But for a philosopher it is really the source of deep and bemusing bewilderment, if not a dilemma: Are we learning to make out the structure of the world, or are we endowing the world with a structure of our making? Is reality gradually revealing the mechanisms according to which it is organized, or is it we who progressively organize the amorphous and continuous flux of our experience? And the Quinean answer, or at least my Quinean answer, is—the latter. That’s what it means to say that we live in a metaphysical desert.11

2. Boundaries in the Sand In some of my writings, I have put this in terms of boundaries.12 There are no boundaries, in the desert. More precisely, there may be lots of

20

Chapter Two

boundaries of various sort, but no natural boundaries, no “joints of reality”. All the boundaries we find are lines we have drawn, artificial fencings that merely reflect of our own demarcations, our classifications, our organizing activity. We may have the feeling that in some cases such demarcations are grounded on natural discontinuities in the underlying reality. But look closely and you’ll see that there are no discontinuities. It’s just sand, and sand, and sand… In the geopolitical domain, where the boundary metaphor finds its origin, this claim is only prima facie extravagant. Prima facie, the distinction between natural and artificial boundaries is perfectly intelligible, and we should be thankful to the British Viceroy of India, Lord Curzon of Kedleston, for having explained it so clearly in his celebrated Romanes Lecture of 1907.13 Think of Ireland, Lake Maggiore, or the crater of Vesuvius versus Austria, Wyoming, Yellowstone Park, or a soccer field. In the latter cases, it is clear that we are dealing with entities whose boundaries are the expression of a collective intentionality that translates—peacefully or by warfare—into political, social, and legal agreements whereby it is determined where a certain territory begins and where it ends. Not so in the former cases, where the boundaries seem to have nothing to do with our organizing activity. We can stipulate that one part of Lake Maggiore belongs to Piedmont and the rest to Lombardy, and the dividing line will be an artefact. But the shoreline—the border of the whole lake—does not seem to depend on us. It’s there regardless, it exists “on its own”. Ditto for the boundaries of a volcano, an island, or even a peninsula, such as Iberia, which although connected to the continent is separated from it by the “admirably fashioned” Pyrenean wall. The distinction between these two sorts of boundary is so compelling, intuitively, as to seem ineluctable. In a physical map we may omit all political boundaries; but a political map will perforce include all physical boundaries—at least, those physical boundaries that are visible at the relevant scale. And yet it doesn’t take long to realize that the distinction is not as ontologically robust as it appears to be. As we fly over lake Maggiore we have the impression of seeing its shoreline, as opposed to the Piedmont/Lombardy boundary. But we all know that, as we go there, ground-level, things are quite different. What looked from the air like a sharp line turns out to be an intricate disarray of boardwalks, stones, cement blocks, musk sediments, marshy spots, decayed fish. Ditto for the celebrated borderline of the Irish island. It’s not just a matter of our disrespect for Nature. Things would not be different if we took a close look at the coast of a virgin island in the middle of the ocean. We may locate the boundary of the island at the water/beach interface, but that

Realism in the Desert

21

boundary is constantly in flux and it is only by filtering it through our cognitive apparatus—it is only by interpolating objects and concepts—that a clear-cut line may emerge. And what goes for lakes and islands goes for all prima facie natural geographical or celestial entities: rivers, craters, mountain chains, continents, whole planets. The distinction between natural and artificial boundaries is by itself intelligible, and it may be extremely helpful in explaining our geopolitical behaviour. (It is one thing to take up arms for the independence of the Irish island, quite another to sacrifice your life for the boundary of Wyoming.) Deep down, however, it is merely a rhetoric distinction: every boundary is, to some degree, the product of our own making. Now, to say that we live in the desert is to say that this picture applies across the board, literally and metaphorically. For boundaries play a central role, not only in relation to the geographic world that we find depicted in ordinary maps and atlases, but at any level of representation of the world around us. The natural/artificial distinction is at work in articulating every aspect of the reality with which we have to deal— objects, events, kinds, properties, whatever. Thus, the spatial boundary separating my part of the desk from my colleague’s, or my head from the rest of my body, would seem to be entirely artificial, whereas the boundary demarcating my whole body, or the interior of this apple from its exterior, would seem to be of the natural sort. The end of a soccer game or my turning 50 years old would be examples of artificial temporal boundaries, whereas my birth and death, or the change between rest and motion, would be obvious candidates for natural boundaries. The properties expressed by disjunctive predicates such as “grue” and “emerose”,14 or by phase sortals such as “student” and “jobless”, would seem to have artificial, de dicto boundaries; those expressed by substance sortals and so-called natural kind terms, such as “cat” or “hydrogen”, would have natural, de re boundaries. It is hard not to see the force of these distinctions. And as we see their force, it is a short step to draw all sorts of consequences in the spirit of a robust metaphysical realism. For if the boundaries that define a certain entity are artificial—if they reflect articulations effected through our human cognition and social practices, including mere linguistic conventions—then we are dealing with a fiat entity, a social or cognitive construction, a product of our worldmaking. But if those boundaries are of the natural sort, it is reasonable to suppose that the identity and survival conditions of the corresponding entity do not depend on us—that it is a bona fide entity of its own. Well, on the Quinean-Humean view I hold, both such conditionals are true. But the latter is only vacuously true. For just as the natural/artificial opposition founders in the geographic case, so

22

Chapter Two

it founders across the board. Just as there literally are no natural boundaries in the Sahara, as Libya and Egypt know all too well, so too in the sandy world we live in. There ain’t no natural joints. Consider the case of material objects, such as this apple. A natural, mind-independent entity? We know very well that the apple is not the solid, spatially continuous thing that it seems to be. On closer look, it is just a smudgy swarm of hadrons and leptons whose boundary is no more settled than that of a school of fish, a swarm of bees, a stock of starlings. On closer look, the spatial boundary of this apple involves the same degree of arbitrariness as that of any mathematical graph smoothed out of scattered and inexact data, the same degree of idealization of a drawing obtained by “connecting the dots”, the same degree of abstraction as the figures’ contours in a Seurat painting. Whether we like it or not, no apple comes with a natural boundary of its own: if it has a surface at all, it is as good a Humean “fiction” as you’d ever meet. Nor is it just a matter of microscopes. Take me, which is to say my body—a paradigm example of a living creature, I presume, hence a good candidate to the status of a bona fide individual. Regardless of my subatomic structure, the question of what counts as me—what lies within the boundary of my body—has no clear, unprejudiced answer. When it was on the table, the apple was not part of me. But now I am eating it and I have a piece in my mouth—Is it part of me? Will it be part of me only after some chewing? Only when I swallow it? Only at the end of the digestive process? You tell me. Surely the stuff composing my body is not a product of your own making; yet what stuff counts as composing it—hence the identity and survival conditions of what you take to be my body—will obtain only relative to some stipulation or other concerning its boundary. And what goes for objects goes for events. The boundary between rest and motion? Yes, it looks like a natural temporal boundary. On closer inspection, however, we know that a body’s being at rest amounts to the fact that the vector sum of the motions of the zillions of restless particles of which the body is composed averages to zero. It makes no sense to speak of the instant at which a body begins to move, even less to speak of it as of a natural boundary. My birth and my death? Yes, they are excellent prima facie candidates of natural temporal boundaries; yet the controversies concerning abortion and euthanasia push in the opposite direction. Sometimes it is a matter of our deciding whether a person is still alive. We decide whether her “vital” functions are still in force, and the criteria for such decisions give expression to our beliefs, our principles, our theories. Similarly, the initial boundary of a person’s life is hardly fixed by Nature alone. Surely it does not coincide with the person’s birth

Realism in the Desert

23

(by itself a rather intricate, messy process), except for the registry office. But neither is there an earlier “moment” that fits the bill easily. We can settle on the time of fertilization, or perhaps on the formation of the zygote, or the beginning of the implantation process, or some other time. The candidates are many and we can base our decision on as many factors as we like, including up-to-date scientific findings. But a decision it is. And if it is a matter of our (arbitrary or informed) decision, then the boundary is not genuinely de re and even a person’s life becomes, to some extent at least, a Humean “fiction”. What about natural properties? Natural kinds? Natural taxa? Here the geographic metaphor is admittedly strained. And yet, again, it is a fact that on closer look our parochial concerns—historical and cultural circumstances, practical interests and limitations, theoretical priorities— play a much bigger role than we might want to admit. No one thinks that emeroses form a natural kind. But why should roses? Why settle on Rosa chinensis? No one thinks that quadrupeds form a natural kind. But it would be equally remarkable, to use Catherine Elgin’s example,15 if a taxonomy that draws the distinction between horses and zebras where we do aligned at all well with categories “fitting the cosmos” regardless of our human faculties and ends. Granted, the thought that biological taxa are mere conventions clashes with the certainty of some biologists concerning “the objective reality of evolution”, as David Stamos has it.16 Yet Darwin himself was adamant that the term “species” is one “arbitrarily given for the sake of convenience to a set of individuals closely resembling each other”.17 And as we open the Annual Review of Ecology and Systematics, we should not be surprised to read that “taxa are human constructs” and “natural taxa are those that are natural to humans”.18 After all, there are horses and zebras, but also zorses and hebras. Didn’t Locke even claim to have seen “the issue of a cat and a rat”?19 On closer look, even our best physical theories deliver microscopic categories that depend on a number of human contingencies. If we construe different isotopes as variants of the same type of atom, say hydrogen, it is because of certain reasonable interests that predominated in the development of those theories. One might as well construe deuterium and tritium as different chemical elements altogether, hence as “natural” kinds of their own. Again, the problem is not that there are no discontinuities in the physical world; the problem is that there are too many discontinuities. And to privilege some over the others is, once again, tantamount to drawing an artificial boundary—a boundary in the sand.

24

Chapter Two

3. Realism Now, the title of this paper says realism in the desert. Why “realism”, if accepting this sort of picture amounts to giving up the thought that the world comes with a robust, objective structure of its own, a structure that is not imposed by us but grounded in the very nature of things? Let me try to answer by refining the picture in four respects—the picture and what truly follows from it.20 First of all, the desert is real. It’s out there and it is what it is regardless of how we feel. To be a Humean fictionalist, or even a social constructivist, is not to be a Berkeleyan idealist. Even assuming that all the “metaphysical” structure we tend to attribute to the world is a projection of our minds and of our organizing practices, and even assuming that it is cognitively biased, artificial, and wholly arbitrary (which is to say, in our arbitrio), it does not follow that the world itself is a projection of the mind. Indeed, the very notion of an artificial boundary is intelligible only to the extent that we acknowledge an appropriate real basis for the demarcations that are effected by our pencils and cookie cutters. That the underlying worldly material should itself be a cognitive construct of sort is a different, stronger thesis that I do not even understand (short of Cartesian scepticism). To put it differently, the view in question does not deny the objective reality of the entities that we carve out by fiat. As Frege put it, the objectivity of the North Sea “is not affected by the fact that it is a matter of our arbitrary choice which part of all the water on the earth’s surface we mark off and elect to call the ‘North Sea’”.21 The Humean fictionalist simply says that every portion of reality is equally objective as any other and that none has greater metaphysical value than any other. In this sense, the desert picture, with all the boundary-drawing and world-making business that comes with it, is not to be mistaken with Goodmanian irrealism, either. For Goodman—and for Rorty—all we learn about the world is contained in right versions of it, and the world itself, bereft of these, is on the whole “a world well lost”.22 On the desert picture, the world is boneless, impoverished, almost bankrupt, but our love for it is not at stake. For Goodman, a world-version need not be a version of the world, just as a Pegasus-picture need not be a picture of Pegasus.23 Not so on the desert view. The metaphysical structure is imposed by us, and therefore fictitious. You can even say that it is a narrative, a story. But it is a story about the world, and all the maps we draw are maps of one and the same reality. Second, the desert is not completely boneless. It’s just that the structure it has is very thin, and is not the sort of structure that traditional and

Realism in the Desert

25

contemporary metaphysical realists—and scientific realists alike—tend to attribute to it. It is, basically, a purely geometric structure. More precisely, it is a mereological structure, the structure that comes with the part-whole relation: some portions of the desert are part of other, larger portions; some parts overlap, some do not; nothing is a proper part of itself, and nothing can have just one proper part; no two things can have the same proper parts; given any number of parts, no matter how scattered or gerrymandered, there is always something that is composed exactly of those parts; etc.24 The desert is bound to be structured this way, and objectively so. It’s not up to us to decide whether or under what conditions x is part of y, just as it is not up to us to decide whether x is the same as y. Similarly—and this is a realist claim that applies already at the ontological level—it is not up to us to decide which parts exist, or which mereological fusions exist. They all exist. What is up to us, if we live in the desert, is simply which parts will come to play a role in our lives, which portions of the desert will be selected and elected to special status by our cognitive and/or social cookie cutters. In this sense, the desert picture differs from Goodman’s irrealism as it differs from Putnam’s brand of anti-realism. For Putnam, the “cookie cutter” metaphor founders on the question, “What are the ‘parts’ of the dough?”.25 No neutral description is available to compare, say, a LeĞniewskian world (with x, y, and the mereological fusion of x and y) and a Carnapian world (with only x and y), and assigning a univocal meaning to the verb “exists” is already “wandering in Cloud Cuckoo Land”.26 For the desert theorist, the metaphor holds and “exists” corresponds to the standard existential quantifier. Which existing entities we are going to select is up for grabs; but the parts of the dough, which is to say of the desert—those parts that provide the appropriate real basis for our carvings—they are whatever they are and the relevant mereology is a genuine piece of metaphysics. (As I mentioned, I personally think that composition is unrestricted, hence the mereology is LeĞniewskian; so either Carnapians are not speaking with their quantifiers wide open, as when we say “There is no beer” meaning “There is no beer in the refrigerator”, or else they are wrong.27 On the other hand, perhaps they are right and I am the one who has it wrong. What is sure is that we can’t be both right.) Third, the desert need not be a boring place. If that is what worries us, it is certainly not a reason to forgo the desert for a garden of Eden, let alone a jungle. Here perhaps the “desert” metaphor is somewhat misleading. For it’s not necessarily all grey. Just look around you. Some portions of reality are red, others are blue. Some are tasty, others insipid.

26

Chapter Two

Some are noisy, others—quiet. The parts of the whole are as colourful and multifarious as they seem to be. It’s just that there is nothing in virtue of which they are as they are. The apple is red, but there is nothing in virtue of which it is red, on pain of Bradley’s regress.28 That was Quine’s point: McX is no better off, in point of explanatory power, for all the occult entities which he posits under such names as “redness”.29 Likewise, in the desert there are dogs and cats, and trees and flowers. It’s just that those are not different kinds of thing, each with its own essence and persistence conditions. Simply, some portions of reality strike us as dogging while others as catting, or treeing, or flowering. In the desert you’ll find Socrates, but not Pegasus, for there is a certain portion that strikes us as socratizing while none is pegasizing. In the desert you may see a white cat facing a dog and bristling, for somewhere it’s catting whitely, bristlingly, and dogwardly.30 Of course, here the details depend on what exactly the desert is like. Some people think it is really made of sand—Democritean sand. In that case, to say that a certain portion is ij-ing is tantamount to saying that there is sand arranged ij-wise. Others think the desert is mere res extensa, stuff, atomless gunk, in which case there is no sand after all; it’s sand all the way down. Or perhaps the desert is entirely made up of tropes. Or it’s the wave function playing itself out in configuration-space. I’m not sure. Depending on the truth, the details may vary. But whatever it is, the desert need not be a boring place. At least, it need not be boring just because the “structure” that we are used to attribute to it turns out to be, on closer inspection, a matter of our own artificial boundary-drawing. Manhattan is as artificial as something could ever be, but I’ve spent eighteen years of my life there and, trust me, I am not even beginning to get bored. Last, but not least, it is for this very same reason that life on the desert need not amount to a Feyerabendian “anything goes”,31 where the difference between knowledge and belief collapses and truth itself becomes an empty notion. I can see how the worry arises. If all interesting structure were the product of our cognitive or social organizing activities, if the lines along which we “carve” the world depended entirely on our cognitive joints and on the categories that we employ in drawing up our maps, then our knowledge of the world would amount to neither more nor less than knowledge of those maps. The thesis according to which all the structure is artificial would take us straight to the brink of precipice, to that extreme form of cognitive or social constructivism according to which “there are no facts, just interpretations”. That was Quine’s own worry in his controversy with Goodman on the pages of The New York Review of Books.32 But whence the worry? The boundaries of Wyoming are

Realism in the Desert

27

artificially drawn, yet that makes no difference to truth as long as we agree on where they are drawn. Either you are inside Wyoming, or you are outside; and if you believe that you are on one side when in fact you are on the other, then your belief is plain false. Truth and falsity work in the desert exactly as they are supposed to work in the garden of Eden. Every predicate has an extension, artificial as it may be; every singular term has a referent, arbitrarily demarcated as it may be. And a subject-predicate sentence is true if and only if the referent of the subject falls within the extension of the predicate. There may be vagueness, of course: not all boundaries are traced with precision. But so be it: precisely because the boundaries are of our own making, any such vagueness would be merely linguistic, or cognitive at large; it certainly isn’t vagueness that bites ontologically. Similarly, if the law forbids you from driving faster than a certain limit, and you do drive faster, than it is a fact that you broke the law and you deserve a fine. You can’t argue that. It is not a natural law, but a law it is nonetheless. It is not fixed by Nature and it does not have the force of necessity, which is to say that if we wish to repeal it, or set a different speed limit, we may do so; yet, as long as it is in force, it rules things. That’s another sense in which the desert is, as a matter of contingent fact, more like Manhattan than like the Sahara. Ferraris is right when he says that the view in question, just like pre-Kantian empiricism, obliterates any substantive differences between laws and train timetables:33 both are in some sense conventional. But as I’ve tried to explain elsewhere, timetables are not drawn up at random.34 They ensue from the necessity to solve, in a conventional but efficient way, coordination problems that can be extremely complex and that could seriously impair our daily deeds. If we come up with a timetable that works poorly, we change it. If a convention fails to measure up to our expectations, we replace it with a new, hopefully better one. Ditto with so-called laws of nature. For even in the desert, not all biological taxonomies (for instance) are on a par. Some are better than others, because they better support the “laws” that govern biology’s coordination game (laws of variation, selection, organic evolution, population growth, etc.). One may object that this sort of pragmatic efficiency calls for more than arbitrary, artificial demarcations. But the burden of proof is on the objector. Linnaeus’s Systema Naturae, the bible of all classical taxonomies, was soaked with essentialism—and the platypus didn’t even fit in.

28

Chapter Two

4. A Good Place to Be Let me conclude with a couple of general remarks on why I think the desert view—and the modest realism that comes with it—is not only a tenable view, but a good one. In a nutshell, I think it is a good view because it is honest. It acknowledges our parochial limits and our cognitive and cultural biases, along with our desperate need to ward off the flux and meet our demand for order and stability, without camouflaging any of this as “natural”. Of course, we all want our perspectives on things to be taken seriously and it is perfectly alright to act in conformity to the maps and laws that they elicit, especially when these appear to serve their purpose well. Just as it would be irrational to bet on the sun not rising tomorrow, so it would be irrational to act as though our most efficient scientific theories were on a par with gimcrack literary fiction. Hume himself was candid about this, as was Quine. But as we rely on our maps, we should not forget what they are and why we have them, on pain of self-deception. We should not think that they are metaphysically transparent and normatively unassailable simply because they help us navigate successfully, on pain of fallacy. It would be unfair to call “dishonest” those philosophies that succumb to self-deception and incur the fallacy. But the desert view is honest precisely insofar as it doesn’t. More importantly perhaps, the desert view requires that we be honest. It’s bad enough to mistake the rational for the real, much worse to sell our views as if they were the views of the world. Yet that is precisely what we tend to do—indeed, what we must do—whenever we act on the basis of a robust realist metaphysics that is meant to cut up the world “at its joints”. Our history is full of horrible things that we did because we followed Plato’s metaphor too blindly.35 We stigmatized and persecuted as “bad butchers”, hence “heretics”, whoever held views that did not match ours. We enforced our rules and precepts upon others as if they came directly from the “tree of knowledge”. We treated people like slaves, if not like animals, because we thought they were different from us in kind. We treated non-human animals as meat because of the privileged status we assign to our own kind. We discriminated human beings on the basis of their gender, race, and sexual orientation as though such features reflected deep metaphysical categories that cannot and ought not be transgressed. Even today it is common to hear that interracial or homosexual relations are “against nature”. All of this is horrible and I do not need to say why. But philosophically it is horrible precisely because it rests on the worst possible form of cheating: to pretend the world is on our side. As I said,

Realism in the Desert

29

this can hardly be avoided if we think and act on the basis of a robust realist metaphysics. In the desert, however, this sort of cheating is not permitted: we are free to fight for the views that we find most attractive, but we can’t spurn the alternatives as being against nature. In the desert we have to be honest and take full responsibility for our actions. This is more than enough, I think, to conclude that the desert is a good place to be. Not that I can say I’m optimistic. We are always going to make mistakes. We are always going to do bad things and pretend we are right. But it makes a big difference whether we do so because we have settled on the wrong criteria, on a poorly drawn map of the world, or because we feel we have to do it, because the world itself drew the lines that way. It makes a big difference for the simple reason that in the first case we can change things and hopefully improve them. If the need arises, we can redraw the map. If a law doesn’t work, we can revise it. It is up to us to do a good job, and it is our duty to try. For if we don’t manage, there’s no one else to blame but ourselves.

References Bradley, F.H. 1893, Appearance and Reality, London: Swan Sonnenschein Broad, C.D. 1923, Scientific Thought, London: Routledge & Kegan Paul Curzon, G.N. 1907, Frontiers, Oxford: Clarendon Press Darwin, C. 1859, On the Origin of Species by Means of Natural Selection, London: Murray Davidson, D. 1979, “Mental Events”, in Foster L. and J.W. Swanson (eds.), Experience and Theory, London: Duckworth, 79 Elgin, C.Z. 1995, “Unnatural Science”, Journal of Philosophy 92, 289 Ferraris, M. 2004, Goodbye Kant!, Milano: Bompiani Feyerabend, P.K. 1972, Against Method, London: New Left Books Frege, G. 1884, Die Grundlagen der Arithmetik, Breslau, Köbner Goodman, N. 1955, Fact, Fiction, and Forecast, Cambridge (MA): Harvard University —. 1968, Languages of Art, Bobbs-Merrill: Indianapolis —. 1975, “Words, Works, Worlds”, Erkenntnis 9, 57 —. 1978, “In Defense of Irrealism”, The New York Review of Books 25/20, 58 Kneale, W.C. 1949, Probability and Induction, New York: Oxford University Press Putnam, H. 1987, The Many Faces of Realism, La Salle (IL): Open Court —. 2004, Ethics without Ontology, Cambridge (MA): Harvard University Press

30

Chapter Two

Quine, W.V.O. 1948, “On What There Is”, Review of Metaphysics 2, 21 —. 1958, “Speaking of Objects”, Proceedings and Addresses of the American Philosophical Association 31, 5 —. 1978, “Otherworldly”, The New York Review of Book, 25/18, 25 —. 1985, “Events and Reification”, in LePore E. and B. McLaughlin (eds.), Actions and Events, New York: Basil Blackwell, 162 Rorty, R. 1972, “The World Well Lost”, Journal of Philosophy 69, 649 Sokal, R.R. 1986, “Phenetic Taxonomy: Theory and Methods”, Annual Review of Ecology and Systematics 17, 423 Stamos, D.N. 2003, The Species Problem, Lanham (MD): Lexington Books Varzi, A.C. 2000, “Mereological Commitments”, Dialectica 54, 283 —. 2006, “The Universe Among Other Things”, Ratio 19, 107 —. 2010, Il mondo messo a fuoco, Rome: Laterza —. 2011a, “Boundaries, Conventions, and Realism”, in Campbell J.K. et al. (eds.), Carving Nature at Its Joints, Cambridge (MA): MIT Press —. 2011b, “On Doing Ontology without Metaphysics”, Philosophical Perspectives 25, 407 —. 2013a, “Fictionalism in Ontology”, in Barbero C. et al. (eds.), From Fictionalism to Realism, Newcastle (UK): Cambridge Scholars Publishing, 133 —. 2013b, “Mereology”, in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy, Winter 2013 Edition —. 2014, “Counting and Countenancing”, in Cotnoir A.J. and D.L.M. Baxter (eds.), Composition as Identity, Oxford: Oxford University Press, forthcoming

Notes 1

Quine 1948, where the “desert” metaphor appears for the first time (at p. 23). The term “jungle” actually comes from Kneale 1949, 12. 3 See Varzi 2011b. 4 Broad 1923, 242. 5 The claim may already be found in Aristotle (Met., II, 2, 89b34-35) and is explicitly endorsed by Aquinas (e.g., S. Theol., I, q. 2, a. 2, ad 2). It was condemned by Duns Scotus (Ord. I, d. 3, 11 e 17) and eventually by Descartes, who said that putting the an sit before the quid sit would violate “the laws of true logic” (Primæ resp. 141). 6 Here I draw on Varzi 2013a, § 4. 7 Treatise, I, iii, 6. 8 Treatise, I, iv, 6. 2

Realism in the Desert 9

31

Treatise, I, ii, 2. Beginning with Quine 1958. 11 I have tried to articulate this view and its ramifications in Varzi 2010. Here I will only be concerned with those aspects of the view that relate directly to the realism/antirealism debate. 12 See e.g. Varzi 2010, chap. 3, and Varzi 2011a. Much of what follows draws on this material. 13 Curzon 1907. 14 The first predicate comes from Goodman 1955, chap. 3: given an arbitrary but fixed time t, “grue” applies to those things that have been examined before t just in case they are green, and to other things just in case they are blue. The second predicate is from Davidson (1970) and is defined in a similar fashion from “emerald” and “rose”. 15 Elgin 1995, 297. 16 Stamos 2003, 131 (n. 35). 17 Darwin 1859, 52 18 Sokal 1986, 424. 19 Essay, III, vi, 23. 20 Here again I draw on Varzi 2011a. 21 Frege 1884, § 26. 22 Goodman 1975, 60. Compare Rorty 1972. 23 Goodman 1968, 21. 24 On mereology I refer to Varzi 2013b. 25 Putnam 1987, 19. 26 Putnam 2004, 85. 27 My personal views on this and related matters may be found in Varzi 2000, 2006, 2014. 28 I mean the regress exposed in the famous argument of Bradley 1893, part 3, § 3. 29 Quine 1948, 30. 30 This is the later Quine; see his 1985, 169. 31 The phrase is from Feyerabend 1972, 23. 32 See Quine 1978 (a review of Ways of Worldmaking), followed by the rejoinder in Goodman 1978. 33 Ferraris 2004, 17. 34 See again the titles cited in n. 12. 35 The recipe in the Phaedrus, 265d, to the effect that we must “cut up the unity according to its species along its natural joints, and try not to splinter any part, as a bad butcher might do”. 10

CHAPTER THREE REALISM AND TRUTH FROM A QUINEAN POINT OF VIEW ANTONIO RAINONE

Realism and truth are closely interconnected in W.V. Quine’s views with a naturalistic approach to human knowledge. From Quine’s perspective (as in Dewey’s) we live in a natural setting that causes, by sensory stimulations, (at least) part of our beliefs; we also have certain ontological beliefs—of natural origins—about certain bodies and objects existing in the external world that make true our beliefs and their linguistic counterparts, namely, the sentences of our conceptual scheme (scientific theories being part of this scheme). But “naturalism”, in a broader sense, means also, for Quine, that the reality can be identified and known only through natural science. How, then, is this interconnection between naturalism, realism and truth to be understood? And in what does Quine’s realism consist?

1. Was Quine an instrumentalist? In order to answer to these questions we have to start from the beginning of the Quinean epistemological and ontological conceptions. The name of Quine, in fact, has seldom joined the debate about realism and anti-realism. This is perhaps not surprising: in the celebrated Two Dogmas of Empiricism Quine seemed rather to defend an instrumentalist (i.e. anti-realist) stance on epistemological matters according to which the physical objects and bodies in the external world are only myths or (cultural) posits—as Quine also claimed there. From the epistemological point of view, Quine said there that physical objects are conceptually imported into the situation as convenient

Realism and Truth From a Quinean Point of View

33

intermediaries—not by definition in terms of experience, but simply as irreducible posits comparable, epistemologically, to the gods of Homer (Quine 1951, 44).

Physical objects as myths? Certainly, this can seem very odd, particularly from a realist point of view. However, Quine claimed that in first place to challenge the Carnapian (and empiricist) attempt to define physical objects in terms of sense data. Quine thinks it is impossible—as the failure of Carnapian Aufbau had demonstrated—to reduce sentences about physical objects to sentences about sense data; it is impossible, in other words, to carry out the program of the early logical positivism. But the impossibility of this reduction does not prevent achieving a better warranty for our knowledge by a physicalistic description of reality. True, physical objects are myths, and so are sense data too. Nevertheless, between the two myths we should prefer, says Quine, the physical objects. Why? The reasons for accepting the language and ontology of physical objects are basically pragmatic: the myth of physical objects is epistemologically superior to most in that it has proved more efficacious than other myths as a device for working a manageable structure into the flux of experience (Quine 1951, 44).

So, we accept (and we should accept) the ontology of physical objects for the convenience and advantages that physical objects bring to our knowledge or “conceptual scheme”. In this connection we might consider Quine’s epistemology, at least as it emerges from Two Dogmas, as a form of instrumentalism or conventionalism. The ontology of physical objects is preferred for reasons of convenience, we could say. But the convenience had been, in fact, the label which Henri Poincaré appealed to in his philosophy of geometry concerning the choice between Euclidean and non-Euclidean geometries. If the choice between two different geometries of the physical space, and two different ontologies of reality, is only a matter of convenience, then there it is nothing we can call the ultimate nature of space or reality. That’s a problem—at least for where to place the conceptions of Quine in the contemporary philosophy of science. True, if his views were analogous to those of Poincaré, then we could label them as a conventionalist conception of the knowledge—very different from the realist one. The distance of Quine from a realist view of knowledge may be noticed not only on the ontological level; it appears also confirmed by his other theses. The most important in this regard are certainly the epistemological holism and the empirical underdetermination of scientific

34

Chapter Three

theories by all available evidence. Even if holism and underdetermination do not coincide, as Quine stresses in his famous On Empirically Equivalent Systems of the World (1975),1 both seem, at any rate, to conflict with realism. Actually, holism—also known as “Duhem-Quine thesis”—is the thesis according to which the scientific hypotheses are not tested individually because our knowledge constitutes a system of highly interrelated theories and statements which compares with the experience only as a whole, as Quine pointed out in his Two Dogmas. For this model the game of science would seem a game of costs and benefits or gains and losses—a game intended to preserve as much as possible most of accepted and entrenched sentences, in order to ensure the internal consistency of our system of knowledge and its overall accordance with experience, so as to get correct observational predictions. In this connection, says Quine in Two Dogmas, we can always save our theories from falsification. Nevertheless, holism may also allow the revocability of principles firmly rooted in our web of beliefs, such as the logical law of the excluded middle. As Quine writes, any statement can be held true come what may, if we make drastic enough adjustments elsewhere in the system. Even a statement very close to periphery can be held true in the face of recalcitrant experience by pleading hallucination or by amending certain statements of the kind called logical laws. Conversely, by the same token, no statement is immune to revision. Revision even of the logical law of the excluded middle has been proposed as a means of simplifying quantum mechanics; and what difference is there in principle between such a shift and the shift whereby Kepler superseded Ptolemy, or Einstein Newton, or Darwin Aristotle? (Quine 1951, 43).

We are told that our scientific theories can be held true notwithstanding experiences to the contrary; and also that we can review any statement of our system to restore an accord with the data of experience and thus simplify our theories. If so, then it could be unlikely that our theories talk about reality: we have no assurance about this matter—and the only criteria for accepting our descriptions of the world out there seem to be consistency and simplicity of our global system of the knowledge.2 Truth, perhaps the most important property of a theory about reality, is in danger of being lost. And an instrumentalist approach to scientific knowledge seems to take the place of the realism and to undermine every chance to talk about the truth of our theories. Indeed, the concept of truth seems to be in danger of being abandoned not only as result of the epistemological (or confirmation) holism, but also as a consequence of the underdetermination thesis. What about the underdetermination, realism and truth? As Quine claims in the first lines of his On Empirically Equivalent Systems of the

Realism and Truth From a Quinean Point of View

35

World, if all observable events can be accounted for in one comprehensive scientific theory—one system of the world, to echo Duhem’s echo of Newton—then we may expect that they can all be accounted for equally in another, conflicting system of the world (Quine 1975, 313).

We have to admit this possibility, Quine adds, “because of how the scientists work”, and the scientists in fact “invent hypotheses that talk of things beyond the reach of observation” (Quine 1975, 313). Given the possibility of equivalent but conflicting systems of the world, underdetermination “sets one to wondering about truth” (Quine 1975, 327). Truth and realism, then, would be out of our reach—and perhaps wrong, irrational, and non-existent aims. We could be satisfied of predicting the same observational sentences from conflicting theories, and we can accept conflicting theories in so far as they have the same observational consequences. Empiricism, then, would exclude realism (but this is not the conclusive claim of Quine, as we shall see). We may here mention another explicit argument of Quine against realism and truth—an argument based on the idea of conceptual scheme, one of the major tenets of the Quine’s philosophy. In “Identity, Ostension, and Hypostasis” (1950), after a brief reference to the “ship of Neurath”, we can read the following remark about realism and truth: We can improve our conceptual scheme, our philosophy, bit by the bit while continuing to depend on it for support; but we cannot detach ourselves from it and compare it objectively with an unconceptualized reality. Hence it is meaningless, I suggest, to inquire into the absolute correctness of a conceptual scheme as a mirror of reality. Our standard for appraising basic changes of conceptual scheme must be, not a realistic standard of correspondence to reality, but a pragmatic standard (Quine 1950, 79).

After this remark, Quine refers in a footnote to the father, as it were, of instrumentalism and holism, Pierre Duhem. Quinean views appear then very devastating for realism and its most famous incarnation, the correspondence theory of truth. But can we actually say that? Certainly, we can; but only for misguided conceptions of realism and truth—realism as a conception regarding the description of an independent reality and the truth as a correspondence between our language and an independent reality. We will see that the Quinean versions of realism and truth are very different from these views. Quine’s arguments go in a different and new direction for a clarification of how we should conceive realism and truth.

36

Chapter Three

2. From posits to internal realism Although we can advance a (at least) prima facie hypothesis in favour of a Quinean instrumentalist conception of the scientific knowledge, there are other arguments of Quine that exclude this hypothesis and suggest a form of realism—if properly understood. But what is realism properly understood? In order to answer this question, we have first to make some reflections on the concept of posit, which is very important in the Quinean perspective. What is a posit? Physical objects are posits, as Quine held in Two Dogmas. Also atoms, electrons and subatomic particles—in other words, theoretical entities—are posits. And classes in mathematics are posits too. If so, then it would seem that nothing is real: all our ontology contains only entities that we posit or invent—entities that could be not really existing out there. Are we perhaps confined to a constructivist epistemology? The problem regarding posits is ontological and epistemological at the same time, how can we expect from Quine’s philosophy. In his Posits and Reality (1955) Quine deals with precisely the problem of the relation between our epistemic necessity of positing certain types of (theoretical) entities and the status of their reality. Are there really existing molecules and physical particles or they are merely fictions? In Posits and Reality Quine states moreover, although only as a provisional hypothesis in the context of his discussion, that the sentences which seem to propound molecules are just devices for organizing the significant sentences of physical theory. No matter if physics makes molecules or other insensible particles seem more fundamental than the objects of common sense; the particles are posited for the sake of a simple physics (Quine 1955, 250).

Is this a verdict of unreality—and fictionalism or instrumentalism— about theoretical entities? Is realism, therefore, a myth to be abandoned? To some extent we could endorse such a verdict, but only with the caveat—noted however by Quine himself—that it should then be pronounced even for ordinary physical bodies of common sense: even these, in fact, are neither more nor less than postulates introduced to organize and simplify the experiential level of our raw sensations— coloured visual stains, tactile and auditory sensations, tastes, smells, etc. The evidence we possess for asserting the existence of physical bodies is basically not much different from the one we have for asserting the existence of molecules: the postulation of physical bodies is only less sophisticated and certainly not intentionally deliberate as that of the

Realism and Truth From a Quinean Point of View

37

molecules. If the molecules are fictions, they are fictions no less than the physical objects of common sense: the positing of molecules—remarked Quine—differs from the positing of the bodies of common sense mainly in degree of sophistication. In whatever sense the molecules in my desk are unreal and a figment of the imagination of the scientist, in that sense the desk itself is unreal and a figment of the imagination of the race (Quine 1955, 250).

But the verdict of unreality, however plausible it may appear given the postulatory character of most of the entities to which our language and our scientific theories refer, is for Quine excessive and misleading. Certainly, postulation of “an esoteric supplement to the exoteric universe” (Quine 1955, 250) made by the physicist, says Quine, is a way to simplify experience and laws of nature, but it would make little sense to consider a figment that esoteric supplement, in the same way in which it would make little sense to consider figments the ordinary physical bodies and objects. Also sense data, Quine writes, are after all posits—“posits of psychological theory, but not, on that account, unreal”: the sense datum “may be construed as a hypothetical component of subjective experience standing in closest possible correspondence to the experimentally measurable conditions of physical stimulation of the end organs” (Quine 1955, 252). If that is the case even for sense data—namely for those which have been considered the most fundamental elements of our knowledge—, then we should acknowledge that posits are real entities, at least insofar as they are component part of our theories about the world. The best evidence for classifying something as real lies in its contribution to knowledge, that is, in its building of theories that are equipped with the features that make them so reliable in the first place from the point of view of simplicity and predictivity. Here’s how Quine summarizes his argument: We can continue to recognize […] that molecules and even the gross bodies of common sense are simply posited in the course of organizing our responses to stimulation; but a moral to draw from our reconsideration of the terms “reality” and “evidence” is that posits are not ipso facto unreal. The benefits of the molecular doctrine […] and the manifest benefits of the aboriginal posit of ordinary bodies, are the best evidence of reality we can ask (pending, of course, evidence of the same sort for some alternative ontology) (Quine 1955, 251-252).

This tenet is even better specified in the first chapter of Word and Object, where Quine returns to the ontological problem discussed in Posits and Reality. In his major book Quine claims that the ontological posits that

38

Chapter Three

have such an important role in our knowledge—from that of the ordinary objects of common sense to the most sophisticated physical theories— cannot in any way be considered unreal: “everything to which we concede existence is a posit from the standpoint of a description of the theorybuilding process, and simultaneously real from the standpoint of the theory that is being built” (Quine 1960, 22). In other words, if the investigation on knowledge reveals the postulatory character of scientific ontologies, in effective scientific investigation postulates are indistinguishable from hypotheses that can be confirmed or invalidated by what is usually called “scientific method”. According to Quine the question of what there is “is simply part of the question […] of the evidence for truth about the world”, and “the last arbiter is so-called scientific method, however amorphous” (Quine 1960, 22-23). Science and the scientific method are the only yardsticks of reality we possess. More importantly, we cannot consider scientific ontologies as figments because “we can never do better than occupy the standpoint of some theory or other, the best we can muster at the time” (Quine 1960, 22). This point of view may perhaps seem too simplistic, but, if we reflect about it closely, we cannot but agree. For, as Quine points out in concluding remarks of Word and Object, we—either as philosophers or as scientists— can never occupy a place outside our conceptual scheme; there is no cosmic exile from which we can judge the correctness of our theories or conceptual scheme (Quine 1960, 275). Our beliefs about how the world is and what there is in it are then internal to our conceptual scheme and our theories; so Quine is proposing once again the argument of “Identity, Ostension, and Hypostasis”: “we cannot detach ourselves from [our conceptual scheme] and compare it objectively with an unconceptualized reality”. “Real” is what to which we acknowledge existence from the point of view of the way in which we structure and organize linguistically the world out there: there is no reality in itself—mind-or conceptual schemeindependent—that we should or could copy, and perhaps such a goal would be totally illusory and irrational because reality is only the reality we can know. This is not constructivism or relativism, but epistemic and internal realism3 (even if someone, Richard Rorty for example, supposed that internal realism is in fact a form of constructivism or relativism). An issue deserves here special attention. We have previously noted that for Quine there is a parallelism between the “imagination” of the scientist and that of the race. If the particles of physics and chemistry are posits— products of the imagination of the scientist—, so, too, are posits physical objects of common sense, products of the imagination of the race. However, this parallelism is in need of some clarification. For posits of

Realism and Truth From a Quinean Point of View

39

bodies and objects of common sense have different status and origin with respect to posits of theoretical entities of mature science. In scientific theories there are reasons to postulate entities in order to simplify our experience and theories themselves. This expedient still does not seem to involve a striking difference between common sense and scientific knowledge and theories: recall in this respect the famous statement of Two Dogmas that “science is a continuation of common sense, and it continues the common-sense expedient of swelling ontology to simplify theory” (Quine 1951, 45). But, actually, there is a difference between common sense and science, as Quine stated in Word and Object; here he distinguishes between two types of theories about reality: the natural theory of common sense realism and the deliberate theory of scientific knowledge: “theory may be deliberate, as in a chapter on chemistry, or it may be second nature, as is the immemorial doctrine of ordinary enduring middle-sized physical objects” (Quine 1960, 11). In spite of his continuity thesis, Quine acknowledges therefore a fundamental difference between the two types of theories. True, the natural beliefs about reality of entities of common sense are beliefs about posited entities, as are scientific beliefs too, but only to a certain extent: in fact, we are creatures naturally disposed to believe in bodies and physical objects, and this disposition is natural because it is based on our biological (and psychological) features. As Quine states in The Roots of Reference, “man is a body-minded animal, among bodyminded animals”, and man and other animals “are body-minded by natural selection; for body-mindedness has evident survival value in town and jungle” (Quine 1974, 54). The same point is stressed even more emphatically in From Stimulus to Science (1995): here Quine argues that in fact, long before the reification by the development of language (definite articles, pronouns, plurals, existential quantification), bodies commanded the special attention of our remote ancestors, as they do that of modern man from early infancy. Natural selection had heightened man’s responsiveness and that of other animals to the sight and smell of bodies, for there is survival value in prompt awareness of prey and predator (Quine 1995, 35).

Hence, we are animals naturally, biologically inclined to believe in the existence and reality of bodies out there; in other words, we are realist from the beginning of our life, at phylogenetic as well as ontogenetic level. Moreover, as Quine had claimed in the famous first page of Word and Object, This familiar desk manifests its presence by resisting my pressures and by

40

Chapter Three deflecting light to my eyes. Physical things generally, however remote, become known to us only through the effects which they help to induce at our sensory surfaces. Yet our common-sense talk of physical things goes forward without benefit of explanations in more intimately sensory terms. Entification begins at arm’s length; the points of condensation in the primordial conceptual scheme are things glimpsed, not glimpses (Quine 1960, 1).

Our realistic attitude is therefore rooted in our biological and psychological-perceptive nature (stimulations of our exteroceptors). We are creatures naturally predisposed to believe in the reality of external objects, regardless of whether the existence of such objects precedes or not our epistemic processes and our language or conceptual scheme, where the realist attitude is manifested through the devices for reification. Mature science is then a conscious and deliberate extension by analogy of our ontological beliefs of common sense; and in this sense scientific theories make broader the domain of reification and objective reference, extending our natural attitude toward entification from bodies to hypothetical entities of physics and chemistry. As Quine argues in From Stimulus to Science, Reification has continued beyond bodies and substances, adding new sorts of objects and new quantified conditionals to tighten the logical structure of science. Atoms tentatively entered the ontology already in antiquity, in easy analogy to primordial bodies. Abstract objects were reified even earlier, it would seem, though there was nothing so suggestive in the way of analogy (Quine 1995, 39).

The continuity thesis, then, presents great plausibility from the linguistic and epistemological point of view. We come to our science, as we know it, by the extension of a natural ontological attitude. It seems, therefore, that the Quinean realism is an example of what Arthur Fine (1984) has called the Natural Ontological Attitude (NOA). It is true that this affinity can seem surprising when one takes into account the fact that Quine—except for a few brief remarks—always kept au dessus de la mêlée with respect to the debate on realism and that there seems to be no direct influence of Quine on Fine. But this, at bottom, makes the affinity even more significant. Actually Fine seems to support a Quinean view on realism, though not an explicitly naturalistic one. According to Fine both the realist and anti-realist (the instrumentalist or the conventionalist), beyond their radically different theories of truth and ontology, share (and should share) a “core position”, a “minimal stance” regarding the acceptance of the truths of common sense and those of scientific theories, that is, the belief in the objects of common sense and of

Realism and Truth From a Quinean Point of View

41

entities postulated by scientific theories (given the reliability of scientific methods to confirm the hypothesis). Here’s how Fine clarifies his view: I certainly trust the evidence of my senses, on the whole, with regard to existence and features of everyday objects. And I have similar confidence in the system of “check, double-check, check, triple check” of scientific investigation, as well as the other safeguards built into the institutions of science. So, if the scientists tell me that there are molecules, and atoms, and /J particles, and, who knows, maybe even quarks, then so be it. I trust them, and thus must accept that there really are such things with their attendant properties and relations. (Fine 1984, ed. 1996, 35)

The NOA, as Fine claims, is common to the realist and the anti-realist: both “accept the results of scientific investigations as ‘true’, on a par with more homely truths” (Fine 1984, 36). This is the “core position” of the NOA, but the realist and the antirealist go on to expand upon it with rhetorical arguments that are only philosophical superfetations, useless additions to the reasonable thesis that reality is, except for errors, that of the common sense and the one to which the science refers. More importantly, Fine sees only “superficial decorations” in the philosophical doctrines that appeal to the correspondence theory of truth as the semantic counterpart of the realism, thus claiming to occupy a point of view outside the “game” of science from which to judge whether the theories correspond, at least approximately, to an independent reality (Fine 1984, 39). But our use of the concept of truth has nothing to do with this unlikely or impossible task, and Fine states that it is sufficiently explained by “Davidsonian-Tarskian referential semantics” (Fine 1984, 41). (But he could also have referred to the Quinean Disquotational Theory of truth).

3. Truth and realism: a minimalist account Quine’s views about realism and Fine’s natural ontological attitude appear very similar in their main assumptions. However, Fine does not consider his NOA as a form of realism, but only as a natural cognitive attitude that is independent from the realism-antirealism debate. And yet it cannot be excluded that differences between NOA and Quinean realism are perhaps only terminological. Fine does not want to enter the realismantirealism debate, which he considers unfounded, nor does he want to take a position in favour of a discredited metaphysical realist conception: he rejects the old and outdated debate on the independent and an sich reality of the external world that our theories should mirror; he claims that we cannot do anything but accept the best available theories—and NOA is

42

Chapter Three

conceived as a kind of third way between realism and anti-realism, something quite different from the traditional solutions in the debate on realism. But this is precisely the argument put forward by Quine when he states, in Word and Object, that “we can never do better than occupy the standpoint of some theory or other, the best we can muster at the time”, being real just those objects to which such theories ascribe existence. And this view is a form of realism—not metaphysical but internal or epistemic realism. And truth? What about the truth? The truth has traditionally been conceived as the mark of realism, as correspondence with an independent reality. We know that Quine has always rejected the correspondence theory of truth, and that for this reason he has often been considered an instrumentalist, as we have already noted. But there is some comments to be made on this point. Reality as we know it is not a structure completely independent from our conceptual scheme—it cannot be so, because we can consider real only what our language and our best theories, confirmed by our sensorial stimulations and scientific method, say that there exists out there and which statements are true. In other words, we cannot have access to a “cosmic exile” from which to assess the correspondence of our theories to an independent reality; nor can we, as Hilary Putnam stressed, take the externalist point of view of “God’s Eye” that could contemplate the way the world is in itself, independently from any observer and free from any contamination of our concepts and conceptual scheme (Putnam 1981, 49). These two metaphors are very similar, and they both suggest the same idea: the impossibility of metaphysical realism implied by correspondence. If we do not want to give up the notion of truth, as does the relativist, we have then to look in another direction—different from the metaphysics of correspondence—for talking of truth. Is it possible to make sense of the realist notion of truth in an internalist conception of the realism? The concept of truth is obviously indispensable, and internal realists, as Putnam and Quine, would not want to give up it. Which is the more plausible internalist surrogate of truth? In his famous manifesto of internal realism Putnam was talking of “rational acceptability in ideal conditions” (Putnam 1981, 55)—an epistemic view according to which a statement is true when it is justified under such ideal conditions. However, this thesis appears somewhat vague; according to some critics it also appears very similar to the concept of the truth that we can find in Charles S. Peirce (1901)—the truth as the property of an ideal theory to which you can approach in the long run, by the scientific method, only as a limit. (But we can recall that Putnam rejects his alleged affinity with Peirce’s theory: see

Realism and Truth From a Quinean Point of View

43

Putnam 1990, VII-VIII). Crispin Wright pointed out the limits of the ideal conditions of which Putnam had spoken and their affinity with the Peircean theory: simply, we could not know, for example, when “any mythical limit” of ideal conditions is realized—and so we would not have necessary and sufficient conditions for stating if our statements are true or not (Wright 1994, 46-47). However, if it is true that when you talk of internal realism you think in the first place of Putnam’s views, in fact the seminal idea of internalism was already in Quine,4 and in the Quinean internalist perspective there do not arise any problems in respect to alleged epistemic ideal conditions under which justify our statements in order to assess their truth. From the Quinean point of view Peirce’s doctrine does not provide a clear and plausible explanation of our use of the predicate “true”; and his objection may perhaps also be applied to Putnam’s theory of ideal epistemic conditions. According to Quine we usually talk, and we can talk, of the truth of our statements or theories without necessarily having to resort to an ideal theory, contrary to what Peirce thought (and later Putnam with ideal conditions). Quine had already stated in the first chapter of Word and Object, significantly entitled “Language and Reality”, that “it is rather when we turn back into the midst of an actually present theory, at least hypothetically accepted, that we can, and do speak sensibly of this and that sentence as true” (Quine 1960, 24). Theories and the scientific method are sufficient for Quine in order to apply the predicate “true” to our statements. Here’s how Quine clarifies his standpoint: Where it makes sense to apply “true” is to a sentence couched in the terms of a given theory and seen from within the theory, complete with its posited reality. Here there is no occasion to invoke even so much as the imaginary codification of scientific method. To say that the statement “Brutus killed Caesar” is true, or that “The atomic weight of sodium is 23” is true, is in effect simply to say that Brutus killed Caesar, or that the atomic weight of sodium is 23. That the statements are about posited entities, are significant only in relation to a surrounding body of theory, and are justifiable only by supplementing observation with scientific method, no longer matters; for the truth attribution are made from the point of view of the surrounding body of theory, and are in the same boat (Quine 1960, 24).

Although laconic and complex, this passage of Word and Object contains at least three fundamental principles of Quinean epistemology: i) an epistemic (and intratheoretical or internalist) view of truth and realism; ii) a semantic definition of the concept of truth—the famous Disquotational Theory of truth; iii) a reaffirmation of the basically holistic character of

44

Chapter Three

scientific inquiry (to which refers the expression “the surrounding body of theory”, with an implicit reference to the “boat” of Neurath). As a matter of scientific rationality, according to Quine, we simply accept as true those statements that our theories and experimental results require us to accept (also in the hypothetical case of competing theories by underdetermination).5 We have no choice in this: scientific inquiry is the most rational means for attain the end of the truth; but consider a sentence true means nothing more than affirm or assert it: as the Disquotational Theory of Truth displays (“p” is true if and only if p), there is no difference between saying that a sentence is true and affirming it. This minimalist account of the concept of truth is “the valid residue of the correspondence theory of truth”, as says Quine in his Pursuit of Truth (1992a, 93), but obviously it avoids the difficulties of the venerable and metaphysical concept of correspondence. Quine is then returned to discuss this topic in Pursuit of Truth, in which he tried to explain it in more detail. “If to call a sentence true is simply to affirm it, then how we can tell whether to affirm it?” (Quine 1992a, 93); Quine asks this question as a possible objection to his theory of truth by those who might remain perplexed by his conception excessively minimalist of truth. His answer, as we might expect, is that in order to assert (as true) a certain statement we have to refer not to a metaphysical and undetectable correspondence of our statements with a mind-independent reality, but simply to our best scientific theories and methodological techniques of testing our hypotheses: “the answer is a general analysis of the ground of warranted belief, hence scientific method” (Quine 1992a, 93). As in Word and Object, Quine is appealing then to the methodological criteria that warrant acceptability of our hypotheses. In particular, successful predictions of observable conditionals deduced from theories and hypotheses—according to the hypotheticaldeductive method—provide the best reasons for asserting (the truth of) a hypothesis. This is a reaffirmation of Dewey’s theory of scientific inquiry, based on the warranted assertibility, an expression that should be preferred according to Dewey to the more ambiguous terms “knowledge”, “belief” (Dewey 1938, 9-10) and perhaps “truth”. Truth is then nothing but disquotation on the semantic level, and (warranted) assertibility on the epistemic one, namely affirmation of a sentence in so far as it is justified by successful results (predictions) of the application of scientific method. Quine’s stance is thus a Deweyan and pragmatic perspective. Supporting a pragmatic perspective does not mean giving up the concept of truth, but solely the traditional and untenable concept of truth as correspondence. As Quine had explained in Word and Object, “within

Realism and Truth From a Quinean Point of View

45

our own total evolving doctrine, we can judge truth as earnestly and absolutely as can be, subject to correction, but that goes without saying” (Quine 1960, 25). Fallibilism is thus a fundamental feature of scientific research of truth, and our attributions of truth are always defeasible: in other words, they are acceptable only until there is no evidence to the contrary. As Quine says in From Stimulus to Science, “we should and do currently accept the firmest scientific conclusions as true, but when one of these is dislodged by further research we do not say that it had been true but became false. We say that to our surprise it was not true, after all” (Quine 1995, 67). And we say this because this is the “idiom of realism”, that “is integral to the semantics of the predicate ‘true’” (Quine 1995, 67). Quinean minimalism about truth is on a par with minimalism about realism: as we cannot hope for something deeper than disquotation and warranted assertibility for the concept of truth, so we cannot require a form of realism deeper than epistemic or internal realism. But why should this be so? The Quine’s answer is naturalism, “the recognition that it is within science itself, and not in some prior philosophy, that reality is to be identified and described” (Quine 1981b, 21). But here’s how Quine explained, in even more explicit and vivid terms, his point of view on realism and its relations with naturalism: Naturalism looks only to natural science, however fallible, for an account of what there is and what what there is does. Science ventures its tentative answers in man-made concepts, perforce, couched in man-made language, but we can ask no better. The very notion of object, or of one and many, is indeed as parochially human as the parts of speech; to ask what reality is really like, however, apart from human categories, is self-stultifying. It is like asking how long the Nile really is, apart from parochial matters of miles or meters. [...] The world is as natural science says it is, insofar as natural science is right; and our judgement as to whether it is right, tentative always, is answerable to the experimental testing of predictions (Quine 1992b, 9).

If this argument is right—as we may grant—then the traditional (i.e. metaphysical) concepts of realism and truth are no more than two other dogmas.

References Caruso, G. 2007, “Realism, Naturalism, and Pragmatism: A Closer Look at the Views of Quine and Devitt”, Kriterion. Journal of Philosophy 21, 64 Dewey, J. 1938, Logic, the Theory of Inquiry, New York: H. Holt and Co.

46

Chapter Three

Fine, A. 1984, “The Natural Ontological Attitude”, repr. in D. Papineau, ed., The Philosophy of Science, Oxford-New York: Oxford University Press, 1996, 21 Hoefer, C. and A. Rosenberg 1994, “Empirical Equivalence, Underdetermination, and Systems of the World”, Philosophy of Science 61, 592 Peirce, C.S. 1901, Truth and Falsity and Error, in Baldwin J.M. (ed.), Dictionary of Philosophy and Psychology, vol. 2, 716; repr. in Peirce C.S., Collected Papers, Cambridge (Mass.): Harvard University Press, vol. 5, 565 Putnam, H. 1979, Meaning and the Moral Sciences, London: Routledge & Kegan Paul —. 1981, Reason, Truth, and History, Cambridge: Cambridge University Press —. 1990, Realism with a Human Face, ed. by J. Conant, Cambridge (Mass.): Harvard University Press Quine, W.V. 1950, “Identity, Ostension, and Hypostasis”, repr. in Quine 1953, second edition revised 1980, 65 —. 1951, “Two Dogmas of Empiricism”, repr. in Quine 1953, second edition revised 1980, 20 —. 1953, From a Logical Point of View, Cambridge (Mass.): Harvard University Press; Second Edition, Revised, 1980 —. 1955, “Posits and Reality”, repr. in W.V. Quine, The Ways of Paradox and Other Essays, Revised and enlarged edition, Cambridge (Mass.): Harvard University Press, 1976 —. 1960, Word an Object, Cambridge (Mass.): MIT Press —. 1970, Philosophy of Logic, Englewood Cliffs (N.J.): Prentice-Hall —. 1974, The Roots of Reference, La Salle (Ill.): Open Court —. 1975, “On Empirically Equivalent Systems of the World”, Erkenntnis, 9, 313 —. 1981a, Theories and Things, Cambridge (Mass.): The Belknap Press of Harvard University Press —. 1981b, “Things and Their Places in Theories”, in Quine 1981a, 1 —. 1992a, Pursuit of Truth (Revised Edition), Cambridge (Mass.): Harvard University Press —. 1992b, “Structure and Nature”, The Journal of Philosophy 89, 5 —. 1995, From Stimulus to Science, Cambridge (Mass.): Harvard University Press Wright, C. 1994, Truth and Objectivity, Cambridge (Mass.): Harvard University Press

Realism and Truth From a Quinean Point of View

47

Notes 1

On this topic see Hoefer-Rosenberg 1994. But we may here recall that in Quine 1970 a revised version of the holism was formulated, less radical and more moderate in respect to that of Two Dogmas. 3 For a similar consideration see Caruso 2007. 4 Besides that in the Kantian theory of knowledge, as Putnam pointed out in Reason, Truth, and History. 5 In this case we may adopt, according to Quine, two different attitudes: a sectarian one or an ecumenical one. For the sectarian attitude we should reject the alien competing theory and retain our own old theory, to be considered as true. For the ecumenical one we could accept both theories. and count both true (Quine 1992a, 99-100). In the latter case we should admit that “we grasp the world variously”, as Quine acknowledges (Quine 1992a, 101). It is no clear, however, what attitude Quine prefer, but we can recall that in concluding “On Empirically Equivalent Systems of the World” he defended the sectarian one. As regard the ecumenical attitude, we can observe here that it is perhaps comparable to the Putnam’s famous thesis of the “equivalent descriptions” (see Putnam 1979, 50), which was, however, influenced by Quine’s theses of underdetermination and of the indeterminacy of radical translation. 2

CHAPTER FOUR WHY ANTIREALISTS CAN’T EXPLAIN SUCCESS MARIO ALAI

1. The “no miracle argument” Scientific realists maintain that the truth of theories is the only plausible explanation for their empirical success: it would be a miracle if theories were as successful as they are, and yet radically false (Smart 1968, 150; Putnam 1975, 73). Antirealists have countered by proposing a number of purported alternative explanations short of the truth of theories. Here I review some of these alternatives, arguing that they all fail. The success of a theory T may be conceived sociologically as widespread acceptance, capacity to draw funds, etc.; but all this follows from the confirmation of its predictions, and this is due, ceteris paribus, to their truth. So, T can be safely considered successful when it makes true rather than false predictions, i.e., predictions of actual phenomena or features of phenomena (Alai 2012, 88-89). When the predicted phenomena are previously known and used in building the theory, there is an obvious explanation of the predictive success of T: the theorist knew them, and wanted T to predict them; hence, she shaped or accommodated T and supplied it with the right auxiliary assumptions so that T yielded precisely those consequences. This explanation does not require that T is true, for in principle any given set of consequences can follow from infinite false theories. However, the above explanation is not available when the predicted phenomenon was not known, or not used (at least not essentially),1 in constructing T. If a new phenomenon is sufficiently homogeneous to those known and used, its prediction by T can still be explained by the general scope of theories: they are not intended to apply only to the restricted set of examples originally considered by the theorist, but also any system or phenomenon sufficiently similar to them; moreover, because of the uniformity of nature, if predictions on the old phenomena are true,

Why Antirealists can’t Explain Success

49

probably those on the new but similar ones will be also true. Or T may predict a phenomenon which is new but sufficiently homogeneous to the old ones, simply because the theorist extrapolated it from the known data, and accommodated the theory to predict it. Hence, these predictions cannot confirm much. For instance, it would be hardly surprising if a theory T based, among other data, on certain chemical and physical properties of carbon-12 and carbon-13, predicted analogous properties for carbon-14. This is to say, of course, that the newly predicted properties in fact are not altogether new (Alai 2014b, § 3.3). Also, the logically weaker is a conclusion, the more are the premises from which it follows (in the limiting case, empty conclusions follow from any premise). So, if a prediction is uninformative or very probable a priori, the theorist may have just picked by chance one of the many theories (even false ones) from which it follows. For instance, the almost trivial claim that there exists a hitherto unknown planet somewhere in the universe (as opposed to the prediction of a new planet (Neptune) with specified mass and orbit in a particular region, as was licensed by Newton’s gravitation theory) would follow from most astronomical theories, including false ones (Alai 2014b, § 3.2; Alai forthcoming, § 4). But some new predictions are also totally unexpected, i.e., both completely heterogeneous from anything that had been observed before (such as, e.g., the bright spot in the shadow cast by an opaque disk predicted by Fresnel’s wave theory) and highly precise and informative, hence a priori improbable (e.g., the precise properties of Neptune, or the value of 1159652359x10-12 predicted by quantum electrodynamics for the magnetic moment of the electron, while that obtained by experiment is 1159652410x10-12: Wright 2002). So, the realists claim that these unexpected predictions cannot be explained by accommodation, extrapolation or chance, but only by assuming that theories are true (at least approximately and/or partially:2 henceforth I will always speak of truth with this implicit qualification). Paradigmatic examples, beside the mentioned ones, are the new chemical elements with precise properties predicted by Mendeleev’s periodic table; the extraordinary and precisely predicted power of atomic reactions; the retard of clocks in motion predicted by Special Relativity; the bending of light in gravitational fields predicted by General Relativity; the background radiation predicted by the Big Bang theory; etc. Here I shall call novel a phenomenon which is (a) unknown or not used essentially in constructing the theory, (b) heterogeneous to old phenomena and (c) a priori improbable. Thus, the kind of success that according to

50

Chapter Four

realists can be explained only by the truth of theories may be schematized as (S1) T predicts the novel phenomenon np. It might be objected that (S1) needs no explanation: for that T predicts the particular phenomenon np, is merely a logical fact about T (White 2003; Rees 2012, 302). But when seeking to explain the success of T, we are not referring to T transparently, as the theory which in fact it is, but opaquely, as the theory which the theorist came up with.3 In other words, the explanation we need is not so much about the theory, as about the theorist which found it: (Q1) how was the theorist able to conceive a theory which (together with appropriate auxiliary assumptions) predicted np, although it was novel? Or in other words, how among the infinite theories compatible with the old data, the theorist picked one which entailed even some novel ones? There is no question that among the infinite possible theories conceived as eternal objects in a Platonic hyperuranion there are some having the particular consequence “np”. But they are an infinitely small minority: so, how was a human theorist able to chose just one of them, without essentially using np? She knew she should pick among those compatible with the old data. But even among them, for any theory predicting np there are infinite others not predicting it. For the realists there is just one possible answer: the theorist picked T because she sought for a true theory, and T is true!

2. Trivial antirealist explanations But antirealists have claimed that the realist explanation is not the only possible one: alternative proposals are for instance (I) Empirical adequacy (van Fraassen 1980, 12): i.e., that the theory is compatible with all the empirical phenomena, or , “what it says about observable things and events […] is true”, or (II) Surrealism (Fine 1984, Leplin 1987): i.e., that all empirical phenomena are as if the theory were true. However, neither (I) nor (II) explain (S1) (that T predicts np), for they just say that T is compatible with all empirical evidence; but a theory may

Why Antirealists can’t Explain Success

51

be compatible with all the data without predicting all the data, in particular, without predicting np. For instance, a theory with no empirical content would be empirically adequate! So, neither (I) nor (II) entail (S1), nor a fortiori they answer question (Q1), nor explain T’s success. The same problem is encountered by the quickest formulation of the realist explanation, that the theorist sought for a true theory: T may be true, without entailing all the truth, in particular, without entailing “np”. So, we shall have to improve that formulation. (S1) would be entailed by assuming (III) Empirical perfection, i.e., that T predicts all the empirical data. But (III) is utterly implausible, for empirical perfection is not feasible for humans. Moreover, there is no need to assume empirical perfection in order to explain just a few actual predictions. Thus, more modest proposals have been advanced, like (IV) Predictive similarity i.e., that T makes predictions very similar to those of the true theory (Stanford 2000, 272 ff.). But which true theory is (IV) about? There are many true theories which do not predict np. So, even (IV) does not entail (S1). One might then try (V) Predictive similarity* i.e., that T makes predictions very similar to those of a true theory which predicts np. But saying that T predicts np because (V) is tantamount saying that T predicts np because T predicts a number of things including np. So, although (V) entails (S1), it does so trivially. Therefore, it is not an explanation, but just a repetition of the explanandum.4 In other words,5 it does not answer our actual question (Q1) how was the theorist able to find a theory predicting np, although it was novel? A possible alternative is (VI) Modest surrealism (Lyons 2002, 78), i.e., that (a) the mechanisms postulated by the theory would, if actual, bring about all the actual phenomena observed, and some yet to be observed, at time t; and (b) these phenomena are brought about by actual mechanisms in the world. In other words, (a) explains that T is successful because it (no matter whether true or false) entails all the phenomena observed up to now

Chapter Four

52

(including np), plus others to be observed later. But this is like saying that T predicts np because it predicts np plus other phenomena, both observed now and to be observed later. So, again, it entails (S1), but rather than explaining, it merely restates it, and it does not answer (Q1). Besides, all the content of (VI) exceeding (S1) is irrelevant to explaining (S1): that T predicts some phenomena to be observed in future does not explain why it predicts past phenomena; and (b), the assumption that the phenomena predicted by T are brought about by actual mechanisms in the world, is just a very preliminary and generic step toward explaining those phenomena (it only says that the full fledged explanation will have to be realistic, not phenomenalistic), but it says nothing on why T predicts them. In general, no “as if” account really explains: “the hypothesis that it is raining explains why the streets are wet—but ‘The phenomena are as if it were raining’ does not” (Musgrave 2006-2007). Similar problems trouble Fine’s (VII) “instrumental reliability” (1986, 1991): for, as noticed by Psillos (1999, 92-93), if “instrumental reliability” simply means that theories are “useful in getting things to work for the practical and theoretical purposes for which we might put them to use” (Fine 1991, 86), it is nothing than success, so the explanation is tautological; if it means a disposition to be instrumentally successful, then the explanation is empty until one spells out the intrinsic property or properties which endow theories with that disposition (a role for which truth would be the most natural candidate). And in any case, one should still explain how instrumentally reliable theories can be found (a question analogous to (Q1)).

3. Attempts to explain success characterized as truth of empirical consequences It might be objected that success could be described in a different way: we have seen that after learning that the novel prediction “np” has been confirmed, we can describe success as (S1), i.e. T’s property of predicting an actual phenomenon although it was not used in constructing T; hence the question is (Q1), how the theorist found a theory with that property. But if we look at T from the point of view of when it was put forward, before its predictions were tested, then its success can also be described as (S2) T’s novel prediction “np” eventually turned out to be true, and then the question is

Why Antirealists can’t Explain Success

53

(Q2) Why did “np” eventually turn out to be true? So, it might be claimed, even if not explaining (S1) or answering (Q1) (how a theory predicting np was found), assumptions (I)-(VI) explain at least (S2), answering (Q2): (I) and (II) say that all empirical consequences of T are true, and since “np” is an empirical consequence of T, this explains why “np” turned out to be true. (IV) and (V) also imply that all empirical consequences of T are true, for T has the same consequences of the true theory, and all the consequences of a true theory are true. Probably the claim that all of T’s empirical consequences are true is too strong to be plausible, but the antirealist could equally well answer (Q2) by the more modest claim that (S3) most of T’s empirical consequences are true. Criticisms of Popper’s notion of verisimilitude (Tichy 1974, Miller 1974a, 1974b) pointed out that if a theory has at least one false consequence, then it has at least as many false consequences as true ones (all the conjunctions of the one false consequence with the true ones); and if a theory has at least one true consequence, then it has at least as many true consequences as false ones (all the disjunctions of the one true consequence with the false ones); so, it might be claimed that (S3) is impossible. But this is a mere logical artifice, by which the one false (or true, respectively) consequence is counted over and over. As argued by Schurz and Weingartner (2010, 425), if T has the true consequence (i) “grass is green” and the false consequence (ii) “snow is red”, the disjunction (iii) “grass is green or the moon is made of cheese” is not a further truth predicted by T beside (i); equally, the conjunction (iv) “grass is green and snow is red” is not a further falsity of T beside (ii). In other words, the set of claims {(i), (ii), (iii), (iv)} has the same content (e.g., it has the same models, or excludes the same possible worlds) as the set {(i), (ii)}. Hence, (iii) and (iv) are not relevant in counting the true and false consequences of T. Therefore (S3) is possible if by “empirical consequences” we mean only the relevant ones, as I shall do in the rest of the paper.6 Still, even an answer to (Q2) would leave unsolved the original question (Q1), namely how the theorist found a theory which predicted a novel phenomenon like np without taking it into account. Moreover, answering (Q2) by (S3) would be like saying that “np” turned out true because (a) “np” is true, and (b) many other consequences of T are true; so, again, it would not be an explanation, but a restatement of the explanandum (S2) with the useless addition (b). In fact, (S3) requires in turn an explanation:

54

Chapter Four

(Q3) why are most of T’s empirical consequences (including novel ones like “np”) true? Realists need not accept a still rather optimistic claim as (S3): for they may be content to claim that T is partially true, so having also many or most false empirical consequences. But should realists grant (S3), they would readily answer (Q3) by explaining that T is largely true, and a largely true theory has mostly true consequences. For the antirealists, however, (Q3) can be answered even without appeal to truth: most of T’s empirical consequences are true because it is a logical fact about T that it has those particular consequences, and they are true because actual phenomena happen to be just as they describe them. But this trivial “disquotation” explanation will not be enough: for recall, in this context we are not considering T transparently, as the theory which in fact it is, but opaquely, as the theory which the theorist was able to conceive. So, our question is now (Q4) how was the theorist able to find a theory most of whose empirical consequences (including novel ones like np) turned out to be true? Of course, if a theory predicted only previously known phenomena, the answer would be easy: by building or accommodating the theory so that it made exactly the predictions known to be true. But when novel phenomena are predicted, this answer is not enough. However, making a true empirical prediction is just predicting an actual phenomenon: so, (Q4) is practically a generalization of (Q1) how was the theorist able to find a theory predicting np, although it was novel? Realists, therefore, have ready the same answer: the theorist found it because she sought for a true theory, she was enough skilled to find one, and true theories have mostly true consequences. Antirealists of course reject this appeal to truth, but as we shall see shortly, there is no alternative explanatory strategy available to them. (III) and (VI) seem to say a little bit more toward an explanation of (S3): (III) says that T predicts all the actual empirical data or phenomena, and (VI) that it predicts all the observed phenomena, and something more. So, they suggest that at least many of T’s empirical predictions correspond to actual phenomena, i.e., they are true. But this is like explaining (S3), the truth of most of T’s consequences, by claiming that (S3') most of T’s empirical consequences correspond to actual phenomena, again a trivial explanation: for then we need to know

Why Antirealists can’t Explain Success

55

(Q3') Why most of T’s empirical consequences (including novel predictions like np) correspond to actual phenomena? which again, on pain of triviality, must be understood as (Q4') how was the theorist able to find a theory most of whose empirical consequences (including novel ones like np), correspond to actual phenomena? which is equivalent to (Q4). Summing up, merely entailing the truth of empirical predictions is not enough to explain success, for the very characterization of success as the truth of empirical predictions is insufficient: the success to be explained is also, to begin with, the making of those predictions (i.e., the finding of a theory which make them) in the face of their novelty.

4. The Deflationist Explanation Another antirealist proposal based on the concept of success as truth of the empirical consequences is the (VIII) Deflationist explanation (Fine 1984, Hacking 1983, 64), i.e., that success is already explained by theories themselves: in fact, the success of T is the truth of its empirical consequences, and this is equivalent to the obtaining of certain phenomena; but the obtaining of those phenomena is already explained by T. In detail, instead of explaining (R) T is successful because T is true, one can explain (DE1) e1, e2, … en take place because t1, t2, … tn, where “e1, e2, … en” stand for a description of the phenomena explained by T, and “t1, t2, … tn” for T’s theoretical claims; but from (DE1) by semantic ascent we get (DE2) “e1, e2, … en” are true because t1, t2, … tn, and from (DE2), by the definition of success as the truth of empirical consequences, we get (DE3) T is successful because t1, t2, … tn. Thus, for deflationists, the claim that T is true, or that the entities it postulates are real, would be an unnecessary addition.

56

Chapter Four

However, X can be said to explain Y in two senses: if X is a fact, an event, a person, or an object, it can explain Y in the sense of being the cause, or part of the cause, or the agent, bringing about Y. Otherwise, if X is a person or a theory, or an account, X can explain Y by describing the cause or agent that brought about Y. So, theory T explains e1, e2, … en by describing the causes that brought about e1, e2, … en, and it does so by claiming that t1, t2, … tn. But claiming that t1, t2, … tn is equivalent by semantic ascent to claiming that “t1, t2, … tn” are true. Hence, theories explain certain phenomena in as much as they claim that a particular account (their own) is true; they cannot be taken as explanations without assuming they are true: as Putnam says (1978, 37), scientific realism is just science taken at face value. Moreover, even if T can explain (in the above sense) why its empirical predictions are true, it obviously cannot explain how it was possible to conceive a theory like T, some of whose consequences, including some novel ones, are true: so, it leaves question (Q4) unanswered. Finally, as we noticed above, the reconstruction of T’s success as the truth of “np” (or of T’s empirical consequences in general) is incomplete, for the success did not consist just in making a true prediction whatsoever, but in predicting the particular phenomenon np, quite improbable and heterogeneous to old phenomena. On this the deflationist proposal says nothing, so it does not answer the crucial question (Q1) how was the theorist able to find a theory predicting np, although it was novel? Since the definition of success as truth of the empirical consequences used in the inference from (DE2) to (DE3) is at least partial,7 so (DE3) is an unsatisfactory answer.

5. The realist explanatory strategy Summing up, explaining success is answering (Q1) how was the theorist able to find a theory predicting np, although it was novel? If empirical success is not limited to “np”, but extends to most predictions of T, then the question is (Q4) how was the theorist able to find a theory most of whose empirical consequences (including novel ones like “np”), turned out to be true?

Why Antirealists can’t Explain Success

57

(Answers to (Q1) and (Q4) would then provide answers to (Q2) and (Q3) respectively). We saw that antirealist proposals do not address these questions, offering only trivial explanations, while I anticipated that realism has plausible answers. Let us now analyze the explanatory strategy of realism, and see why no parallel strategy is available for antirealists. As noticed, ad hoc accommodation, extrapolation from old data, and chance are not viable explanations for the prediction of novel phenomena (i.e. not used essentially in building T, a priori improbable and heterogeneous to old phenomena). Maher’s (1988) and White’s (2003) explanation is that the theorist aimed at a true theory, and in so doing she employed a reliable method. Barnes’ (2008, 137) explanation is that in choosing to endorse T the theorist relied on true background theories. These explanations of course can be combined: the theorist sought a true theory by reliable methods and relying on true background beliefs. But this explanation is still incomplete: first, as noticed above, a theory might be true, but fail to have many empirical consequences, especially novel ones like “np”: only when the theoretical content is deep and fruitful (of great generality, scope, unifying power, etc.) theories will have many far reaching and unforeseen empirical consequences. Second, just reliably searching for a true theory per se does not make it probable finding a theory with true consequences: it makes it probable only in as much as it makes probable finding a true theory. Moreover, reliability is not infallibility: one might employ true background beliefs and reliable methods, but fail to achieve the truth or anything close to it, and so fail to make true predictions like “np”.8 So, we need to assume something which directly confers some probability to (S1) T predicts the novel phenomenon np, or more generally, if this is the case, to (S3) most of T’s empirical consequences are true. That is, we need to assume that (R1) the theorist actually found a theory with enough true, deep and heuristically fecund theoretical content to have the novel “np” as a consequence, or, if it is the case, (R1') the theorist actually found a theory with enough true, deep and heuristically fecund theoretical content to have mostly true empirical consequences, including some novel ones like “np”.

58

Chapter Four

However, Fine objected that in itself truth explains success just in the sense that it entails empirical adequacy, or instrumental reliability, and these entail success; hence truth is superfluous, one could just assume empirical adequacy or instrumental reliability (1986, 153-154; 1991, 8283). This is right; but the problem is that neither truth by itself nor empirical adequacy or similar notions answer (Q1) and (Q4), i.e., how theories with such special properties are found. Now, we shall see that the assumption of truth is a by-product of the answer to (Q1) and (Q4). Besides, finding a theory with some (or much) true content is difficult, much more than finding an empirically adequate one: since in the hyperuranion of logically possible theories, for any theory with a given true content there are infinitely many totally false but empirically equivalent ones. So, such a theory cannot be found by chance.9 Moreover, theoretical truth cannot be empirically ascertained, like empirical adequacy. Therefore, (R1) and (R1') would be unsatisfactory and implausible, unless we can in turn explain them. In fact, they still don’t answer (Q1) and (Q4): i.e., they don’t explain (Q5) how has the theorist picked a theory with enough true and heuristically fecund theoretical content to predict the novel phenomenon np, rather than one of the infinite theories compatible with the old data but too poor of true content to predict np? or (Q5') how has the theorist picked a theory with enough true and heuristically fecund theoretical content to make mostly true predictions, including novel ones like “np”, rather than one of the infinite theories compatible with the old data but for the rest making mostly false predictions? Here is where Maher’s, White’s and Barnes’ answers play an important role: the theorist found such a theory because (R2) The theorist (i) aimed at a true—but also deep and heuristically fecund—theory, (ii) skilfully employing a reliable method (i.e., truth-conducive and heuristically effective) (iii) and relying on mostly true background beliefs But if the reliability of scientific method is not to remain an empty explanation, like Molière’s virtus dormitiva, it should be explained in its turn. For instance, it should be spelled out as follows: (R3) Scientific method is reliable because

Why Antirealists can’t Explain Success

59

(i) it does not seek only synthetic descriptions of experience, but theoretical explanations with consequences reaching well beyond the observed phenomena; (ii) it respects empirical constraints; (iii) assuming that nature is simple and uniform, it reconstructs the unobservable natural systems by analogy, abduction and inductive extrapolation from observable ones; (iv) nature actually is simple and uniform (Doppelt 2007, 102). So, (R3) explains and justifies (R2), (R2) explains and justifies (R1) and (R1'), thus answering (Q5) and (Q5'); finally, (R1) and (R1') answer respectively (Q1) and (Q4), thus explaining success. Notice, (R3) is not a question-begging assumption, but a postulate required to explain the unquestionable phenomenon of novel success. In other words, (R3) is not a premise of my (inference to the best explanation) argument for realism, but part of its conclusion.

6. Why antirealists lack a similar strategy We shall now see that nothing analogous to the realist explanatory argument (R1)-(R2)-(R3) is available to antirealists. In fact, we saw that the only contribution offered by antirealists to an explanation of T’s success were statements on the scope and truth-value of T’s empirical consequences: (I), (II), (IV) say that T predicts np because its empirical consequences are true (a non sequitur); (III), (V), (VI), and (VII) say that T predicts np because it predicts a number of phenomena including np, or because it is instrumentally reliable (a repetition which fails to answer (Q1)); (I)-(VI) also say or imply that “np” is true because all (or most) empirical consequences of T, including “np”, are true (a trivial answer that leaves (Q4) unanswered); (VIII), the deflationist explanation, fails to dispense with the appeal to truth, and again to answer (Q1) and (Q4). So, how should antirealists go about explaining (Q1) and (Q4)? They reject any appeal to truth, and so the realist key assumptions (R1) the theorist actually found a theory with enough true, deep and heuristically fecund theoretical content to have the novel “np” as a consequence, and (R1') the theorist actually found a theory with enough true, deep and heuristically fecund theoretical content to have mostly true empirical consequences, including some novel ones like “np”.

60

Chapter Four

In their place, they can only use (A1) the theorist actually found a theory with some novel empirical consequences, including “np”, and (A1') the theorist actually found a theory with mostly true empirical consequences, including some novel ones like “np”. .

So, while (R1) ad (R1') raised the questions (Q5) how has the theorist picked a theory with enough true and heuristically fecund theoretical content to predict the novel phenomenon np, rather than one of the infinite theories compatible with the old data but too poor of true content to predict np? and (Q5') how has the theorist picked a theory with enough true and heuristically fecund theoretical content to make mostly true predictions, including novel ones like “np”, rather than one of the infinite theories compatible with the old data but for the rest making mostly false predictions?, (A1) and (A1') raise the parallel questions (Q6) how has the theorist picked just a theory predicting the novel np, rather than one of the infinite theories compatible with the old data but not predicting np? and (Q6') how has the theorist picked a theory with mostly true predictions, including novel ones like “np”, rather than one of the infinite theories compatible with the old data but for the rest making mostly false predictions? Realists answer (Q5) and Q5') by (R2) The theorist (i) aimed at a true—but also deep and heuristically fecund—theory, (ii) skilfully employing a reliable method (i.e., truth-conducive and heuristically effective) (iii) and relying on mostly true background beliefs. But antirealists cannot answer (Q6) and Q6') by the parallel assumption (A2) The theorist

Why Antirealists can’t Explain Success

61

(i) aimed at a theory predicting all previously known data, plus many novel ones, (ii) skilfully employing a reliable method, (iii) and relying on mostly true background beliefs. They cannot answer this because, while finding a theory which accounts for known data is fairly easy, there are only two methods for seeking a theory which makes true predictions on unknown data: one is looking for a true, deep and fruitful theory, as the realist suggests. The other is inductively extrapolating the unknown data from the known ones. In fact, those who try to explain science’s success without realism propose exactly this latter method: recognizing regularities in the observed phenomena by the inductive methods of agreement and difference or of controlled and double-blind experiment (Laudan 1984, § 5), and trusting that, thanks to the stability and regularity of nature (Rees 2012, 302), past regularities will to go on in the future (Wright 2013, caps. 3, 4: see Alai 2014a, 125-126). But this cannot explain the prediction of novel phenomena, for by hypothesis they are heterogeneous to the old ones, hence cannot possibly be extrapolated from them. Although nature is regular, it is not prima facie empirically uniform, for often we encounter phenomena utterly unexpected and unlike earlier ones. That often science manages to predict them, cannot be explained by empirical regularity and inductive extrapolation. For Rees (2012, 294) we find theories making novel predictions because, as suggested by Kuhn (1970, 169), new theories are chosen not only for their capacity to account for old phenomena, but for their potentiality to capture new phenomena. This is right, but this potential is provided only by their theoretical assumptions, and it can be assessed only by looking at those assumptions and taking them seriously. So, it would be a miracle if this method actually singled out predictively fecund theories, but their theoretical assumptions were largely false. Thus, antirealists cannot point at any method for finding theories which make novel predictions, and must say that such theories are found by sheer luck. But this is utterly implausible, since by hypothesis novel predictions also a priori improbable; so, finding theories with novel success would be a cosmically rare coincidence, while such cases are relatively frequent in science.10 This explains why, contrary to Fine’s claim (here, § 5), even if instrumental reliability already entails success, truth is not superfluous: because when novel predictions are concerned, one can find empirical adequacy or instrumental reliability only by searching (and finding) truth.

62

Chapter Four

Faced with their inability to explain predictive success, eventually antirealists can only reject the request for an explanation as illegitimate (van Fraassen 1980, 2.4, Rees 2012, 304.). But this is a clearly ad hoc move, for they standardly accept explanations and inferences to the best explanation in the observable realm. Stanford (2000, 281) holds that once realists abandon the untenable claim that theories are completely true, they cannot anymore explain success, since the probability to yield true novel predictions is high for completely true theories, but may be very low for partially false theories. But the realist need not assume that such a probability is high, only that it is enough to explain why theories succeed in making novel predictions from time to time. On the contrary, the probability that a radically false theory, assembled just to accommodate known data, predicts novel phenomena is negligible (Alai 2012, 88). Stanford suggests that this is not the case, because thanks to “the systematic relationship among phenomena within the same domain of inquiry (e.g., optical phenomena, chemical properties, descent with modifications in organisms”, even a radically false theory which saves the phenomena is likely to make novel predictions in that domain (ibid.). So, looking for empirical adequacy (i.e., in practice, inductive extrapolation) would be a method for finding theories with novel predictions. But either “systematic relationship” means homogeneity, hence the predictions can be derived from past observations by inductive extrapolation, but then they are not actually novel; or the new phenomena are actually heterogeneous, and “systematic relationship” means the underlying theoretical mechanisms linking the old with the new phenomena: and if so it would be a miracle if a theory predicted the novel phenomena without getting those mechanisms right, i.e., without being true. For example, the false corpuscular theory of light saved a lot of optical phenomena, but it could not predict novel phenomena within the same domain; Fresnel’s wave theory, instead, predicted the new white spot effect, because it got right the wave mechanism from which it depends.

7. The evolutionary explanation The failure to explain how theories making novel predictions can be found ((Q1) and (Q4)) undermines also another famous antirealist account of success: (IX) “Natural” selection (van Fraassen, 1980, 40): i.e., our theories are successful, simply because the unsuccessful ones were dropped.11

Why Antirealists can’t Explain Success

63

In fact, this explains only (1) why we have only successful theories (what in evolutionary theory would be a phenotypic explanation); not (2) thanks to which characteristics these theories are successful (the genotypic explanation), and even less (3) why and how we found theories with those characteristics.12 Just in the same way, to the question “why (most) birds can fly?” it is not sufficient to answer (1) “because (most of) the nonflying birds were wiped out by natural selection”.13 Darwinian theory also answers question (2), explaining that they fly thanks to the genes endowing them with wings, feathers, hollow bones, etc., and question (3), explaining that those genes were produced by chance mutation and selected by the harsh struggle for survival and reproduction (Kitcher 1993). Realism too answers all of the three questions: (1) we have only successful theories because the unsuccessful ones were dropped; (2) our theories are successful because they are true, deep and heuristically fecund; (3) they have these properties because theorists employ a reliable method and have largely true background beliefs. Antirealists of course answer (1) by van Fraassen’s explanation, and can answer (2) by claiming that theories are successful because they entail all the known data, plus many novel ones; but this is a trivial explanation, and then they stop here, without answering (3),14 i.e., our questions (Q1) and (Q4). I thank those who personally or in public discussions helped me with useful comments and suggestions: Claudio Calosi, Vincenzo Fano, Giovanni Macchia; Pieranna Garavaso, Carl Hoefer, Matteo Morganti, Marco Panza, Alberto Cordero; David Miller, Ilkka Niiniluoto.

References Alai, M. 2012, “Levin and Ghins on the ‘No Miracle’ Argument and Naturalism”, European Journal for Philosophy of Science 2 (1), 85 —. 2014a, “Explaining the Novel Success of Science”, Metascience, 23 (1), 125, http://www.springerlink.com/openurl.asp?genre=article&id=doi:10.10 07/s11016-013-9838-7 —. 2014b, “Novel Predictions and the No Miracle Argument”, Erkenntnis, 79 (2), 297, http://link.springer.com/article/10.1007%2Fs10670-013-9495-7 —. forthcoming, “Defending Deployment Realism against Alleged Counterexamples”, in Bonino, G., J. Cumpa and G. Jesson (eds.)

64

Chapter Four

Defending Realism. Ontological and Epistemological Investigations, Frankfurt a.M.: Ontos Verlag Barnes, E.C. 2008, The Paradox of Predictivism, Cambridge: Cambridge University Press Clarke, S. and T.D. Lyons (eds.), 2002, Recent Themes in the Philosophy of Science. Scientific Realism and Commonsense, Dordrecht: Kluwer Doppelt, G. 2007, “Reconstructing Scientific Realism to Rebut the Pessimistic Meta-induction”, Philosophy of Science 74, 96 Fine, A. 1984, “The Natural Ontological Attitude” in Leplin J. (ed.) 1984, 83 —. 1986, “Unnatural Attitudes: Realist and Instrumentalist Attachments to Science”, Mind 95, 149 —. 1991, “Piecemeal Realism”, Philosophical Studies 61, 79 Hacking, I. 1983, Representing and Intervening, Cambridge: Cambridge University Press Kitcher, P. 1993, The Advancement of Science, Oxford: Oxford University Press Kuhn, T. 1970, The Structure of Scientific Revolutions, Chicago: University of Chicago Press Ladyman, J. 2002, Understanding Philosophy of Science, London, New York: Routledge Laudan, L. 1984, “Explaining the Success of Science” in Cushing J.T., C.F. Delaney and G. Gutting (eds.), Science and Reality: Recent Work in the Philosophy of Science, University of Notre Dame Press: South Bend, 83 Leplin, J. (ed.) 1984, Scientific Realism, Berkeley: University of California Press Leplin, J. 1987, “Surrealism”, Mind 96, 519 —. 1997, A Novel Defence of Scientific Realism, Oxford: Oxford University Press Lipton, P. 1991, Inference to the Best Explanation, London: Routledge (20042) Lyons, T.D. 2002, “The Pessimistic Meta-Modus Tollens” in Clarke, S. and T.D. Lyons (eds.) 2002, 63 Maher, P. 1988, “Prediction, Accommodation and the Logic of Discovery”, PSA 1, 273 Miller, D. 1974a, “Popper’s Qualitative Theory of Verisimilitude”, The British Journal for the Philosophy of Science 25, 166 —. 1974b, “On the Comparison of False Theories by their Bases”, The British Journal for the Philosophy of Science 25, 178

Why Antirealists can’t Explain Success

65

Musgrave, A. 1988, “The Ultimate Argument for Scientific Realism”, in R. Nola (ed.), Realism and Relativism in Science, Dordrecht: Kluwer, 229 —. 2006-2007, “The ‘Miracle Argument’ for Scientific Realism”, The Rutherford Journal 2, http://www.rutherfordjournal.org/article020108.html Niiniluoto, I. 1987, Truthlikeness, Dordrecht: Reidel —. 1999, Critical Scientific Realism. Oxford: Oxford University Press Psillos, S. 1999, Scientific Realism. How Science Tracks Truth, London: Routledge Putnam, H. 1975, Mathematics, Matter and Method, Philosophical Papers vol. 1, Cambridge: Cambridge University Press —. 1978, Meaning and the Moral Sciences, London: Routledge & Kegan Paul Rees, P. 2012, A Critique of The Arguments for Scientific Realism, Reading: Cranmore Publications Schurz, G. and P. Weingartner 2010, “Zwart and Franssen’s Impossibility Theorem Holds for Possible-World-Accounts but not for ConsequenceAccounts to Verisimilitude”, Synthese 172, 415 Smart, J.J.C. 1968, Between Science and Philosophy, New York: Random House Stanford, K.J. 2000, “An Antirealist Explanation of the Success of Science”, Philosophy of Science 67, 266 Tichý, P. 1974, “On Popper’s Definitions of Verisimilitude”, The British Journal for the Philosophy of Science 25, 155 Van Fraassen, B.C. 1980, The Scientific Image, Oxford: Clarendon Press White, R. 2003, “The Epistemic Advantage of Prediction over Accommodation”, Mind 112 (448), 653 Worrall, J. 1985, “Scientific Discovery and Theory-Confirmation”, in Pitt J. (ed.) Change and Progress in Modern Science, Dordrecht: Reidel —. 2005, “Prediction and the ‘Periodic Law’: a Rejoinder to Barnes”, Studies in History and Philosophy of Science 36, 817 Wright, J. 2002, “Some Surprising Phenomena and Some Unsatisfactory Explanations of Them”, in Clarke, S. and T.D. Lyons (eds.) 2002, 139 —. 2013, Explaining Science’s Success. Understanding How Scientific Knowledge Works, Durham: Acumen Zwart, S.D. and M. Franssen 2007, “An impossibility Theorem for Verisimilitude”, Synthese 158, 75

66

Chapter Four

Notes 1

See Alai 2014b, § 3.1. Worrall 1985, 319; 2005, 819. Leplin 1997, chaps. 2, 3. But this detail will be irrelevant here. 2 More precisely, at least those tenets of the theory which were essentially employed in deriving such predictions must be (approximately) true: see Psillos 1999, 108 ff. 3 White 2003. More implicitly, Laudan 1984, 92. 4 In Musgrave’s words, “It is like explaining why some crows are black by saying that they all are” (1988, 242). Stanford’s explanation “is just a tautology” (Rees 2012, 310). See also Leplin 1987, 522; 1997, chap. 1; Stanford (2000, 268) raises this criticisms against the empirical adequacy explanation, without noticing that it also applies to the predictive similarity explanation. See Musgrave 2006-2007 for a detailed confutation of Stanford along these lines. 5 As granted by Stanford himself: 2000, 276. 6 Providing a general definition of what is relevant or irrelevant in this intuitive sense is not easy: it can be done in terms of consequences (as in Schurz and Weingartner’s 2010, 426), or of models (Zwart and Franssen 2007) or of state descriptions or constituents (Niiniluoto 1987), but each of these approaches encounter technical and conceptual problems. Still the basic idea is clear enough to show what is wrong with Tichy’s and Miller’s paradoxes, and to make sense of the evaluation and comparison of the quantity of true consequences of theories, at least in many typical cases. 7 As noticed at the end of § 3. 8 In Leplin’s words, “Why does this theory work, while others equally the product of diligence and preferred methods fail?” (1997, 8); see Rees 2012, 301. 9 See also Stanford 2000, 282. 10 An analogous conclusion is reached by Barnes 2008, § 4.3.2.2. 11 In particular, thanks to the filter provided by the scientific community, peerreview processes, etc.: Rees 2012, 291-292. 12 See Laudan 1984, 92; Psillos 2009, 96; Stanford 2000, 270. 13 For similar examples see Leplin 1997, 9; Kitcher 1993; Stanford 2000, 269-272. 14 See also Lipton 1991, 170 ff., 194-195; Ladyman 2002, 217; Niiniluoto 1999, 197-198; Psillos 1999, 96-97, etc.

CHAPTER FIVE THE CONTENT CHALLENGE TO NOMINALISM ABOUT MATHEMATICS MATTEO PLEBANI

1. Hard vs. easy roads to nominalism Nominalists think that there are no mathematical objects. According to received wisdom, our best scientific theories contain claims (such as “the atomic weight of beryllium is 9.012182” or “there is a function that maps spacetime points into real numbers”) that imply the existence of mathematical objects. It seems, therefore, that a nominalist cannot accept our best scientific theories in their entirety. The problem for her then is to specify the part of these theories she retains (the so-called “nominalistic content” in the theory). The difference between hard- and easy-road nominalists (Colyvan 2010) lies in their different stances to this problem. According to hard-road nominalists, the only acceptable way of conveying the nominalistic content of our mathematical theories is by taking a detour through a theory spelled out in nominalistic language, i.e. in a language where quantification over and reference to mathematical objects are absent. This strategy is championed in Harry Field’s work (1980, 1989). Easy-road nominalists (ERN), in contrast, think that this is not necessary. In particular, Yablo’s logical subtraction-based strategy (Yablo 2012) and Melia’s weaseling strategy (Melia 2000) are two ways to implement the idea that one can spell out the nominalist content more directly by just withdrawing the unwanted implications.1 Introducing the practice of weaseling, Melia says the following: In general, we assert sentences in order to present a picture of the way we think the world is. We normally think of each successive sentence in our story as adding a further layer of detail […] But […] Why can’t we understand some of the later sentences [of a theory] as taking back things that we said earlier in the theory? Why can’t we understand the later

68

Chapter Five sentences as erasing some of the implications of the earlier part of the story? Or as changing some of the implications of the previous parts of the theory? (Melia 2000, 467).

Considering the case in point, to spell out the nominalistic content of a scientific theory T by weaseling amounts to saying: “T, but there are no mathematical objects” (I will label such statements as “weaseling statements”). Melia (2000) notes that in many cases, it is unproblematic to interpret weaseling statements in a non-contradictory manner. If somebody claims that all Fs are Gs except Harry, who is F but not G, it charitable to interpret this person as claiming that for all x such that x is different from Harry, if x is F, then x is G. That much is unproblematic. Problems arise when the only way of expressing the nominalistic content of a given scientific theory T is via a weaseling statement.

2. The content challenge One objection against easy road nominalists is that in the absence of a paraphrase that explains its content, a weaseling statement cannot but be interpreted as a contradiction. Colyvan formulates the challenge in as follows: I simply do not know what to make of sentences such as (4) [There exists a differentiable function that maps from the spacetime manifold to the real numbers, but there are no mathematical objects] where no obvious paraphrase presents itself. […] How does [Melia] respond to someone who simply does not understand sentences like (4), except as the obvious contradiction? One way of clarifying things would be to provide the appropriate translation. Indeed, this is the only way of replying that is acceptable. […] So we see that this easy-road strategy is dependent on the success of a hard-road strategy (Colyvan 2010, 295-296).

A similar problem is pressed by Azzouni (2009, 156-159). The previous quote by Colyvan suggests to seeing the challenge as in the form of a dilemma: (D) either (1) ERN provides an alternative to weaseling for conveying the nominalistic content, or (2) the nominalist content is rendered unintelligible or contradictory. Colyvan thinks that both horns of the dilemma are unattractive: accepting (1) is unattractive for ERN because it makes easy-road strategies parasitic on the hard-road ones, and accepting (2) means refusing to meet

The Content Challenge to Nominalism about Mathematics

69

they very reasonable demand of a clarification. Raley (2012) is completely dedicated to the discussion of this problem. Raley argues that Melia’s nominalist account of mathematics is “incoherent” and that the incoherence of Melia’s position can be appreciated by noticing that Melia ought to respond to the content challenge. According to Raley, Melia could try to respond to this challenge in three ways. Yet she claims: the problem remains that given all of Melia’s claims, none of these additional options leaves the original strategy intact. Any workable interpretation of Melia’s position forces him to take back at least some of his own assertions (Raley 2012, 342).

In what follows, I will argue that there are ways to resist this conclusion. First, Raley has underestimated the resources that Melia is allowed to use for replying to the content challenge. The three options considered by her are not at all exhaustive. Second, at least two of Raley’s ways out are perfectly compatible with Melia’s general strategy. Raley’s way of conceiving the dialectic between Melia and sceptics of weaseling is highly influenced by the way she interprets the content challenge. I use her paper as a case study to elucidate the general significance of that challenge.

3. Raley’s menu According to Raley (2012), Melia has only three options available to reply to the content challenge. They are as follows: we can say that, contrary to appearances, the statements taken back are not implications of the theory. This is because we can understand “taking back” as changing our logic (in particular, as having changed the implication relations between theories and sentences) (Raley 2012, 343).

Or: We take back and deny the truth of the theory within which the implications were couched (Raley 2012, 343).

Or: We separate out and then deny only the existential statements (or the offending subset of these existential statements) of our Best Theory (Raley 2012, 344).

Chapter Five

70

Raley’s method of conceiving the dialectic between the content challengers and Melia is contingent on the moral she draws from the content challenge: If all Melia is saying is that in some cases, we don’t mean what we said and can illustrate this by providing an alternate expression, then Melia is not offering a methodology that is different from paraphrase. But we already know that he does not think the paraphrase move is promising […] If, on the other hand, Melia claims that we didn’t mean what we said but an alternate expression cannot be provided, then why shouldn’t we just conclude that the speaker is contradicting himself? (Raley 2012, 342).

4. Partial and complete paraphrases A first, partly terminological issue is clarifying what Raley means when she uses the phrase “alternate expression”. The last quote suggests that “alternate expression” is synonymous of “paraphrase”. However, there is an important distinction to be drawn between partial and complete paraphrases. “Paraphrase” can mean i) an “alternate expression” for weaseling statements that helps clarify their content or ii) a complete paraphrase that maps every statement of T into a statement expressed in nominalistically kosher language. This distinction is relevant when evaluating Melia’s options for responding to the content challenge: Melia is committed to the impossibility of providing a paraphrase in the sense of (ii) on the pain of falling back into the hard road, but this does not prevent him from offering “alternate expressions” in the sense of (i). Liggins (2012) suggests that a partial paraphrase of our best scientific theories could be adequate to meet the content challenge. If this were the case, easy-road nominalists could meet the content challenge while maintaining that the only way to express the entire nominalistic content of a theory within a finite compass is via weaseling. Why should we need a complete paraphrase of our scientific theories to meet the content challenge? The challenge for the nominalist is to render non-obscure the contents of the parts of scientific theories that cannot be paraphrased nominalistically. Consider the following claim: (S) The number of cats is even. One way of conveying the nominalistic content of (S) would be using a sentence such as the one that follows: (S*) Either there are two cats or there are four cats or there are six cats…2

The Content Challenge to Nominalism about Mathematics

71

It is true that this is not a complete sentence, at least in the sense that it is not equivalent to any finite disjunction of sentences (maybe it would be complete if the nominalist’s vocabulary includes expressions such as dots “...”). It is in this sense that the nominalistic content of (S) cannot be paraphrased completely. So what? So long as the content of S* is nonobscure, the nominalist has succeeded in meeting the content challenge relative to sentence S. Note that anyone who understands (S*) is in a position to explain why the claim that Santa Claus exists is not part of the content of S*. As long as such a move is not ruled out in other similar cases, there is reason to think that nominalists can reply to the content challenge without abandoning weaseling-like strategies. This relates to Raley’s second interpretation of Melia, according to which he would reply to the content challenge by denying the truth of our original scientific theory T. It is somewhat strange to put forward this move as a “way out” of the content challenge. Given that Melia’s account of mathematics is nominalistic and the theory T is platonistic, Melia is already committed to denying the truth of T, regardless of the content challenge. If T were completely true, there would be numbers, and, according to nominalists such as Melia, there aren’t. If Melia could not deny the truth of T without making his position incoherent, he would be in trouble. According to Raley, the problem with denying the truth of T is as follows: that’s not an option because […] In the absence of paraphrase, […] we have no way of separating the true from the false statements in a theory (Raley 2012, 343).

Of course, one reply to this is that all statements entailing the existence of numbers are not completely true, but each statement S has a true part, which is S minus the claim that there are abstract mathematical objects. The sum of these parts is the nominalistic content of the theory, the part that even a nominalist retains. Of course, one might raise the problem of elucidating the nominalistic content of our scientific theories. Melia claims this cannot be done by offering a nominalistic replacement of our theory T. However, this does not mean he cannot meet the demand in another way, i.e. by elucidating the nominalistic content using the sort of partial paraphrases presented here. One could still complain that a partial paraphrase is much more similar to a complete paraphrase than it is to weaseling. Considering Azzouni’s proxy strategy (2009, 152-9), which is somewhat close to partial paraphrase, Raley (2012, fn. 2) remarks that this strategy is very different from weaseling. Using proxying, one does not vindicate the weaseling

72

Chapter Five

strategy but simply adopts a different strategy. This is true, but misleading as far as it gives the impression that weaseling and partial paraphrase are in competition, implying that one approach excludes the other. Why should the fact that nominalistic content can be expressed by a partial paraphrase entail that weaseling is not a legitimate move? Rather, the initial supposition should be the opposite. As a matter of historical record, people such as Yablo (2012, MS, see also the introduction of 2010) have defended weaseling-like strategies and denied that nominalistic content can always be expressed in nominalistic language, while trying to approximate nominalistic sentences expressing nominalistic content (see Yablo 2002, 2005). The same holds for Melia (2000) and Liggins (2012), whose defence of the weaseling strategy is coupled with partial paraphrases of platonistic claims. Probably, the thought here is that in the presence of a proxy strategy such as that proposed by Azzouni (2009), in which not only some but every platonistic claim D stands proxy for a nominalistic claim D*, weaseling would become redundant. This conclusion, too, can be resisted. By focusing on a trivial example such as (S), one can find reasons to combine the weaseling strategy with partial paraphrase: the weaseling statement is expressed in the finite compass, whereas the proxy statement associated with S is either an infinite locution or a locution that expresses only a part of the nominalistic content.3 Moreover, if we consider not just one statement but a theory in its entirety, it is clear that asserting “T, but there are no mathematical objects” is quite a shorthand for replacing T with its nominalistic counterpart T* (the set of statements D*). In short, it is unclear as to why an ERN could not accept the first horn of (D) and question the claim that the only other way to express the nominalistic content apart from weaseling is to provide a complete paraphrase of our best scientific theories. Melia is only committed to the impossibility of providing a complete paraphrase of our scientific theories, so providing “alternative expressions” to the weaseling statements is a live option for him. It is important to note here that the program of providing “alternate expressions” need not be similar to Field’s strategy (Colyvan 2010, 287), so that the tenability of easy-road strategies need not necessarily presuppose the success of hard-road strategies. The project of providing partial paraphrases is much less demanding than Field’s hard road. We have just seen that ERNs are allowed to indulge in infinite disjunctions, whereas Field is not. Moreover, Liggins (2012, 1002-3) offers examples of cases where Melia is entitled to use resources richer than Field’s in dealing with quantitative statements. Modal language is another resource that Field’s strategy precludes but ERNs are entitled to

The Content Challenge to Nominalism about Mathematics

73

use: see Dorr (2010) for a defence of the proposal of replacing scientific theory T with MŠT (where M is the mathematical part of T). The concern about the weaseling strategy voiced in Colyvan’s and Azzouni’s papers remains. The charge this time is not that weaseling is impossible or redundant but rather that the way Melia presents it, i.e. as admissible even in the absence of something like the proxy strategy, is irresponsible (Azzouni 2009, 159). It is clear that sometimes we can retract implications of what we earlier said. Apart from Melia’s own examples, consider cases where somebody asserts a conjunction for withdrawing one of the conjuncts later: one asserts that Grass is green and snow is black MINUS snow is black. However, in other cases, this subtraction operation seems not to work. For example, try to subtract the fact that somebody is a philosopher from the fact that Plato is a philosopher, that a given tomato is red from the fact it is scarlet, or that my arm went up from the fact that I raised my arm (Yablo 2012). It is evident that there are cases where weaseling does not yield any evaluable content. Azzouni (2009, 158-159) presents a hilarious example taken from a passage of Calvino’s Cosmicomics and rightly asks how to distinguish Melia’s weaseling from Calvino’s weaseling. Colyvan (2010) rightly asks why trying to subtract the assumption that there are numbers from the content of mathematics would be different from trying to subtract the assumption that there are Hobbits from the content of The Lord of Rings. These are perfectly legitimate demands, but they do not show in any way that Melia’s weaseling is irresponsible. Rather, they call for a theory of logical subtraction, the practice of scaling down part of the content of one’s statement. This theory will yield an answer to the question about the distinction between responsible and irresponsible weaseling. However, that is very different from assuming from the outset that weaseling as applied to mathematical claims falls on the irresponsible side. According to one theory of logical subtraction (Yablo’s theory), the truth is rather the opposite. It is now time to consider this theory.

5. Logical subtraction We can approach the issue of logical subtraction by noticing that another way to call into question (D) is to attack (2). First, there is an ad hominem argument against Colyvan. At a certain point, Colyvan concedes that there are cases in which the only way of expressing certain content is the use of metaphorical language: “non-literal language is present in our scientific discourse—average stars, and so on—and such usage is arguably ineliminable” (2010, 298).

74

Chapter Five

This concession seems highly plausible, but in such cases, something very close to weaseling would be unavoidable. As Colyvan admits, we would like to be able to retract the commitments engendered by the use of figurative language. We would like to say that someone is an angel even though there are no such creatures as angels. However, this seems to give rise to a new version of the content challenge: how can we specify the content of metaphorical statements that we cannot endorse literally but cannot paraphrase either? Colyvan is aware of the problem. His rejoinder (Colyvan 2010, 299 ff.) consists of allowing for essentially metaphorical language as far as descriptive use of the language is considered, but demanding a paraphrase when language is used to formulate explanations. I do not want to discuss here the tenability of Colyvan’s proposal, but Colyvan’s qualification (that the content challenge is a problem only when we are dealing with mathematical explanations) is telling. A more general reply to (2) would be that weaseling is an application of a general procedure that allows for the scaling down of certain content from a larger hypothesis. Yablo (2006, 2012, MS) has drawn from linguists’ works on the notion of logical subtraction. His idea is to make the nominalistic content the result of scaling down from a scientific statement the presupposition that there are mathematical entities. As long as logical subtraction is legitimate, there is no reason for nominalists to not use it. As noticed in the previous section, it is clear that in some cases, logical subtraction is legitimate. If you say that snow is white and hot, I can say that apart from the part about the temperature of the snow, your statement is correct. If your statement is P (= snow is white) & Q (= snow is hot), I want to endorse is P&Q minus Q. In other cases, though, logical subtraction seems to give awkward results, for example, if we try to subtract the claim that somebody is a philosopher from the claim that Plato is a philosopher. It is perfectly natural to ask for a criterion to distinguish between cases where logical subtraction is possible from cases where it cannot be performed sensibly. Therefore, it is clear that logical subtraction makes sense in some cases but seems insensible in other cases. What remains unclear is whether there is a criterion to set the good cases apart from the bad cases. In this context, it is natural to interpret the content objection expressed by Azzouni (2009) as a challenge to find an alternative criterion to this: (C1) A-B has definite content only when this content can be expressed with non-problematic sentences—what Azzouni calls the “targets” of

The Content Challenge to Nominalism about Mathematics

75

the proxy statements—and these unproblematic sentences meet certain constraints.4 For instance, in discussing the proxy strategy, Azzouni sets forth these requirements for the targets of proxy statements: (The proxy norm) (i) The representation and deduction roles of D can be interpreted as ones in which D is proxying for D*, (ii) Where disputes arise about whether discourse containing D has such and such implications, or applies in such and such circumstances, researchers can determine answers via access to D* (Azzouni 2009, 152). […] whether condition (ii) of the proxy norm is satisfiable […] depends on how informative the existence proof (of D* on the basis of D) is. If the information the proof yields suffices—in principle—to adjudicate disputes about the (implicational) properties of D*, then a strict demand that D* be exhibitable is unjustified. D can still be treated as proxying for the— strictly inexpressible—D*. Call this requirement, on the relationship of proxies to their targets, the “informativeness condition on proxy/target relations” (Azzouni 2009, 156).

A reply to this way of expressing the content challenge involves replacing (C1) with another criterion. Yablo (2012, MS) developed a theory of logical subtraction to account for the difference between cases where it is possible to subtract B from A and cases where it is not. If correct, this would explain why it is possible to subtract from our scientific theories the hypothesis that there are mathematical objects, even though it is not possible to subtract from Tolkien's The Lord of the Rings the assumption that there are Hobbits (see Colyvan 2010). It is worth spending some time looking (rather superficially) at Yablo’s theory. The guiding idea (Yablo 2012, 1015-1019) is that A-B is evaluable in a world w' where B is not true depending on the status at w' of this conditional: BŠA. The question to be asked is what are the reasons for the conditional being true at w'.5 The next step is to compare them with the reasons why A-B is true in a world w where A is true. If the reasons for the conditional being true in w and w' can be the same, then there is an evaluable logical remainder of A-B, otherwise there is not. Testing this account against a case where A is of the form p&q and B is p, we get a positive result. pŠ p&q is true in a world where p&q is true because q is true in that world,

Chapter Five

76

and the same reason can hold in a world where p is false. In contrast, if we have A = this tomato is scarlet and B = this tomato is red, there is no evaluable remainder: in a world where the tomato is scarlet, the tomato is scarlet-if-red because it is scarlet and the same fact cannot hold in a world where the tomato is not scarlet. Now, to understand whether T-M is evaluable, the proposed test asks us to focus our attention on what makes it true, in the actual world, i.e. the following conditional: MŠT. If the reason why such a conditional is true involves only nominalistic facts, then the conditional is evaluable in worlds where there are no mathematical objects as well.6 If this were the case, logical subtraction of the hypothesis that there are mathematical objects from the content of our scientific theories would yield a definite result, thus vindicating the claim that Melia’s weaseling is responsible. There is certainly one kind of platonists, the so-called heavy duty platonists (Field 1989, 186-93), who think that facts about the physical world do not suffice to make the conditional “if M then T” true. They maintain the following: the truth of T at w depends on the obtaining there of some fundamental relations between concrete and mathematical entities—relations which do not supervene on the totality of facts either about the concrete realm or about the narrowly mathematical relations with which M is concerned (Dorr 2010, 148).

Yablo’s defence of the permissibility of Melia’s weaseling will have no bite against a heavy duty platonist, as Yablo acknowledges (2012, fn.16). However, I consider it as a piece of common wisdom that heavyduty platonism is not a popular option, so the real question is whether any other kind of platonist is in a position to deny the existence of nominalistic truthmakers for the conditional MŠT. If we admit that in the presence of mathematical objects, some features of the physical world suffice to make T true, how can we deny the intelligibility of the notion of nominalistic truthmakers for MŠT? Of course, one could think that this does not address the point Azzouni highlighting in advocating an informativeness constraint on the existence of proxies.7 However, the question is whether the request remains reasonable in the present context. The proxy strategy with its constraints was submitted as a criterion to distinguish between bad and good cases of logical subtraction. We have now seen that an alternative criterion is available, which gives the result that Melia’s weaseling is responsible for

The Content Challenge to Nominalism about Mathematics

77

mathematical statements even in the absence of targets, let alone targets meeting Azzouni’s constraints. This does not mean that such targets cannot be found. For instance, Dorr (2010) advocates the translation of a scientific theory T into T , where8 (T )

MŠT.

This seems to be in line with the use of the logical subtraction just sketched and seems to provide a way to satisfy the proxy norm and the informativeness constraint. We can put the point dialectically in the manner discussed by Liggins (forthcoming). Recent work on grounding (Clark and Liggins 2012) starts from the consideration that some facts hold because of or in virtue of the obtaining of other facts: it is true that dogs bark or fly because it is true that dogs bark, {Socrates} exists because Socrates exists, etc. Confronted with a mathematical claim S implying that an abstract object o and a concrete object c are so-and-so related, a nominalist is entitled to ask platonists about the grounds for the obtaining this relation. The most reasonable option is to think that it is some nominalistic property of the relevant concrete object that explains why it stands in a relation with a certain abstract object. Therefore, the platonist herself should acknowledge that platonistic facts are grounded in nominalistic facts. If this line of thought is correct, then even platonists (or Azzouni) are committed to the existence of such nominalistic facts.9 Furthermore, once the existence of such facts is established as a common ground, I see no reason why a nominalist cannot take these facts as the content that is asserted when someone utters a mathematics-infused sentence S.10 Granted Yablo's account is controversial, but it points towards the fact that there is no reason to think that we should always be able to express nominalistic content in nominalistic language. As Yablo (2012, 1014) aptly notes, the task of expressing nominalistic content is that of representing a world in which mathematical objects are absent, not that of representing a world without using mathematical vocabulary. The fact that we need mathematical language to represent a nominalistic word is no greater a problem for those who believe in a number-free world than the fact that we need natural language to represent the world at the time of Mesozoic is problematic for those who believe in a human-free world. The idea that, in the absence of a paraphrase, the content of a weaseling statement is doomed to obscurity should therefore be regarded with suspicion.

78

Chapter Five

6. A note on implication Discussing possible ways out for Melia, we saw that Raley considers the following option: we can say that, contrary to appearances, the statements taken back are not implications of the theory. This is because we can understand “taking back” as changing our logic (in particular, as having changed the implication relations between theories and sentences). On this interpretation, Melia has dropped classical logic (Raley 2012, 343).

Interestingly, ERNs can make sense of a notion of implication according to which scientific theories do not entail mathematical objects in any interesting sense. Moreover, this does not require abandoning classical logic. The idea is to distinguish between entailment as truth-value preservation and entailment as truth-value and subject matter preservation (see Yablo 2009, MS). This distinction can be used to explain why we feel uneasy about the following argument: (P) I have two hands, therefore (C) I am not a brain in a vat. If we distinguish between entailment as truth-value preservation and entailment as truth-value and subject matter preservation, we can say that the argument presented above is correct in the sense that it preserves truth from the premise to the conclusion. Still, we feel uneasy about it owing to a shift in the subject matter from the premise, which does not seem to be about the exclusion of sceptical scenarios, to the conclusion, which is precisely about that issue. The conclusion is not a consequence of the premise, if implication is understood as including subject matter preservation. Applying this to the ontology of mathematics, the inference from (A) The number of stars is greater than 1000, to (B) There are numbers causes a shift in subject matter because (A) only presupposes the existence of numbers to assert something about the concrete world, whereas (B) takes the existence of mathematical objects as its assertive content.

The Content Challenge to Nominalism about Mathematics

79

A strategy for defining subject matters is present in the work of David Lewis and is developed further in Yablo (2012, MS). We do not need to go into the details here. The basic idea is that subject matters are treated as equivalence relations on the set of worlds (as a partition of that set). Given a subject matter m, two worlds are in the same m-cell if they agree on m. If we take as our subject matter the concrete world, w and w' will be in the same cell if and only if they are concretely indiscernible. A certain proposition about a certain subject matter m is true in a world w if it is true in a world w' that belongs to the same m-cell as w.11 A certain proposition is wholly about a subject matter m if and only if its truth value does not change between m-equivalent worlds. Whether a proposition S wholly about a subject matter m is true in w can be settled by simply looking at w's m-conditions. Accordingly, the nominalistic content of a theory T is the set of implications of T that are wholly about the concrete world (Yablo 2012, 1012-1014). Assuming that the conditions of the physical world do not suffice to settle the question as to whether there are mathematical objects, the claim that there are mathematical objects is not part of the theory’s nominalistic content. In this sense, it is not an implication of the theory. There could still be a worry here that we have “changed the implication relations between theories and sentences” and so “dropped classical logic” (Raley 2012, 343). However, this need not be the case. It is easy to prove that if a sentence A is true about a subject matter m and that A implies (in the classical sense) B, B is true about m as well. This means that if the subject matter under discussion is fixed, we can safely reason classically. However, some conclusions reached by reasoning in this manner are pretty pointless and irrelevant for the subject matter under discussion (because they are not wholly about it). A is true about m if mconditions do not preclude A from being true. If m = the concrete world and A = there are numbers, both A and -A turn out to be true about m, which is why we naturally hear someone claiming that there are numbers as not talking about the concrete world. Melia can thus claim that the claim that there are numbers is not part of our scientific theories in any interesting sense without committing himself to the abandonment of classical logic.12 Scientists are free to reason classically under the presupposition that there are numbers, and later discard as irrelevant those consequences of their reasoning that do not bear any relevance to the subject matter they are interested in.

80

Chapter Five

References Azzouni, J. 1997a, “Applied Mathematics, Existential Commitment and the Quine-Putnam Indispensability Thesis”, Philosophia Mathematica 5, 193 —. 1997b, “Thick Epistemic Access: Distinguishing the Mathematical from the Empirical”, Journal of Philosophy 94, 93 —. 2004, Deflating Existential Consequence: A Case for Nominalism, New York: Oxford University Press —. 2009, “Evading truth commitments: The problem reanalyzed”, Logique et Analyse 206, 139 —. 2011, “Nominalistic content”, in Cellucci, C., E. Grosholz and E. Ippoliti (eds.), Logic and knowledge, Newcastle upon Tyne: Cambridge Scholars Publishing Clark, M.J. and D. Liggins 2012, “Recent work on grounding”, Analysis 72, 812 Colyvan, M. 2010, “There is no easy road to nominalism”, Mind 119, 285 Dorr, C. 2010, “Of Numbers and Electrons”, Proceedings of the Aristotelian Society 110 (2pt2), 133 Field, H. 1980, Science without Numbers: A Defence of Nominalism, Oxford: Blackwell —. 1989, Realism, Mathematics and Modality, Oxford: Blackwell Fine, K. 1994, “Essence and Modality”, Noûs Supplement 8, 1 Hellman, G. 1989, Mathematics Without Numbers, Oxford: Oxford University Press Liggins, D. 2012, “Weaseling and the content of science”, Mind 121, 997 —. 2014, “Grounding and the indispensability argument”, Synthese http://link.springer.com/article/10.1007/s11229-014-0478-2 Melia, J. 2000, “Weaseling away the indispensability argument”, Mind 109, 455 Miller, K. 2010, “Three routes to contingentism in metaphysics”, Philosophy Compass 5, 965 Raley, Y. 2012, “Why the weasel fails”, Philosophia Mathematica III (20), 339 Yablo, S. 2001, “Go figure: a path through fictionalism”, Midwest Studies in Philosophy 25, 72 —. 2002, “Abstract objects: a case study”, Philosophical Issues 12, 220 —. 2005, “The myth of the seven”, in Kalderon, M. (ed.), Fictionalism in Metaphysics, Oxford: Clarendon Press, 88

The Content Challenge to Nominalism about Mathematics

81

—. 2006, “Non-catastrophic presupposition failure”, in Thomson J.J. and A. Byrne (eds.), Content and Modality: Themes from the Philosophy of Robert Stalnaker, Oxford: Oxford University Press —. 2009, “Must existence-questions have answers?”, in Chalmers D., D. Manley and R. Wasserman (eds.), Metametaphysics, Oxford: Oxford University Press, 507 —. 2010, Things: Papers on Objects, Events, and Properties, Philosophical Papers, vol. II, Oxford: Oxford University Press —. 2012, “Explanation, Extrapolation, and Existence”, Mind 121 (484), 1007 —. 2014, Aboutness, Princeton: Princeton University Press

Notes 1 It is worth noting that this does not apply to the easy-road strategy presented by Jody Azzouni (1997a, 1997b, 2004), which is not based in any way on retracting implications of our best scientific theories, but rather on the denial of Quine’s criterion for ontological commitment. 2 Numerical quantifiers (“There are n Fs”) are definable in the usual recursive manner without quantifying over numbers: there are 0 Fs = there are no Fs; there are n + 1 Fs = there is a y such that Fy and there are n x such that x is different from y and Fx. 3 I am ignoring here the possibility of treating expressions such as “etc.” or “…” as part of nominalistic language. 4 See Azzouni (2009, 159): “we can tell the difference between the genuine communication of something consistent by something sounding contradictory, and actual contradiction when and only when the proxy norm is satisfied (along with the informativeness condition)”. 5 I am considering only that case where the remainder of A-B (if there is such a remainder) is true. 6 I am assuming a conception of logical space according to which the existence of mathematical objects is contingent relative to the way the concrete world is. That this can be reconciled with the assumption that mathematical entities are either necessary or impossible has been argued by many: see Yablo (2012, 1013) for references. See also Miller (2010) for a discussion of contingentism related to the existence of mathematical objects and Fine (1994) for the notion of “possible relative to…”. 7 It is worth noting that Field (1989, 198) makes a point similar to that of Azzouni. 8 This strategy resembles that of Hellman (1989), as Dorr himself (2010, fn. 21) notes. 9 In the case of Azzouni, one should avoid talking about platonistic facts and rather say that Azzouni should acknowledge the grounding of mathematical truths in nominalistic truths.

82

Chapter Five

10 Part of the problem here seems to be the fact that according to Azzouni (2009, 173) mathematical fictionalists deny that mathematical claims are truth-apt. This is misleading: as Yablo (2010, 3-6) clarifies, according to hermeneutic fictionalism, when uttering a mathematics-infused sentence S, we quasi-assert S to assert its nominalistic content ||S||. ||S|| is defined as the set of worlds w such that w is concretely indiscernible from a world w' in which S is true. This means that in most cases, ||S|| has one and only one truth-value, which seems to be adequate for making it truth-apt. 11 The part of the content of S that is about m is the proposition true in w iff S is true about m in w. Remember there can be cases where there is no such sentence S* such that S* expresses the part of S that is about m. 12 It is worth noting that this also suggests that Melia can accept Raley’s third way out—retracting the sentences that existentially quantify over numbers. For Raley and Azzouni (2011, sect. 5), making this move would be tantamount to rejecting Quine’s criterion of ontological commitment. In a sense, this is true: one of the points Melia makes in his paper is that we should not believe in the existence of those entities that we indispensably quantify over in our best scientific theories. However, this does not mean that we reinterpret the meaning of the existential quantifier when it is used inside a mathematics-infused sentence. Melia and Yablo’s point is not that existential quantification does not express ontological commitment, but that uttering “there are numbers” in the context in which we are talking about the concrete world and presupposing the existence of numbers it’s claiming something trivial, similar to “the concrete world is compatible with there being numbers” or “on the assumption that there are numbers, there are numbers”.

CHAPTER SIX IS THERE SUCH A THING AS METAPHYSICAL KNOWLEDGE? CLAUDINE TIERCELIN

Philosophers have always wanted to answer such acute challenges as those raised by relativisms and idealisms of all stripes, which constitute a permanent threat to our genuine knowledge of the world. Such is the “Integration Challenge” which any rational being has to face, whenever he wishes to clarify the link there is between the ways in which we come to know the truths of such and such domain and the ways in which our beliefs hook onto the facts of such and such domain and may be then entitled to be taken as proper “knowledge” claims (Peacocke 1999; Tiercelin 2012). Indeed, we have to determine what we know, when and if we do know something. In what follows, I would like to make some brief suggestions to start bridging the gap between our epistemology and our metaphysics. Viewing things mainly from the metaphysical side of the challenge, I would like to argue that even if the path is rather narrow, between false modesty or cowardice, and temerity or arrogance, it is indeed not only legitimate, but necessary to follow a deliberately metaphysical stance, along, I shall suggest, dispositional realistic lines. In so doing, I hope to give some reasons to prefer boldness to humility in our claims to a genuine metaphysical knowledge of things, of their real nature and of their properties.

1. Some good arguments in favour of humility No doubt, humility in metaphysics has, for it, a rich history and its days of glory, in particular, with the empiricists (Hume, Locke), and also, of course, with the Kantians and contemporary neo-Kantians. We cannot go beyond ideas derived from our sensible impressions (empiricist and

84

Chapter Six

sceptical motto); our knowledge is situated within the limits of all possible experience and mere phenomena (Kantian version). The very nature of things is forbidden to us, in terms both of our access to it, and of its contents. Kantians and Humeans are on the same boat: what we come to know are not the things in themselves, their intrinsic properties, but mere phenomenal appearances. Either because of the receptivity of our senses, or due to our finite cognitive faculties, we cannot get out of the constraining frames of representation. Whoever dreamt of natures, substantial forms, or for-ever fixed essences, would live in the wrong century. Even if all things are not totally construed or made up, the relativists say, what we deal with are family resemblances and games, floating and vague identities, mere analogies, bits of things which we cannot approach except through our descriptions, fictions, and tales. At best, when the tales are being told by science, as some philosophers of science tell us, what we have some access to, is some sort of “veiled” reality. But the heart, even more, the cement of things will remain forever unknown to us. A widely shared view represents the world as a set of atomic entities, with a spacetime location, endowed with irreducible fundamental properties, made of solitary mosaics, peopled with individuals or substances, characterized by their intrinsic properties. In fact, this may be traced back to Aristotle. However, the world then was constituted by dynamical substantial forms. Through the sieve of mechanism, we are now facing a world of passive entities, linked one to the other through merely contingent relations. Which deepens the gap between epistemology and metaphysics, and induces, in the end, such forms as those illustrated by Kantian humility or Humean scepticism. Thus, it is not surprising that a common favourite target, for those who, in contemporary metaphysics, prefer powers and capacities, should aim at Hume’s (as much as Kant’s) model, in the guises it has taken in David Lewis’s approach.

2. From Kantian Humility to Lewisian Humility Indeed, an acute awareness lies at the heart of Kantian humility, either of two classes of things: phenomena (which are knowable), and things in themselves (which are unknowable), or of two forms of epistemic access, according to which the only possible access is constituted by an empirical relation of affection through appearances (Langton 1998, Whittle 2006, Lewis 2009). But we find also, through a more realistic reading of Kant, the recognition of two classes of properties of things: the relational ones (irreducible to their relata, contrary to what Leibniz might have thought)

Is There Such a Thing as Metaphysical Knowledge?

85

which we can know, and the intrinsic ones, the knowledge of which is forbidden to us (Langton 1998, 12-13). Hence, Kant’s “epistemic humility” is somewhat moderate: our knowledge is constrained by a limited accessibility. While ignoring the nature of substances and of their intrinsic properties, we know that there are independent substances thanks to their relational properties which do affect us. The situation is worse with those philosophers who, even when they consider that relational properties and, in particular, causal powers melt into or supervene on intrinsic properties, find that such a connexion is contingent and does not provide any knowledge whatsoever of intrinsic properties (Langton 1998, 175-180). See, e.g., F. Jackson: When physicists tell us about the properties they take to be fundamental, they tell us what these properties do. This is no accident. We know about what things are like essentially through the way they impinge on us and our measuring instruments. It does not follow from this that the fundamental properties of current physics, or of “completed” physics, are causal cum relational ones. It may be that our terms for the fundamental properties pick out the properties they do via the causal relations the properties enter into, but that at least some of the properties so picked out are intrinsic. They have, as we might put it, relational names but intrinsic essences. However, it does suggest […] the uncomfortable idea that we may know next to nothing about the intrinsic nature of the world. We know only its causal cum relational nature (Jackson 1998, 23-24).

Clearly, it is neither idealism nor relativism here which lead to such a conclusion, but rather, some conception of the links between relational and intrinsic properties. And this is what leads Lewis to humility, too. If one thing is the basis, another one the causal power, and if the link between the two of them is contingent, then, indeed, we have no access to the intrinsic properties; we face a something-we-know-not-what, and we can only identify the properties through the role they occupy (Langton 1998, 176; Whittle 2006). Lewis (2009) does not seem to be moved by this. But is he right? On the epistemic level, he may be: fundamental properties are taken to be simple (not complex) entities, or again, “structural properties”, i.e. properties composed of other properties (Lewis 1986, 8), of which it is impossible, more or less, to say much more. After all, explanations have to stop somewhere. If we were to reach the bedrock of fundamental properties, we would have reached the final end of scientific explanation: Optimists hope and expect that we will discover the final theory someday soon, or anyway someday. I share their hope and expectation, but I am not

86

Chapter Six assuming it. Maybe scientific research may go out of fashion. Maybe the task of fully understanding the workings of nature is just too hard for us […] (Lewis 2009, 219).

Now, as everybody familiar with Lewis’s epistemology knows, it is implied by the final scientific theory that the universe should not be an indefinitely complex location; otherwise, science would never be completed. No doubt then, some kind of humility is up to the point, since we could never reach a complete knowledge of the properties. However, this would not be because of the intrinsic nature of the properties, which we can never, in principle, discover; rather it would be because, as far as we might go in scientific reasoning, we could only reach structural properties which we would have to analyse further on, ad infinitum. Since Lewis does not grant such a hypothesis, he must admit that the fundamental properties do not have a complex nature and are simple. Let us note, incidentally, that the properties might have a complex causal nature, because of the nexus of complex powers which they endow the particulars they instantiate with. However, as Lewis endorses the Humean principle of quidditism, he cannot accept such a hypothesis. Since the property which realizes a certain nomological role can realize different roles according to the possible worlds, the causal features of a property cannot be part of its intrinsic nature (as rightly pointed out by Whittle 2006). So, it may well be the case that we do not know such a simple nature for the fundamental properties, which would be identifiable in this world and in other possible worlds. But we have the means to describe them through the nomological roles which they occupy. Hence ignorance is not such a big deal. When it comes to the metaphysical level, the situation may be less simple. Let us assume, indeed, as a primitive fact, such a Humean universe of intrinsic physical properties, contingently, although regularly, distributed along spacetime points (see Beebee 2006). How then, if there is nothing more in the nature of such properties, although they are supposed to serve as a basis of supervenience, can we explain their having so different nomological roles? Is it not surprising that such properties, in their intrinsic nature, in their identity, should have no causal bearing on what they do, and more generally (a fortiori if they are taken as such entities as are assumed by science), on what takes place in the world? Properties are supposed to have a functional role, which allows them to contribute to the causal nexus as a whole; but how should we understand the relationship between the intrinsic nature of a property and its functional aspect? It may be the case that the laws of nature are called in, to serve as an answer. However, if such laws are mere regularities

Is There Such a Thing as Metaphysical Knowledge?

87

included within some Lewisian “best system”, the natures may happen to combine with very different functional roles (Whittle 2006). Then, to use Sellars’s terms, the intrinsic natures of properties seem to be as “substitutable the one to the other, as pennies, whatever the context may be” (1948, 300). Humean metaphysics claims to be an empiricist one and to reject any kind of entities that might not be within reach of scientific characterization. Yet, it looks as if it does generate the greatest amount of mysteries. If rationality is something we are aiming at, then, we have to make another choice.

3. Another route Hence, I suggest another route along dispositional and realistic lines (as the ones I have detailed in Tiercelin 2011, 247-359), which relies on four main assumptions: 1. A causal theory of properties. 2. A conditional dispositionalist account of laws. 3. Some kind of aliquidditism (or thin essentialism). 4. An emphasis put not only on efficient causation but on other, maybe teleological, aspects of causation. In a nutshell: first, such a dispositional realism (hereafter DR) is committed to the view that there are real universals, mainly understood not as Platonic independent, uninstantiated universals (metaphysical realism), but rather along Aristotelian or more precisely Scotistic lines: the real is mainly defined as that which signifies something real. Secondly, DR goes together with a semantic realism according to which one should first clarify the meaning of all our concepts, in particular the concept of causation, and determine the meaning of our dispositional attributions, so as to understand, for example, why reducing dispositional attributions to conditionals does not work, and why reduction sentences cannot tell “everything” that is meant by our dispositional predicates. Thirdly, DR should take the form of a scientific DR, which aims at finding real properties and not mere predicates. Fourthly, such a DR assumes scientific realism, as an abductive hypothesis called for by the explanatory necessity of science, and assumes certain real universals. Last but not least, such a scientific DR is essentialist, although not substantialist; it calls, in particular, for a new definition of “essence” (though a “thinner” one), viewed less as a static quidditas, or pure natural kind or mere bundle of habits, than as a habit-dispositional aliquid, close, in many respects, to a form of relational or structural realism. Within the limits of this paper, I can only elaborate a little more this latter feature. In order to be able to defend DR, we have to be clear about the fundamentum of the dispositional realities we put forward and, to that end, have to favour some form or other of thin essentialism. Indeed, if we

88

Chapter Six

conceive essentialism merely as a natural methodological strategy (or because, as George Molnar 2003, for example, claims, anti-essentialism sounds “so counter-intuitive”), we merely go half of the way. On the other end, if we adopt such a view as Brian Ellis’s scientific essentialism (2001), for instance, it may well turn out also that we do not get exactly what we are looking for, not so much because Ellis’s position, despite its great merits, relies on a rather obscure notion of essence, nor for the reason advocated by Mumford that it remains possible to accept natural kinds into one’s ontology without accepting their corresponding essences, as because the concept of essence Ellis relies on has more to do, as Alice Drewery (2005) convincingly argued, with issues related to necessity than with what one might expect from a genuine definition of essence, and cannot, in particular be used to ground what Ellis claims it does, namely the necessity of the laws of nature. Indeed, we should obviously distinguish, which Ellis and many contemporary essentialists do not, between: Df1: F is a necessary property of a iff a has F in all possible worlds that include a. Df2: F is an essential property of a iff being F is constitutive of the identity of a. Again, the natural necessities introduced by Mumford (2004) to ground his lawless realism do not give exactly what we need to identify the essentially dispositional nature of our properties, and it is also doubtful that causal powers should by themselves be sufficient and could dispense us with laws. We do need laws, and even more, if some or other kind of dispositional essentialism is right, most probably (at least for some of them) necessary, not contingent, laws of nature. In particular, Bird (2007) is right in his objections to Mumford’s criticism of laws according to which laws should be understood as governing rather than descriptive laws and that science would not have a really unified concept of law (so that, for Mumford, we could just as well give it up too!) In that respect too, Bird’s essential or “relational” dispositionalism is more congenial to the kind of dispositional realism I would like to defend, all the more so as Bird is inclined to consider structural properties as being fundamentally relational, thus in perfect convergence with what science tends to show. And it may well be the case that, contrary to what is most often assumed today, essentialism should develop along relational lines rather than in the wake of substantialist models which remains closer to traditional Aristotelian logic than to what contemporary logic has taught us, by

Is There Such a Thing as Metaphysical Knowledge?

89

insisting, in particular, on the importance of relations, and, at all events, on the limits encountered by any simple subject-attribute conception. To begin with, it might be worth reminding the ways in which some scholastics, and Duns Scotus, for example, viewed essentialism, in a somewhat different way from Aristotle. Following Avicenna, he insisted on the neutrality or irreducible and positive indeterminacy of essence. Crucial to Avicenna was the view, not so much that essence, as such, could be considered under two headings: in things and in the intellect, as the fact that it could be taken as such, in its pure essentiality, neither universal nor singular. The essence (in Avicenna’s or Scotus’s terms, the “Common Nature”) was characterized by such neutrality or indifference to any further possible determinations. For Scotus there are formal or metaphysical realities which are not to be viewed as what we call today “primitive thisness”, precisely because they are, so to speak, awaiting further both physical and logical determination. The focus is less on thinking of essence independently of its properties (which belong to it properly), i.e. in distinguishing the essence from what makes it a particular substance, than on showing how what is more an aliquid than a quidditas or a substratum without substance is necessary in order to ground, then, on the logical level, logical universality, and on the physical level, the quiddity of things. So, to be a realist means neither to hypostasize Platonic essences, nor to develop a form of essentialism simply devoid of the Aristotelian substantialist shape: it means first and foremost, to distinguish logical reality and real community, to grant the irreducibility of a Common Nature which, in itself, is neither universal nor singular, although it is universal in the mind, and singular in the things outside the mind. For sure, “quidditism” is not viewed as a very attractive position nowadays. As John Hawthorne (2001) has noted, for causal structuralists in particular, quiddities are a “will o’ the wisp”: or a way to say that I could have been a poached egg, no matter, so long as my haecceity was present. But there is a confusion, here, or even a full misunderstanding of what, for the scholastics, quiddity and haecceity meant. Contrary to the view we now have of haecceitism, haecceitas was introduced by Scotus precisely to differentiate, formally, the singular from the universal, or the Common Nature. Again, in order to be clear about the various categories that populate our world, we should be careful not to confuse the logical, the physical and the metaphysical levels of our investigation, but also, to establish the right alphabet of being (and it may turn out that there are more than one or two kinds of “essential properties”). In particular, as far as the latter point is concerned, one should note that, even if (as both Scotus and someone like C.S. Peirce, who was a close follower of him, for

90

Chapter Six

example, argue), material essences are dispositional, it does not necessarily follow, however, that all dispositional properties are essential. The fact that “X is hard” need not be essential to X, even though hardness is a dispositional property causing X to behave in certain predictable ways (Raposa 1984, 158). More specifically, we might identify the real nature of a thing (the essential “meaning” of the concept of that thing) with the complete set of habits that govern its behaviour (thus no longer distinguishing between the essence and the accidents of the thing). But then, in order to have a well realized essentialist world, and not a mere mosaic of essences either viewed as static quiddities (or as mere natural kinds) or as “bundles of habits” fixed by some mysterious “glue”, as Peirce tried to build in a consistent way out of the Scotistic model, one should be able to account for the real “binding” between the various essences. If, in the end, all properties are essentially defined as dispositional, namely as entirely reducible to sums of causal powers, how are such powers in turn supposed to be linked together, and again, how are they supposed to be really causal? It may well be that in order to account for this, something more is needed than first, mere natural kinds, second, mere efficient causation. Let me elaborate this a bit more. Indeed, if we only have natural kinds and not essences, then it becomes difficult to understand what is the real source of the intelligibility of a thing (“operari sequitur esse”). Now the purpose of the quiddity is precisely to specify, for any given object, the kind of thing it is. “The very meaning of a word or significant object ought to be the very essence of reality of what it signifies”, Peirce says (5.429). Now, as he himself claimed, against the too static view of essence as defined by Scotus, it is not the behaviour of a thing but rather its habit of behavior that constitutes the intelligible nature or real essence (2.664). Such a habit is a general disposition affecting the way that an object would tend to behave under certain types of circumstances. Both Scotus and Peirce distinguished between the essence and the activities of a thing. But although Scotus and the medieval logicians were able to deal with propositions that involved monadic predicates (like “… is hard”) they were unable to deal with those that involved relational predicates (such as “… is a lover of …”, “ or “… gave … to …”) (3.464 ff; Raposa 1984, 151). Hence, they could talk of specific classes or collections of things, each class being comprised of all the subjects bearing a particular monadic predicate, which allowed them to say something about the relation of similarity (i.e., the sharing of a “common nature”) between the members of a given class. But their logical analysis did not go far enough. As Peirce showed in his logic of relatives,

Is There Such a Thing as Metaphysical Knowledge?

91

we wish to analyse relationships other than that of resemblance of a certain object to the various members of its class. We want to make out the way in which laws govern the interactions between objects within a meaningful process. And the analysis of such a process or “system”, for Peirce, involved the use of dyadic and triadic predicates: to claim for example that “X is hard” is to do more than simply ascribe a particular quality; rather, it is to assert that under certain specifiable conditions, X will (or rather “would”) tend to behave in a certain specifiable manner. Thus, “hardness” is to be regarded as a dispositional property, and a real “habit” or “law” must govern the behaviour of those objects within which it inheres, thus defined as a “would-be” operating within the actual world of objects and events (Raposa 1984, 152). What is important is not to specify the generality that characterizes a collection of objects having some quality in common (what Scotus does), but to account for the infinite number of real possibilities, i.e. the real and continuous relationships that exist between any two members of a class, between an object and its successive actualizations in time, between the interacting fragments of a system. Peirce, incidentally following Francis Abbot at the time, was clearly arguing not only that there are real relations, but that relations comprise the real natures of things. Habits account for an object’s essential intelligibility. Habits are laws that govern objects by relating certain types of behaviour to specific kinds of circumstances. As a consequence, the essence of a thing is defined not by any particular relationship or activity within which the thing actually participates, but by a general habit that determines those relations and activities to which, given the appropriate conditions, that thing would be disposed. Such a habit is not simply essential to, but rather, must be of the essence of the thing, namely it must be predicated of the thing “per se primo modo” (2.361) (Raposa 1984, 154). From this, a major lesson may be drawn: if we are looking for the essence of a thing, and if that essence is no collection of properties, but rather, a special “habit of action”, more specifically “a bundle of habits” for a “law-cluster”, so that what a thing is, is intimately connected with what that thing is for, we may need more than mere efficient causation to account for the way in which it exerts its causal power as a whole.1 It means, first, that we should view matters in terms of a “specific idea” or “final cause” specifying the general patterns of behaviour that a given object or organism will tend to manifest, which amounts to saying that what a thing is may be best defined by what that thing is to become, or that the causal function of the essences of things may be appropriately defined in terms of both formal and final causation. But it also means, in

92

Chapter Six

the second place, that we may have to view the binding of all the objects itself—as Peirce and more recently Brian Ellis suggested—in terms of some final (or intentional?) causation.2 At all events, this needs to be carefully elaborated (as well as the exact roles played both by dispositions and by laws of nature in the intelligibility of nature). My contention here is that both are required: dispositions would find their intelligibility in the conditional necessity of laws: but laws would only be a true description of the world provided they were grounded in what things can do, or rather would do (in the sense of real metaphysical possibilia, which are necessary although discovered a posteriori.

4. Some conclusions Let me draw some conclusions from this, which, hopefully, will highlight the merits of DR over Kantian, Humian or Lewisian stances, as a means to reach genuine metaphysical knowledge. If the fundamental physical properties are categorical and intrinsic, we cannot even know that they are so. In which case, what reason do we have to suppose it? On the contrary, if we characterize properties in dispositional terms, then we have a cognitive access to their properties, through the effects they produce. If, following such philosophers as Shoemaker (1980), Mumford, Bird or Ellis, we emphasize that properties are dispositions or causal powers, we can show how they produce other properties, through the effects they generate, and also how they do it, not regularly, but necessarily, because they are not mere potentialities but real powers from which laws of nature proceed, in some cases necessarily, and not in an artificially imposed manner. Could we dream of a better agreement between metaphysics and epistemology? For sure, such a dispositional realist model for properties as the one I have offered is not the only one that can be provided, when one wishes to reject the neo-Humean model. In particular, it resembles some kinds of structuralism, which prefer a characterization in terms of structures rather than properties, be they categorical or dispositional. Like structuralism, whatever its variants might be, DR intends to follow the lead of science and the results of contemporary physics, while avoiding any straightforward deference to science and various traps into which some scientistic attitude might lead (Tiercelin 2011, 97-186; Psillos 2006). It is precisely in that first respect, that DR separates itself from ontic structural realism (OSR), which, being strengthened by an underdetermination of individuality, has become a kind of metaphysics of fundamental physics

Is There Such a Thing as Metaphysical Knowledge?

93

(non relativistic quantum mechanics, quantum field theory, and general theory of relativity mainly) (Esfeld 2004; Esfeld and Lam 2006; French and Ladyman 2003; Ladyman and Ross 2007, chaps. 2-4; French 2010). Secondly, when pushed too far, structuralism tends to be counterproductive: if there is nothing in the world but structure, what will it be supposed to be opposed to (Psillos 2006, 2012)? In general, structures in science were designed as entities with blank places which objects could occupy. But if these must be “reconceptualized” or are meant to have a mere “heuristic” function (French 1999, 204) or even to disappear (French and Ladyman 2003, 37), it is hard to understand what role the structure itself can still play (Psillos 2006; Chakravartty 2003). Thirdly, conceived at first along mathematical terms (French 1999), and with an eagerness to “erase the lines between mathematics and physics” (French and Ladyman 2003, 41), structuralism intends structures to play two roles: but is it successful in this? In other words, is it abstract enough to be independent of the concrete physical systems, and at the same time, concrete enough to be part of the causal identity of those systems (Psillos 2006; 2012, 170)? Most of all, even if dispositional realism asserts the non supervenience of relations on the objects, and claims that objects do not have any existence or identity independently of the relations they have with one another (Ladyman and Ross 2007, chaps. 3.4; French 2010), it is not ready to accept the pure disappearance of the objects, which is advocated by some eliminativist ontic structural realists, who claim that relations are not merely primary in relation to objects which are literally constituted by them or simple “nods” within structures in a relation of asymmetric dependence (Ladyman and Ross 2007, chaps. 2 to 4). For them, there are no objects any longer: only relations or structures, i.e. relations without relata. “Relational structures are ontologically more fundamental than individual objects: all there is, is structure” (Da Costa and French 2003, 189). Now, several reasons (and not only common sense) militate in favour of maintaining such a category in our ontology. A metaphysical reason, first: without relata, relations have no reason of being; even if such relata have not necessarily any intrinsic identity; an empirical reason, then: the physical characteristics on which one relies do not in the least suggest to abandon such a commitment for objects in the fundamental physical world; and finally a logical reason, which has to do, as Esfeld and Lam have mentioned, with quantifying over objects in standard first order logic and the apparently unavoidable use of set theoretic concepts in physical theories (Esfeld and Lam 2011). If one tries to pull too far the very meaning of our primitive concepts of “real” and of object, we run the risk of rendering the world simply unintelligible (see Heil 2003, 58-60).

94

Chapter Six

Hence, DR clearly has affinities with causal structural realism (CSR), which is, in many respects, more convincing than “ontic” structuralism, in particular in its “moderate form” (Esfeld 2004; Esfeld and Lam 2006). Indeed, it has two main virtues: while giving ontological priority to relations, it does not deny that properties and objects are part of a fundamental ontology; simply, as I would claim myself, such properties are not intended to be intrinsic, they may be relational or extrinsic. If there are physical relations between objects or relata, such objects have themselves relational properties. The universal context of entanglement and non separability in quantum mechanics is fully assumed. However, a principle of weak discernability, although much disputed, is also admitted, viewed as a symmetric and irreflexive relation between two objects (hence there are two objects and not only one). A second important virtue of CSR is to remind us how important it is for properties to have a causal profile. Now, OSR was aware of the need to introduce some “modality” within structure, but it failed in trying to do so (French 2006, 18). One could hardly see whether the type of causation introduced applied to particular relations or to an underlying causal activity which, so to speak, would endow the relevant relations with causal powers (and if so, how?) Hence, as noted by Psillos 2012, it made it hard to choose between the first way, leading to a causal structuralism simpliciter, and the second one, condemning either to mystery (so as to give the needed modal force to the relational structure) or to causal hyperstructuralism (Hawthorne 2006, 223). Causal profiles thus became purely structural, and one ended into a merely formal structure devoid of any causal profile. In short, one was brought back to one’s starting point. In so far as it stresses that properties are well identified through their causal roles and that the structures are defined as a network of causal relations among properties, hence by the causal powers which they confer to their possessors, CSR has a twofold merit (which I have underlined myself): it avoids both the circularities and mysteries of OSR and of Lewisian humility: if all the properties do not have any intrinsic property beyond their causal profile, if all the causal powers are the kind of entities which make things happen, and if knowledge requires some causal contact with the things we know, then our knowledge of properties is in principle possible. However, and this is, I claim, a crucial flaw, it remains to be shown how it handles a problem which any kind of dispositional monism has to face, obliged as it is, to follow a holistic model. Now, causal structuralism is indeed a structuralism, rejecting any form of quidditism, or the view according to which there would be “something” beyond the causal profile

Is There Such a Thing as Metaphysical Knowledge?

95

which, independently of it, in so far as it might exist, would make of that property what it is.3 If no property can be identified unless all the others are, it looks as if none of them can be identified simpliciter. We hoped to understand the identity of properties while avoiding unknowable quiddities, we have merely moved the problem to another place, since what we come to is a holistic network of relations among properties which seems even more mysterious and which is not more able to identify the properties. Quiddities have not disappeared: they have become, a global “totusity” (Psillos 2012). It is not only the case that our reasons for humility have not been overridden; we should rather say that humility has now turned into temerity. This may explain, for example, why Mumford, in order to break the circle to which his causal dispositionalism commits him, appeals to the view that some powers, may, at the end of the day, be identified by the effects that the “phenomenal appearances have on us, and on our sensory modalities” (2004, 187). In the same way, Chakravartty observes that “any case of warranted attribution of a causal property is facilitated by some properties which are being known independently of a knowledge of their other effects” (2007, 136). This seems plainly to grant that, in all cases, the conditions of individuation of the causal powers which are assured by the place they occupy in the global network can only be so provided that the identity of “some properties or relations is fixed independently of the place they occupy in that network”. And it also means that causation itself must be a relation identified independently of the role it plays in the causal network, even if one runs then the risk, from the point of view of structuralism, to turn this time into some kind of “hypostructuralism” (Psillos 2012). Again, there is something preposterous to consider that “one can in principle discover what properties are through the effects they produce” (Esfeld 2009, 184), and that this applies to “all” the properties. First, to suppose that the real is knowable, at least in principle, does not imply that “everything” in the real is. There are “ultimate” facts which any one, be he a man of science or any man in the street has to take account of (Peirce, 1.405) and in particular, such isolated facts do not imply any explanation whatsoever (Peirce, 7.200; 7.194). Secondly, one should never underestimate the length, the complexity and sometimes even the tricks of the various chains through which we come to discover the causal properties, some being too far from one another, some being hidden by the screen some others may constitute. Besides, even granted that the total network of the causal profiles might be knowable, how could we ever know that it is indeed such properties that play such role in the totality? Finally, what the limitations of CSR show, is also, I think, to what extent it is illusory and

96

Chapter Six

mistaken to think that one can, in the end, do without quidditism. In the moderate version Esfeld has come to, he writes that “physical structures are networks of qualitative and concrete physical relations between objects which are nothing more than what lie in those relations, in other words, which do not possess any intrinsic identity over the relations in which they are” (2009, 180), but he still holds that relations do need relata and that objects are necessary and insists on a simple modal or conceptual (though not ontological) distinction between the relations and the objects (Esfeld and Lam 2011). However, I doubt that such a concession is enough to provide genuine identity conditions, and to allow, in particular, to distinguish between the essential and the accidental parts of causal powers. Without some kind of “aliquidditism”, one cannot go all the way down: to say what the fundamentum of things is, one needs more than modal or conceptual distinctions, even in a Spinozist guise. It takes more than conceptualism to be able to say in what a thing consists, what its real being is, to talk like John Locke. Such a real being, its identity is what makes the thing, the thing it is. Any radical anti-essentialism would take us to such a global anti-realism, that is would surely be incoherent (Lowe 2007, 92). Without a minimal essentialism, or a “serious essentialism” (2007, 86), viewed not as an ersatz essentialism of possible worlds, nor as an essentialism of act and potency, but as what allows to specify, for each object, the very being of the reality it signifies, which was Locke’s (Essay, III, 3, § 15), as well as Aristotle’s definition, I doubt very much that we might claim, not even to know, but merely to understand what is at the root of the intelligibility of things. This is why, relying as it does, on a causal and dispositional theory of properties, on a conditional dispositionalist account of laws, together with a scholastic realism involving a certain aliquidditism and a new account of causation, a version like the one I have tried to sketch in terms of a scholastic dispositional realism, may be a way to overcome metaphysical humility and to promote a reasoned form of metaphysical boldness, or at least, against all kinds of Humean, Kantian, Stroudian scepticisms or subtle forms of neo-Pyrrhonism, a way to favour metaphysical engagement, optimism and satisfaction.

Is There Such a Thing as Metaphysical Knowledge?

97

References Beebee, H. 2006, “Does anything hold the world together?” Synthese 149, 509 Bird, A. 2007, Nature’s Metaphysics: Laws and Properties, Oxford: Clarendon Press —. 2009, “Structural Properties Revisited”, in Handfield, T (ed.), Dispositions and Causes, Oxford: Oxford University Press, 215 Chakravartty, A. 2003, “The structuralist conception of objects”, Philosophy of Science 70, 867 —. 2004, “Structuralism as a form of scientific realism”, International Studies in the Philosophy of Science 18,151 —. 2007, A Metaphysics for Scientific Realism: knowing the unobservable, Cambridge: Cambridge University Press Da Costa, N. and S. French 2003, Science and Partial Truth, New York: Oxford University Press Dipert, R.R. 1997, “The mathematical structure of the world: the world as a graph”, Journal of Philosophy 94, 329 Drewery, A. 2005, “Essentialism and the Necessity of the Laws of Nature”, Synthese 144, 381 Duns Scotus, J. Ordinatio (Ord.) and Lectura (Lect.), in Opera omnia Civitas Vaticana: Typis Polyglottis Vaticanis, 1950-; Reportatio I-A (Rep.), The Examined Report of the Paris Lecture. Reportatio I-A, A.B. Wolter and O. Bychkov (eds.), vol. 2, St. Bonaventure, N.Y.: Publications of the Franciscan Institute, 2008; Quaestiones super Metaphysicam and Quaestiones super Praedicamenta are taken from his Opera philosophica [OPh], St. Bonaventure, N.Y.: Publications of the Franciscan Institute, 1997-2006 Ellis, B. 2001, Scientific Essentialism, Cambridge: Cambridge University Press —. 2005, “Universals, the Essential Problem and Categorical Properties”, Ratio 18, 462 Esfeld, M. 2004, “Quantum entanglement and a metaphysics of relations”, Studies in History and Philosophy of Modern Physics 35B, 601 —. 2009, “The Modal Nature of Structures in Ontic Structural Realism”, International Studies in the Philosophy of Science, 23, 179 Esfeld, M. and V. Lam 2006, “Moderate structural realism about spacetime”, Synthese 160 (1) 27 Esfeld, M. and V. Lam 2011, “Ontic structuralism as a metaphysics of objects”, in A. and P. Bokulich (eds.), Scientific structuralism, Dordrecht: Springer, 143

98

Chapter Six

French, S. 1999, “Models and Mathematics in Physics”, in Butterfield J. and C. Pagonis (eds.), From Physics to Philosophy, Cambridge: Cambridge University Press, 187 —. 2006, “Structure as a Weapon of the Realist”, Proceedings of the Aristotelian Society, 106, 1 —. 2010, “The interdependence of structures, objects and dependence”, Synthese 175 (1) suppl., 89 French, S. and J. Ladyman 2003, “Remodelling structural realism: quantum physics and the metaphysics of structure”, Synthese 136, 31 French, S. and M.L.G. Redhead 1988, “Quantum physics and the identity of indiscernibles”, British Journal for the Philosophy of Science 39, 233 Hawthorne, J. 2001, “Causal Structuralism”, Philosophical Perspectives: Metaphysics 15, 361 —. 2006, Metaphysical Essays, Oxford : Oxford University Press Heil, J. 2003, From an Ontological Point of View, Oxford: Clarendon Press Jackson, F. 1998, From metaphysics to ethics. A defence of conceptual analysis, Oxford: Oxford University Press Ladyman, J. 1998, “What is structural realism?”, Studies in History and Philosophy of Modern Science 29, 409 Ladyman, J. and D. Ross with D. Spurrett and J. Collier 2007, Every Thing Must Go. Metaphysics naturalized, Oxford: Oxford University Press Langton, R. 1998, Kantian humility. Our ignorance of things in themselves, Oxford: Oxford University Press Langton, R. and D. Lewis 1998, “Defining ‘intrinsic’”, Philosophy and Phenomenological Research 58, 333 Lewis, D. 1986, “Against Structural Universals”, repr. in Philosophical Papers, Vol. 2., Oxford: Oxford University Press —. 2009, “Ramseyan humility”, in Braddon-Mitchell D., Nola R. and Lewis D. (eds.), The Canberra programme, Oxford: Oxford University Press, 203 Lowe, E.J. 2007, “La métaphysique comme science de l’essence”, in Garcia, E. and F. Nef (eds.), Métaphysique contemporaine. Propriétés, mondes possibles et personnes, Paris: Vrin, 85 Molnar, G. 2003, Powers: A Study in Metaphysics, Oxford: Oxford University Press Mumford, S. 2004, Laws in Nature, London: Routledge Peacocke, C. 1999, Being Known, Oxford: Clarendon Press

Is There Such a Thing as Metaphysical Knowledge?

99

Peirce, C.S. 1931-1958, The Collected Papers, 8 vols., Cambridge (MA): Harvard University Press (all references are cited by number of volume followed by number of paragraph) Psillos, S. 2006, “The structure, the whole structure and nothing but the structure”, Philosophy of Science 73, 560 —. 2009, Knowing the Structure of Nature: Essays on realism and Explanation, Palgrave: MacMillan —. 2012, “Adding Modality to Ontic Structure: an Exploration and Critique”, in Landry E. and D. Rickles (eds.), Structure, Object, and Causality, Structural Realism, Western Ontario Series in Philosophy of Science; Dordrecht: Springer, Vol. 77, 169 Raposa, M. 1984, “Habits and Essences”, Transactions of the C.S. Peirce Society, vol. XX (2), 147 Sellars, W. 1948, “Concepts as Involving Laws and Inconceivable Without Them”, Philosophy of Science 15, 287 Shoemaker, S. 1980, “Causality and Properties”, in van Inwagen P. (ed.), Time and Cause, Dordrecht: Reidel, 109, repr. in Shoemaker S. 1984, Identity, Cause and Mind, Cambridge: Cambridge University Press Stroud, B. 2011, Engagement and Metaphysical Dissatisfaction Modality and Value, Oxford: Oxford University Press Teller, P. 1986, “Relational holism and quantum mechanics”, British Journal for the Philosophy of Science 37, 71 Tiercelin, C. 2011, Le Ciment des choses, Paris: Éditions Ithaque —. 2012, La connaissance métaphysique, leçon inaugurale du Collège de France, Paris: Éditions Fayard Whittle, A. 2006, “On an Argument for Humility”, Philosophical Studies 69, 1

Notes 1

“Efficient causation is that kind of causation whereby the parts compose the whole; final causation is that kind of causation whereby the whole calls out its parts. Final causation without efficient causation is helpless […] efficient causation without final causation, however, is worse than helpless, by far; it is mere chaos; and chaos is not even so much as chaos, without final causation; it is blank nothing” (1.220). Hence, a genuine scientific explanation of a system will involve the discovery of the final end or cause that defines it, rather than the detailed analysis of how its components function (as “efficient causes”) to bring about its normal operation. 2 This may have to do with what Ellis calls his compatibilist thesis, i.e. the view that “a property can have a causal power without either being a causal power, or

100

Chapter Six

being ultimately reducible to causal powers. For even the most fundamental causal powers in nature have dimensions […] For the dimensions of the causal powers are not properties that things could have independently of the powers. They simply would not exist if the powers did not exist. Nevertheless, most of the powers that do exist in the world could not exist without them […] we do not live in a dimensionless world, and how things act and interact with each other depends on how they are distributed and oriented, and what powers, capacities and propensities they all have” (Ellis 2005, 470-471; italics mine). 3 Causal dispositionalism, as Mumford e.g sees it, claims that “the identity of a property is fixed by the causal role it plays in relation to other properties”. Indeed, if the identity of a property is “fixed by the relations it has to other properties, its existence does not ontologically depend on those properties” (Mumford 2004, 171). However, it remains true that holism has to be granted, namely “a world viewed as a simple whole, composed of properties whose essence and identity are determined by their place in that whole” (164). Or again, “the properties that are real in a world must […] form a network of interconnections: a system where no property is alone or outside”. We face a similar difficulty in the graphical model (inspired by Peirce and Dipert 1997) proposed by Bird to account for the relations between causal powers or potencies (potencies) while disarming all the objections that any holistic view has to face in relation to the individuation of properties (Bird 2007, 142; 2009).

CHAPTER SEVEN THE IDENTITY OF INDISCERNIBLES AND THE PRINCIPLE OF NO CO-LOCATION ROBERTO CASATI AND GIULIANO TORRENGO

1. Black’s two spheres The most famous counterexample to the principle of identity of indiscernibles PII is Black’s perfectly symmetric universe containing only two perfectly similar spheres. (PII) For every x and y, if for every property P, x has P if and only if y has P, then x = y. In Black’s world the spheres are (numerically) distinct because they are spatially separated from each other, and yet indiscernible because they are similar under any respect. Thus, Black’s argument holds if spatial separation entails distinctness, but not discernibility.1 This may seem rather counterintuitive at first glance, but Black shows that our intuitions here are not particularly firm. Black seems to support the idea that spatial separation entails distinctness by appealing to the rather intuitive principle governing multiple location: (PNML) No individual material body can be in two distinct places at the same time. What this principle says2 is that for every x and y, if x = y then x has a null distance from y, and thus—by contraposition (PNML') Distance (x,y)  0 Š x  y

102

Chapter Seven

which allows Black to derive distinctness from spatial separation. Black uses this principle as a premise, but it denies the stronger principle to the effect that spatial separation entails discernibility. (DD) Distance (x,y)  0 Š Discernible (x, y) He defends his denial of DD by arguing—in a verificationist mood— against the substantival view of space, and by appealing to the hypothesis of a perfectly symmetric universe.3 (This is fortunate, because Black’s opponent in the dialogue is clearly on the verificationist side: he is actually arguing in favour of PII in order to vindicate certain verificationist intuitions). It is easy to see why a substantival view of space would make trouble for Black's account. According to substantivalism, places are discernible in virtue of nothing else than themselves, their individuality being their order and position (Nerlich 2003). Thus, if p and p' are nonoverlapping places, then p and p' are discernible even if the entities a and b that occupy p and p' (respectively) are otherwise perfectly similar. It follows that two perfectly similar entities occupying two discernible places are in turn discernible (and not just distinct) from each other. It is easy to see why: ex hypothesis, sphere a occupies a distinct place from sphere b, the two places x and y are discernible entities according to substantivism; hence the two spheres can be discerned by the relation of occupation that they bear to two discernible entities. More generally, the argument relies on the following principle: if a bears relation R to x, b bears R to y, but it is not the case that a bears R to y and b bears R to x, then if x is discernible from y, it is also the case that a is discernible from b. However, if substantivalism is false, spatial relations are not relations between places, but rather between material objects like the two spheres in Black's thought experiment. And given the perfect similarity of the spheres and the perfect symmetry of their universe, there is no spatial relation in virtue of which we can discern them. If we assume a relationalist view of space, then spatial separation does not (obviously) entail discernibility, and PNML (or, more precisely, PNML') can be used against PII—at least in a perfectly symmetric universe. There is a trade-off between intuitions here: given that the intuitive appeal of the relational view of space is roughly on a par with that of the substantival view, Black’s conclusion is not so far-fetched in terms of intuitions. There are two strategies to defend PII against Black’s attack: either we argue that Black fails to show that spatial separation implies distinctness (in other words, in the universe that he describes there is only one sphere), or we concede distinctness but maintain that spatial separation entails

The Identity of Indiscernibles and the Principle of No Co-location

103

discernibility too (in other words, in the universe that he describes there are two spheres, but they are discernible).4 The first strategy has been pursued along two directions: either by arguing that Black has not specified a situation in which a sphere stands at a certain distance from another sphere, and his description of a two-sphere world is indeed a description of a one-sphere world (Hacking 1975 and Adams 1977), or by arguing that in the specified situation a single sphere stands at a distance from itself.5 In either case PNML' is rejected, albeit for different reasons. According to the first construal of the first strategy, PNML' fails as a necessary principle, because it has to be restricted to Euclidean spaces. In a closed curved space, being at a non-null distance from oneself is a welldefined notion, and what in a flat space is described as a two-sphere universe may turn out to be, in a non-flat space, a universe containing only one sphere. According to the second construal of the first strategy, PNML' is false even in a Euclidean space, since if individuals are bundles of universals, they can—as each of the universals forming the bundle—exist in different places at the same time. Since PNML' seems highly intuitive, the first strategy comes with a high counterintuitive cost. What about the second strategy? It can be pursued by denying Black’s relational view of space and its verificationist underpinnings. As we have seen, if substantivalism is true, then PII is shielded from Black’s counterexample, because the two spheres are discernible in virtue of the discernibility of the regions in which they are located. And given that we do not seem heavily biased towards either theory of space, such a position is no more costly in terms of intuitions than Black’s.

2. The argument from superposition Another attack on PII comes from quantum mechanics (QM). Allegedly, QM provides us with situations in which we have reason to think that two superposed particles exist. The reason why we should take them to be distinct is not their spatial separation, since there is none, but the fact that they have inverse spin with respect to each other (French 1989, 2006). However, this difference does not entail that the two spheres are discernible, since, considered separately, each spin’s direction is indiscernible from the other. Each particle, thus, has a spin opposite to the spin of some other perfectly similar particle, just as in Black’s case each sphere has the property of being separated from a perfectly similar sphere. Thus, each particle is numerically distinct from the other but also indiscernible from the other, since the only spin property that both possess is having a spin

104

Chapter Seven

opposite to the spin of some other perfectly similar particle. Note that in the superposed case we are not giving up PNML', but rather its converse, the principle of no co-location: (PNC) No two material bodies of the same kind can occupy the same place at the same time. If PNC is true then, for any x and y, if x  y, then there is a spatial separation between x and y, and by contraposition (PNC') Distance(x,y) = 0 Š x = y Within QM, giving up the intuitions behind PNC' seems theoretically rewarding. We can deny that x being at a null distance from y implies x being identical to y, since we have independent reasons (their relative difference in spin) to take x to be distinct from y (although not indiscernible from y). Now, what happens if we “turn” the QM case into a Black-type argument involving macroscopic bodies? It turns out that we obtain a version of the original argument that appears to be immune to any counterexample based on the assumption that space is substantival, does not require any assumption about the symmetry of the universe, and does not lead us to the denial that spatial separation does not entail discernibility. Consider a universe in which two perfectly similar iron spheres are perfectly superposed. If this were the case, then PII would fail. But that scenario is possible only if PNC' is false, and regardless of whether we reject the relational view of space or PNML' holds. Besides, given that the two superposed spheres share all qualitative properties by assumption and bear the same spatial relations to any other distinct spatial point, they are not discernible, even if the universe that they inhabit is not symmetric. Thus, the price to pay here is to abandon PNC', which is a highly intuitive principle, but it was not among Black's assumptions. Indeed, PNC' does not play any role in Black’s original argument, the main support for Black's point that the spheres are distinct being PNML'—which can be maintained in the version of the argument with superposed spheres too. We want to focus on a puzzle about the intuitions behind the two arguments—Black’s original one, and the one assuming the possibility of two superposed iron spheres. PNML' figures as a premise in Black’s argument. It is an intuitive principle, but also the stronger principle DD that says that spatial separation entails discernibility has a strong intuitive appeal. Black’s argument has to deny DD. The revised version does not

The Identity of Indiscernibles and the Principle of No Co-location

105

rely on PNML', but it does not entail its negation either. Rather, it assumes the negation of PNC' as a premise. Assuming that DD and PNC' are equally intuitive, we should expect a stalemate of intuitions here. However, the revised version looks far less intuitive than Black’s original argument. The puzzle is that it looks difficult to explain why it is so. Note also that Black’s argument is less conservative than the revised version, since it requires us to make two further assumptions (relativism about space, and the possibility of a symmetric universe). It may be objected that as the physical story goes, two macroscopic coincident spheres are not acceptable—and that is enough to explain the limited intuitive appeal of the revised argument (Della Rocca 2005). But we observe that we ourselves have no problem imagining two superposed spheres, just as we have no problem imagining the spheres in Black's original example, no matter how contrived it may appear from a physical point of view. As we noted, denying the entailment from spatial separatedness to discernibility is as costly as denying PNC' in terms of intuitions. Critics of the possibility of two superposed spheres should indeed be aware of the risk that their argument against the superposition of macroscopic objects assumes PII. If we consider the property of being in a certain place (at a certain time) as a property that only one material body can instantiate, then each material body will have at least one property that no other material body can have (O’Connor 1954). If a material body x has all the properties that a material body y has, then x will have the property of being at the same place and time y is at. Since this is a property that only one material body can have, x turns out to be identical to y. So, does PNC' imply PII? No. If Black is right, at least within the relationalist view of space, we are entitled to derive only distinctness and not discernibility from being in different places—and thus PNC' may be true, while PII fails. In other terms, it may be true that for every place p and time t, only one material body instantiates being in p at t, while two distinct material bodies occupying two distinct places share all their properties. Of course, if PII holds, then one can argue that we have good reason to consider being in a certain place at a certain time as a uniquely instantiated property. At least, given that sharing all the properties boils down to identity, there cannot be two perfectly similar bodies sharing their location. But then it is not clear whether there are reasons for PNC' that are independent from PII, apart from its being intuitively true in our ordinary experience.

106

Chapter Seven

3. Impure properties and bare places Some might object that the puzzle about intuitiveness that we have raised has an easy solution. In discussing his example, Black also appeals to the principle that in formulating PII we have to rule out “impure” relational properties, i.e. properties that refer to individuals, or extensionally defined sets of individuals (Adams 1977; see also Cross 1995). If we allowed for such properties, PII would immediately follow. Imagine a Black-world in which sphere a is spatially separated from sphere b. Sphere a exemplifies the impure property of being identical to a, and sphere b exemplifies the (different) impure property of being identical to b; hence, a is discernible from b. But this “defence” of PII would trivialize the principle and beg the question against Black’s point. If we do not assume that a is discernible from b, then the two properties would not count as discernible either. However, even if we allow only for impure properties other than these peculiar ones, Black’s counterexample would fail. Given the spatial distance between the two spheres, we could discern them by appealing, for instance, to the property of being one hundred miles from a, which b, but not a, exemplifies. The rationale against the introduction of these properties as relevant for discernibility is that if we do not assume a and b to be discernible, then the property being one hundred miles from a would not be discernible from the property being one hundred miles from b. Thus, Black seems to appeal to verificationist intuitions against the introduction of impure properties of this kind, intuitions that are very similar to the ones supporting relationism about space. And it may seem that we can only resort to such a verificationist principle in order to rule out the substantivalist’s objection: if we do not allow for impure properties, then relations to places would not be allowed either (as means of discernibility). Indeed, from the point of view of the overall dialectic, defending PII by appealing to “bare places” as what allows us to discern the two spheres may turn out to be slightly problematic. If we put the issue of PII in terms of whether there are “thisnesses”, which can be individuated independently from their “suchnesses” (as per Adams 1977), a bare place indeed looks like one of these thisnesses. To be sure, if adherence to PII is motivated by verificationist worries about the metaphysics of material bodies and places—as in the case of the “original” opponent of Black—this line of defence is almost self-defeating. But this is too hasty. Places are a different sort of entity (if they are entities at all) than the material objects that occupy them. In particular, the distinction between distinctness (their “thisness”) and discernibility (their “suchness”) does not apply to them, as it does to material objects. The only

The Identity of Indiscernibles and the Principle of No Co-location

107

“individuation principle” for places substantivally construed is their position. Thus, even if we banish more general impure properties (such as being spatially separated from sphere a), a substantival view of space would support Black’s opponent. Indeed the refutation of Black and a defence of PII would be very convenient: being at place p is a different property than being at place p' if both p and p' are spatially separated and the substantivalist is right. And given the sui generis status of places as individual entities, it is not at all clear that banishing impure properties would block being at a place p as discernible from being at a place p'. Therefore, in Black’s argument both the relational view of space and the refutation of impure properties are required in order for the argument to hold. In the superposed spheres case, on the other hand, we can drop both restrictions: the substantivalist cannot resort to two different properties being at place p and being at place p' to discern the two spheres, since they share the same location. Again, the fact that the revised version is less intuitive than Black’s original one appears to be unjustified by our ordinary means of assessment of intuitions.

4. A decision tree for intuitions To sum up, we present a quick decision tree. Is space relational? YES: Does PNML hold? YES: a and b are not discernible; but they are distinct, hence PII fails. NO: a and b are not distinct; PII is safe. NO: Does PNC hold? YES: a and b are discernible; PII is safe. NO: a and b are not discernible (when superposed); PII fails. Note that the requirement of a perfectly symmetric universe and the confinement to non-impure properties can be dropped in the last branch of the decision tree; congruent superposition guarantees indiscernibility. Our line of reasoning indicated that the viability both of Black's criticism of PII and of a defence of PII rests upon the acceptance or the rejection of some contrasting principles about co-location and multiple location, about the nature of space, and of assumptions about symmetries in the universe. However, both versions of the thought experiment are based on the rejection of one or other intuitive principle (DD and PNC' respectively). The revised version is based on fewer assumptions than the

108

Chapter Seven

original version and yet is less intuitive. This is an interesting discovery, we maintain, since it shows that a good deal of counter-intuitiveness, once “unpacked”, may turn out to be utterly ungrounded. The situation is somehow the reverse of a familiar one in logic, in which apparently obvious theorems sometimes require complicated demonstrations, based on non-obvious axioms. In conclusion, the rejection of PNC' has the advantage of freeing the argument from the requirement of a relationalist view of space and the hypothesis of a symmetric universe, but it is arguably much more counterintuitive than its acceptance. The counterintuitiveness of perfect superposition by bodies of the same kind is in the end the only serious threat to Black's expanded argument. As we observed, one should be careful not to ground PNC' on considerations that may depend on PII. PNC' is thus best understood as a cognitive bias towards spatial separation. As such, its evidential weight is simply the same evidential weight we would put on any of our ungrounded intuitions.6

References Adams, R.M. 1977, “Primitive Thisness and Primitive Identity”, The Journal of Philosophy 76 (1), 5 Black, M. 1952, “The Identity of the Indiscernibles”, Mind 61 (242), 153 Cross, C. 1995, “Max Black on the Identity of Indiscernibles”, The Philosophical Quarterly 45 (180), 350 Delle Rocca, M. 2005, “Two Spheres, Twenty Spheres, and the Identity of Indiscernibles”, Pacific Philosophical Quarterly 86, 480 French, S. 1989, “Why the principle of the identity of indiscernibles is not contingently true either”, Synthese 78, 141 —. 2006, “Identity and Individuality in Quantum Theory”, E. Zalta (eds.) Stanford Encyclopedia of Philosophy (http://plato.stanford.edu/archives/spr2006/entries/qt-idind/) Hacking, I. 1975, “The Identity of Indiscernibles”, The Journal of Philosophy 72 (9), 249 Hawley, K. 2009, “Identity and Indiscernibility”, Mind 118, 101 Nerlich, G. 2003, “Space-time substantivalism” in Loux, M.J. and D. Zimmerman (eds.) The Oxford Handbook of Metaphysics, Oxford: Oxford University Press, 281 O’Connor, D.J. 1954, “The Identity of Indiscernibles”, Analysis 14 (5), 103 O’Leary-Hawthorne, J. 1995, “The Bundle Theory of Substance and the Identity of Indiscernibles”, Analysis 55 (3), 191

The Identity of Indiscernibles and the Principle of No Co-location

109

Rodriguez-Pereira, G. 2004, “The Bundle Theory is compatible with distinct but indiscernible particulars”, Analysis 64 (1), 72

Notes 1

Or, at least, that is what Black’s “B” character in the dialogue assumes in response to A’s objections. See Black 1952. 2 We frame PNML' in terms of distance in order to avoid begging the question of how to individuate distinct places; but nothing depends on this particular reading. Della Rocca 2005 argues that Black is appealing to brute facts about distinctness (either of material objects or places); we take brute facts to be about spatial relations (either between material objects or places). 3 More precisely, Black’s “B” concedes that there are places, but he maintains that “x is in a different place than y” is analytically equivalent to “x differs from y”. 4 A third strategy, proposed by Hawley 2009, is to maintain that only the sum of the two spheres (or rather, the entity that is the sum of what its parts would be were Black right) exists. We will not consider this proposal in what follows. 5 If individual material bodies such as Black’s sphere are bundles of universals, they can be at a spatial distance from themselves because they can be multilocated. See O’Leary-Hawthorne 1995; but see also Rodriguez-Pereyra 2004. 6 For financial supports, Giuliano Torrengo thanks the projects FFI2011-29560C02-01, FFI2011-25626, and CSD2009-00056 of the Spanish Ministerio de Ciencia e Innovacion (MICINN).

CHAPTER EIGHT NEW REALISM, DOCUMENTALITY AND THE EMERGENCE OF NORMATIVITY1 MAURIZIO FERRARIS

1. New Realism The first move of what I proposed to call “New Realism”2 is conceptual clarification, which focuses on stressing the difference between ontology (what there is, which is independent of our representations) and epistemology (what we think we know, which may be dependent on our representations—but what makes our statements true are not our representations, but that to which those representations relate). Now, it makes perfect sense to assume that there is a conceptual action going on when I recognize a constellation,3 or when, looking at three objects, I believe—like LeĞniewski—that for every two objects there is one which is their sum, multiplying the number of objects.4 But this conflict can be explained by the simple consideration that we cannot see properly neither the constellations nor LeĞniewski’s objects, but only the stars and the three objects recognized by common sense. This is not to argue that constellations are not real, but rather to draw a difference (which obviously stems from the difference between ontology and epistemology) between two layers of reality that fade into each other. The first is what I would call İ-reality, meaning by this “epistemological reality”, or what the Germans call “Realität”. It is the reality linked to what we think we know about what there is (which is why I call it “epistemological”). This is the reality referred to by Kant when he says that “intuitions without concepts are blind”; or by Quine when he says that “to be is to be the value of a variable”. But next to, or rather below, the İreality I also set the Ȧ-reality in the sense of ੕ȞIJȦȢ (I use the omega just to make a distinction): the ontological reality, or what the Germans call

New Realism, Documentality and the Emergence of Normativity

111

“Wirklichkeit”, which refers to what there is whether we know it or not, and which manifests itself both as a resistance and as positivity. The second move made by New Realism, after that of conceptual clarification, is an empirical observation. There is a class of representations that I think will never be able to accompany: that of the infinite number of things that existed before any I think. I call this argument “pre-existence”:5 the world is given prior to any cogito. Then there are classes of representations that, even though accompanied by the I think, seem to resist it, without regard of the so-called “representational dependence”; I call this argument “resistance”:6 reality may resist our conceptual schemes. And then it often happens that the I think successfully interacts with beings presumably devoid of any I think, for example with animals; I call this argument “interaction”:7 beings with different conceptual schemes can interact in the same world. I collect these empirical circumstances— which, however, have a transcendental role, since they define, even though in retrospect, our possibilities of knowledge—under the name of “unamendability”:8 the key feature of what there is is its prevalence over epistemology, because it cannot be corrected—and this is, after all, an infinitely more powerful necessity than any logical necessity. On the topic of unamendability two clarifications are necessary. The first is that reality cannot be corrected in the sense that things continue to exist regardless of what we think of them (while we can affect their existence through action, both in the social world and in the physical). Second, the notions of pre-existence, resistance, interaction and unamendability refer to a basic ontological factor: that is, causality. Things in the world act causally (and thus in a pre-existing, resistant, interactive and unamendable way) upon us, and we are things in the world as well. Hence the third move of New Realism: affordance. The real does not only manifest itself as resistance and negativity: every negation entails a determination and a possibility. The world exerts an affordance,9 through the objects and the environment, that qualifies as a positive realism.10 The thesis I defend through the argument of affordance11 is that we should start from the objects (an area in which, as I said, subjects are also included), so as to reduce the gap between our theories and our experience of the world. This is not meant to be a futile worship of objectivity (which is a property of knowledge, not of being), but a due recognition of the positivity on which we all rely, but upon which we rarely reflect. This does not only apply to physical experiences: the way in which beauty, or moral value or non-value come forward is clearly the manifestation of something that comes from outside of us, surprising and striking us. And it has value precisely because it comes from outside:

112

Chapter Eight

otherwise it would be nothing but imagination. That is why, contrary to what is often said, one cannot distinguish the value from the fact: trivially, this is because the fact is itself a value, and the highest one, i.e. positivity, 12 which in turn is the condition for the possibility of each value. In this regard it is necessary to clarify a statement that might seem too trenchant. Facts have the characteristic of being independent of us and of affecting us (i.e., of acting causally on us), and this, in fact, is their fundamental value. However, “value” is usually a standard against which to assess things and facts. Beauty is a value in the sense that it allows us to evaluate something as more/less beautiful, and similarly justice allows us to evaluate something as more/less just. Instead, facts are not “more/less real” but simply real, and they do not allow us to evaluate anything other than themselves. Another way to express this difference is to say that facts exist exerting causality while values exist exerting normativity. However, as will be clear in the part of this essay dedicated to emergence, normativity derives from the sphere of facts, so causality is the necessary (though not sufficient) condition of normativity. At this point—and it is the fourth move of New Realism—it becomes possible to articulate the characteristics of the affordance that comes to us from the objects. We need to begin by introducing, next to the categories of natural objects (which exist in space and time independently of the subjects) and ideal objects (which exist outside of space and time, independently of the subject), two new categories: that of artefacts, which exist in space and time depending on the subjects for their genesis, and that of social objects, which exist in space and time depending on the subjects for their genesis and their persistence.13 From this point of view, it is entirely legitimate to assert that the stock market or democracy are representationally dependent (I will soon try to clarify this term since, as we have seen, it is rather obscure) on our collective beliefs. But this does not mean in any way that dinosaurs have some degree of dependence with respect to our collective beliefs. If anything, dependence concerns professorships in paleontology. The fifth and final move of New Realism consists in isolating the concept of environment: everything, including corporations, symbolist poems and categorical imperatives, has its origin in the affordance offered by the environment. A cave offers affordances for different types of beings and serves as a shelter because it has certain characteristics and not others. Ecosystems, state organizations, interpersonal relationships: in each of these structures (infinitely more complex than a cave) we find the same structure of resistance and affordance, which from causality may (though not necessarily) evolve into normativity. I define “environment” every

New Realism, Documentality and the Emergence of Normativity

113

sphere in which these interactions take place, from an ecological niche to the social world—of course, each with its own characteristics. The perspective suggested by “To exist is to resist in an environment” (which can be articulated in positive as: “To exist is to produce effects— causal or normative—in an environment”) is that of a structurally opaque existence that manifests itself first of all in its persistence and possibly in its acting in an environment, without further qualifications. In other words, the field of sense is in the environment and not in the head; it is in the affordance and not in the concepts. Obviously, starting from the objects and from the opacity of existence involves being aware that there can never be a full totality, and rather that our relationship with the world lies on a confusing balance between ontology and epistemology. This, however, does not mean that the positivity of objects is precluded to us. Indeed, it is this very positivity that allows us to dwell in the world despite the fact that our notions are rarely clear and distinct.

2. Documentality I call “Documentality”14 the environment in which social objects are generated. In the definition of documentality the concepts of craft, writing and recording are central. Craft is, strictly speaking, any possibility of recording, which heralds the possibility of iteration: that is, the most manifest form in which technology comes into our experience. Now, the craft of all crafts, in our historical experience, is precisely writing. And contrary to what was posited by Plato, external writing has three incalculable advantages compared to inner writing, which takes place on the soul. First, public accessibility. No one can look into the minds of others, but to read the texts of others is more than possible; contracts, money, encyclopaedias: all of the social world and the world of knowledge require this resource. Second, while internal writing is destined to disappear with us, external writing can survive even without us. Third, the ability to produce more copies of the same entity (a form of repetition that it would perhaps be more appropriate to define “instantiation”15). Especially, note this: if we look at its essence, any form of recording is a kind of writing. A video or voice message that you can play as you like (which today is technically very easy) are forms of writing, just as a computer file or a piece of paper. If we consider that recording, as permanence of the memory trace, is also the condition of possibility of thought—as suggested by the old metaphor of the mind as a tabula, as a support for registration—it is not difficult to recognize the centrality of this category, which was instead systematically neglected in favour of

114

Chapter Eight

others (mind, action ...) in the analysis of the construction of social and mental reality.16 From this point of view, it is possible to assert that society is not based on communication, but on recording. Since there is nothing social outside the text, papers, archives and documents constitute the foundation of the social world. I repeat: society is not based on communication, but on recording, which is the condition for the creation of social objects. Man socializes through recording. Bare life is but a remote start: culture begins very early—manifesting itself through recordings and imitation (language, behaviour, rituals). This explains why writing and the sphere of recordings that precedes and surrounds writing are so important. Equipped with these tools, we now have a system that explains both the media and social reality, even in its bureaucratic dimension. The constitutive law of social objects is Object = Inscribed Act. That is to say that a social object is the result of a social act (such as to involve at least two people, or a delegated machine and a person) that is characterized by being recorded, on a piece of paper, on a computer file, or even only in the minds of the people involved in the act. Social objects are divided into documents in a strong sense, as inscriptions of acts, and documents in a weak sense, as recordings of facts. Once recorded, the social object, dependent on minds as to its genesis, becomes independent as to its existence—the same thing happens in the case of artefacts, with the only important difference that an artefact can offer its affordance even in the absence of minds (a table can be a shelter for an animal), while a document cannot. It can certainly be argued that documental recording is not sufficient without a normative social practice behind it. Banknotes become waste paper if there is no normative institutional background to support them. In fact, if mere recording were enough for the existence of social objects, then the old European currencies would continue to be valid even when they are actually worth nothing because in the meantime their regulatory environment has changed. But if—as I articulate in the section dedicated to the emergence—documentality is the source of intentionality and normativity, then I think that this objection loses its validity, as evidenced by the fact that a collective amnesia (i.e., the disappearance of documentality) would coincide with the disappearance of intentionality and normativity. In other words, the institutional background that supports documentality is in turn a documental background, and so on ad infinitum. Documentality has an immediate heuristic advantage. A theory of mind-dependence will always have intrinsically obscure aspects because it does not entail a simple causal dependence. For social objects to exist, it is

New Realism, Documentality and the Emergence of Normativity

115

necessary that there are at least two minds and normally, in complex phenomena, there are many more. In such complex social cases, usually many minds do not think in any way about the object and yet they interfere with it, while many others do think about it and yet are unable to successfully interfere with it (think of a financial crisis, or a war). Apparently, we are dealing with a puzzle: social objects, as we have seen, are dependent on the mind, but they are independent of knowledge (i.e., even of consciousness). A marriage that nobody knows anything about did still take place; in the same way, there may be a recession even though no one suspects it. How is this possible? Does this not mean to argue that social objects are both dependent on and independent of the mind? No, it does not. The contradiction would present itself only if “mind dependence” were understood as dependence on one mind, as if anyone could determine the course of the social world. But this assumption is contradicted by any experience of the social world (my mind does not make the laws, nor the prices of goods, at most it can write this article), as well as by the fact that in many circumstances our own mind seems to be independent of itself, such as when we develop obsessive thoughts that we would rather not have. If we no longer have a contradiction between “dependence on the mind” and “independence from knowledge”, we still have to explain how social objects can persist even when we do not have consciousness or knowledge of them. That is why I argue that the foundation of the social environment is documentality. In fact, when dealing with social objects we do not have to do with a series of intentionalities that consciously keep the object alive, so to speak, as if we all thought at the same time about the Constitution. It is not so: the Constitution is written, and at this point it is valid even if no one thinks about it (which in fact happens all too often). In addition to solving the puzzle of mind-dependence and independence from consciousness, documentality also allows us to provide a more solid basis for the constitutive rule proposed by the most influential theorist of social objects, John Searle: namely the rule “X counts as Y in C” (the physical object X counts as the social object Y in the context C). The limit of such proposal is twofold. On the one hand, it does not seem able to account for complex social objects (such as businesses) or negative entities (such as debts, in which case it seems difficult to find a corresponding physical object). On the other hand, it makes the entire social reality depend on the action of a completely mysterious entity (as opposed to documents), that is, collective intentionality, which allegedly manages the transformation of the physical into the social.

116

Chapter Eight

According to the version that I propose, on the contrary, it is very easy to account for the totality of social objects, from informal promises to businesses and even negative entities such as debts. In all these cases there is a minimal structure, which is guaranteed by the presence of at least two people who commit an act (which may consist of a gesture, a word, or writing) that can be recorded on some kind of writing surface, even if it were only the tabula of human memory. In addition to accounting for the physical basis of the social object—which is not an X available for the action of collective intentionality, but a recording that can take place in multiple ways—the rule that I propose (and which I call the “rule of documentality” as opposed to the “rule of intentionality”) has the advantage of not making social reality depend on a function, i.e. collective intentionality. In fact, such function is dangerously close to a purely mental process: this has led Searle to make a statement that is anything but realistic, namely that the economic crisis is largely the result of imagination.17 From my perspective, on the contrary, since this is a form of documentality, money is anything but imaginary, and this circumstance allows us to draw a distinction between the social (what records the acts of at least two people, even if the recording takes place in the minds of those people and not on external documents) and the mental (which can take place only in the mind of a single person). Of course, it could be noted that the Constitution applies even if nobody thinks about it, but it ceases to have any value if there is no one who is able to read it or is simply willing to follow its dictates. The validity of what is written in a document remains dependent on a system of practices that makes the content of the document normative. Otherwise, it would be impossible to distinguish between a current Constitution and a Constitution no longer in act, since both are written down somewhere. The theory of documentality, if taken to an extreme, could risk not being able to explain this fundamental distinction. At this point, however, it is worth pointing out that the social cannot be merely “that which records the acts of at least two people” but it must also include the practice that supports this recording making it valid and normative. That is precisely what I now propose to illustrate through the thesis of the emergence of intentionality and normativity from documentality. Incidentally, this is intuitively known by every advertising investor, who hopes that the iteration of a message can generate desires and lifestyles (i.e., intentionality and normativity) related to the advertised product.

New Realism, Documentality and the Emergence of Normativity

117

3. Emergence It is in what I call environment that the emergence of thought from being occurs; such a process can be regarded as the development of an (intelligent) epistemology on the basis of an unintelligent ontology, a competence that precedes comprehension. The fact that the meaning is not in the head,18 but in the world, is a principle well-illustrated in the social world and its media, which I regard as equivalent and locate in the sphere of documentality. In this regard I would like to suggest two reflections. The first one concerns the externalism of meaning, which seems to be realized in the Web. The original idea of the semantic web was that programmers should classify what they put on the network—a clearly unfeasible project. The conception of the environment as a sphere in which emergence takes place proposes a Copernican revolution whose basis is the thesis, brought forward by Wittgenstein, that “meaning is use”. Concretely, let us take the uninterrupted stream of texts (20 billion documents) that is the Web, see the exchanges between people and, starting from use, derive the meaning. The second one concerns the equivalence between social and media. Of course it could be argued that between social and media there is at least a quantitative difference. The media seem to ensure a much larger degree of iterability, replicability and instantiability than the social, in the strict sense. The media might be seen as a powerful enhancement of the documental structure that constitutes the social world. The media introduce iteration methods that are more and more refined (notational rather than purely mechanical) and increasingly pervasive (with phonography, photography and cinema one can not only iterate conceptual contents, but also perceptual ones). What this objection does not seem to take into account is that, with the new media, the difference between social and media completely disappears, since the same technical devices are delegated to the production of both social objects and media content. With the same tablet you can get a plane ticket or a certificate, watch a sport event, or produce social and media content yourself through a social network. The Web is a recording system that generates a superorganism that evolves autonomously, just like a termite mound, structuring complex articulations in the total absence of a central system19. Now, in the light of the Darwinian theory of the mind proposed by Daniel Dennett, we might ask ourselves: is it not exactly the same thing for the human brain? Just like ants (or like computer memories and the whole Web), single neurons

118

Chapter Eight

do not “think”, but “download”. Yet the whole of them constitutes consciousness and thought. I do not mean at all to consider the Web as some sort of macroconsciousness or macro-brain, with yet another restatement of the thesis— which in retrospect proved to be fallacious—that the Web is a “collective intelligence”.20 I simply assert that the Web, as a macro-archive and macrocommunity, presents the same mechanism that takes place in superorganisms or in intelligence (natural or artificial), so that organization precedes and produces understanding. From this point of view, one could even speak of a computer evolutionism, which depends on computers much more than it depends on designers, revealing the real needs of society: a calculation tool turned into an archive tool, an isolated machine became a machine connected to the web. Something similar also happened with the cell phone, which was thought of as a tool to talk but tuned into a writing device, with the ultimate convergence between phone and computer. It is in the environment of documentality that the social “we” takes place, through the genesis of what I have proposed we call “documental community”.21 Again, we are dealing with a difference with respect to collective intentionality, regarded as a sort of natural primitive instinct that makes us say “we” instead of “I” in a number of situations, and that is allegedly the basis of the construction of the social world. I have more than one doubt with regard to this, because in fact the “we” is only reached through training. It is true that a group of people on a trip can say “we are walking”, but is it still “collective intentionality” when those who are walking are a group of prisoners held at gunpoint? In the perspective of documentality, it is through the sharing of documents and traditions that a “we” is constituted. It is precisely for this reason that society has adopted writing and archives so early: in order to ensure that the spirit can manifest itself and become recognizable, gaining visibility and permanence in time. From this point of view, the most transparent form of the “we” is a document bearing signatures and exhibiting with honesty the terms, boundaries and objectives of the “we”—which, in this version, appears as the conscious agreement between a defined number of people for a specific purpose. Of course one could argue that this kind of sharing requires that the “we” be already constituted. But here I would like to point out that the process I described does not constitute a rigid and unidirectional determination so that documentality leads to the “we”, but rather a virtuous circle for which collective interaction (sharing, made possible by iterations, recordings, imitations and education, i.e. various forms of documentality) supports the production of documents that enhances

New Realism, Documentality and the Emergence of Normativity

119

collective interaction itself, which in turn enhances the production of documents. If postmodernists claimed that fiction cancels reality, with a constructionist hyperbole, I say that intentionality derives from documentality. The prospect of documentality begins with the theory that—from its ancient to its modern supporters—conceives of the mind as a tabula on which to lay inscriptions. In fact, as we have seen, there is a powerful action of inscriptions in social reality: social behaviours are determined by laws, rituals and norms; social structures and education form our intentions. Imagine an Archi-Robinson Crusoe as the first and last man on the face of the earth. Could he really be devoured by the ambition to become an admiral, a billionaire or a court poet? Certainly not, just as he could not sensibly aspire to follow trends, or to collect baseball cards or still lives. And if, say, he tried to fabricate a document, he would be undertaking an impossible task, because to produce a document there must be at least two people, the writer and the reader. In fact, our Archi-Robinson would not even have a language, and one could hardly say that he would “think” in the usual sense of the term.22 And it would seem difficult to argue that he was proud, arrogant or in love, for roughly the same reason why it would be absurd to pretend he had friends or enemies. We thus have two circumstances that reveal the social structure of the mind. On the one hand, the mind cannot arise unless it is immersed in the social, made up of education, language, communication and recording of behaviours. On the other hand, there is the huge category of social objects. Rather than sketching a world at the subject’s total disposal, the sphere of social objects reveals the inconsistency of solipsism: the fact that in the world there are also others in addition to us is proven by the existence of these objects, which would not have a raison d'etre in a world where there was only one subject. There is no thought without social normativity, nor is there normativity without the social.23 This brings me to my second argument: namely that normativity derives from documentality. 24 To put things in a more all-encompassing way, as suggested by the German mediologist and philosopher Friedrich Kittler,25 the media are not so much an extension of man (according to the optimistic view proposed by McLuhan), but rather man is the result of the media. Such thesis is all the more true now that, as I am suggesting, any difference in principle between the media and documentality has ceased to be. Everything that becomes powerful can follow its own logic;26 there is nothing surprising about this: the general thesis is that any system of emancipation is at the same time a control system. Machines emancipate people from physical fatigue but deliver them to industrial work. Internet appeared, at its outset,

120

Chapter Eight

as the liberation from work and as a countervailing power; in reality, as was perfectly imaginable, it introduced a new work and a new power. This takes nothing away from the merits of the Internet, just as Taylorism takes nothing away from the merits of the machines, but it is an element that cannot be underestimated. This is the dark side of the Web, that we need to make explicit. Surprisingly, the greatest thinkers of the power of the Internet are, in my opinion, two figures that have never known it, and perhaps not even suspected its future existence: Schmitt, who emphasized that the essence of power lies in bureaucracy, and Jünger, who theorized the total mobilization and militarization as the essence of the modern world. The most important aspect, in my opinion, is the way in which documentality becomes a source of normativity. If it were not possible to keep traces, there would be no mind—and it is not by chance that, as I have noted, the mind was traditionally depicted as a tabula rasa, a support on which impressions and thoughts are inscribed. But without the possibility of inscription there would not even be social objects, which consist precisely in the recording of social acts, starting from the fundamental one of the promise. If this is the case, perhaps we should translate Aristotle’s sentence that man is a zoon logon echon as: “man is an animal endowed with inscriptions”, or rather (since one of the meanings of logos in Greek is “promise”, “given word”) as “man is an animal that promises”. “The breeding of an animal that can promise—is not this just the very paradox of a task which nature has set itself in regard to man?” 27 From recording thus derives a mobilization and especially a total responsibilization. The Internet is an empire on which the sun never sets: at any time we can receive a request for work to be done, and at all times we are responsible for it, with a process that extends indefinitely the duration of work and the principle of responsibility (because the request is recorded). Here is what we learn from what, to me, is more than an example, and with which I wish to conclude. Imagine an old amnesic phone, in pre-answering machines and pre-cell phone times. It would ring, we were not at home, we would come back and not know about it—we lived happily and without obligations. Today it is no longer the case. Each “missed call” is recorded on the phone, and this call generates an obligation to respond, it makes the phantom quiver, it raises the pang of remorse in what we call “soul”. The very fact of recording makes us responsible: a promise made between amnesic people would not be a promise, it would be a series of empty words. This is why the world is filled with papers, files, archives. Moral responsibility, at its core, is just that: inscription and recording. It is by no chance that divine omniscience

New Realism, Documentality and the Emergence of Normativity

121

and omnipotence are represented as the holding of a book in which everything is written, nothing is hidden or forgotten.

References Beuchot, M. and J.L. Jerez 2013, Manifiesto del nuevo realismo analógico, Buenos Aires: Circulo Erméneutico Brandom, R. 1994, Making it Explicit: Reasoning, Representing, and Discursive Commitment, Cambridge, Mass: Harvard University Press Dawkins, R. 1976, The Selfish Gene, Oxford: Oxford University Press Ferraris, M. 1997, Estetica razionale, Milano: Raffaello Cortina —. 2001, Il mondo esterno, Milano: Bompiani —. 2002, “What is Like to be a Slipper”, The Dialogue 1, 164 —. 2005, “Diversity of Social Objects. Outlines of a Theory”, in Bussani M. and M. Graziadei (eds.) Human Diversity and the Law, BrusselsBerne-Athens: Stämpfli-Bruyland-Sakkoulas, 135 —. 2006, “Causality and Unamendableness”, Gestalt Theory 28 (4), 401 —. 2007, “Documentality or why nothing social exists beyond the text”, in Kanzian C. and E. Runggaldier (eds.) Cultures. Conflict - Analysis Dialogue (Proceedings of the 29th International Ludwig Wittgenstein Symposium in Kirchberg, Austria), Frankfurt: Ontos, 385 —. 2009, “Documentality, or Europe”, The Monist 92 (2), 286 —. 2010a, “Community”, La Repubblica 14 Luglio 2010 —. 2010b, “Social ontology and documentality”, in Pozzo R. and M. Sgarbi (Hg.) Eine Typologie der Formen der Begriffsgeschichte, Archiv für Begriffsgeschichte: Sonderheft 7, 133 —. 2011a, “Nuovo Realismo”, Rivista di estetica, 48 (3), 69-93 —. 2011b, Anima e iPad, Parma: Guanda (French translation forthcoming for Presses de l’Université de Montréal) —. 2012a, “Esistere è resistere”, in De Caro M. and M. Ferraris (eds.), Bentornata realtà, Torino: Einaudi, 139 —. 2012b, Documentality. Why is necessary to leave traces, New York: Fordham University Press —. 2012c, Manifesto del nuovo realismo, Roma-Bari: Laterza (German transl. Frankfurt a.M.: Klosterman 2014; English transl. New York: SUNY forthcoming) —. 2013, “Reality as Unamendability”, in Cataldi Madonna, L. (ed.) Naturalistische Hermeneutik, Wuerzburg: Koenigshausen & Neumann, 113 —. 2014, “New Realism as Positive Realism”, META. Research in Hermeneutics, Phenomenology and Practical Philosophy

122

Chapter Eight

Fichte, G. 1796, Grundlage des Naturrechts (“Zweiter Lehrsatz”), ch. 1, § 3, Gesamtausgabe der bayerischen Akademie der Wissenschften, Stuttgart-Bad Cannstatt, Frommann-Holzboog, vol. I/3, 342 Gabriel, M. 2013, Warum es die Welt nicht gibt, Berlin: Ullstein Gibson, J.J. 1979, The Ecological Approach to Visual Perception, Boston: Houghton Mifflin Goodman, N. 1978, Ways of Worldmaking, Indianapolis: Hackett Hölldobler, B. and E.O. Wilson 2008, The Superorganism: The Beauty, Elegance, and Strangeness of Insect Societies, New York: W.W. Norton & C. Kelly, K. 2010, What Technology Wants, New York: Viking Press Kittler, F. 1999, Gramophone Film Typewriter, Stanford: Stanford University Press Lévy, P. 1994, L’Intelligence collective. Pour une anthropologie du cyberespace, Paris: La Découverte Lewin, K. 1926, “Untersuchungen zur Handlungs- und AffektPsychologie. I. Vorbemerkung über die psychischen Kräfte und Energien und über die Struktur der Seele”, Psychologische Forschung 7, 294 Meillassoux, Q. 2008, After Finitude: An Essay on the Necessity of Contingency, London: Continuum Nietzsche, F. 2003, The Genealogy of Morals, New York: Courier Dover Publications Putnam, H. 1975, “The Meaning of ‘Meaning’”, in Mind, Language and Reality. Philosophical Papers, vol. 2., Cambridge, Mass.: Cambridge University Press, 215 —. 1987a, “Truth and Convention: On Davidson’s Refutation of Conceptual Relativism”, Dialectica 41, 69 (reprinted in Putnam, H. 1990, Realism with a Human Face, Cambridge, Mass.: Harvard University Press, 96) —. 1987b, The Many Faces of Realism, LaSalle: Open Court Rickert, H. 1915, Der Gegenstand der Erkenntnis, Tübingen: J.B.C. Mohr Scarpa, R. 2013, Il caso Nuovo Realismo. La lingua del dibattito filosofico contemporaneo, Milano-Udine: Mimesis Searle, J. 2010, Making the Social World: The Structure of Human Civilization, New York: Oxford University Press

New Realism, Documentality and the Emergence of Normativity

123

Notes 1

I prepared this article in Bonn with the support of the Käte Hamburger Kolleg “Recht als Kultur”. 2 With a debate started in 2011—which summed up the topics I have been working on for the past twenty years (see Ferraris 1997)—and that is still ongoing, both in Italy and on an international level (for a full overview see nuovorealismo. wordpress.com; for an analysis of the debate see Scarpa 2013). Among the numerous contributions there are several of my own writings on the topic (I particularly refer the reader back to Ferraris 2011a and 2012c) as well as other works such as Gabriel 2013 and Beuchot-Jerez 2013. For a more detailed analysis of some of the moves proposed in the present article, see Ferraris 2012b. 3 Goodman 1978. 4 Putnam 1987b, chaps. 1, 2; and Putnam 1987a. 5 Meillassoux 2008. 6 Ferraris 2012a. 7 Ferraris 2001. The Gedankenexperiment through which I develop the argument of interaction appears in English in Ferraris 2002. 8 Ferraris 2006 and 2013. 9 By using the term “affordance” I am referring to a notion that has been widely popular last century: see Gibson 1979; Lewin 1926. Fichte already spoke of an “Aufforderungskaracter” of the real, see Fichte 1796, chap. 1, § 3. 10 Ferraris 2014. 11 I have extensively dwelt on this in Ferraris 2012b. 12 See Rickert 1915. 13 See Ferraris 2012b and 2005. 14 Apart from the already mentioned Ferraris 2012b, I refer the reader back to Ferraris 2007, 2009 and 2010b. An important development of documentality was provided by Barry Smith with his analysis of “Document Acts”: see http://ontology.buffalo.edu/smith/articles/documentacts.pdf 15 In this sense, writing makes social objects structurally similar to living entities. It is in this sense that, in Ferraris 2012b, I refer to the parallelism offered by Dawkins between genes and “memes” (see Dawkins 1976). 16 See Ferraris 2011b. 17 “It is, for example, a mistake to treat money and other such instruments as if they were natural phenomena like the phenomena studied in physics, chemistry, and biology. The recent economic crisis makes it clear that they are products of massive fantasy” (Searle 2010, 201). 18 Cf. Putnam 1975, 227. 19 Hölldobler and Wilson 2008. 20 Lévy 1994. 21 Ferraris 2010a. 22 In agreement with the argument against private language proposed by Wittgenstein. There must be at least two people not only to produce a document, but also to have a language and, more generally, a rule.

124 23

Chapter Eight

Brandom 1994, cap. I. For a more detailed articulation of these two fundamental theses, see Ferraris 2011b. 25 Kittler 1999. The German original dates back to 1986. 26 As is suggested by Kelly 2010. Nevertheless, it is an old idea that goes from Spengler to Jünger and Heidegger. 27 Nietzsche 2003, 34. On this acknowledgment of the human specific ability to make promises (and, more generally, commitments and entitlements) I refer the reader again to the first chapter of Brandom 1994; Brandon refers in turn to Kant’s idea that humans are essentially “distinctively normative, or rule-governed, creatures” (Brandom 1994, 9). 24

CHAPTER NINE WHY SOCIETY IS NOT MADE OF DOCUMENTS RICHARD DAVIES

The present paper seeks to bring to light the mythical nature of a certain proposal in the theory of social reality that has been crystallising over recent years and that finds systematic formulation in Maurizio Ferraris’ 2009 book Documentalità, translated into English in 2013 under the title Documentality. In line, I hope, with the editors’ choice of “myth” in the present volume’s title, to provide reasons for thinking that the theory of documentality does not give a true account of social ontology while trying to do justice to some ways in which the proposal of it has at least some of the characteristics of the myths we find, for instance, in some much-discussed passages of Plato, such as the story about the formation of the world in the Timaeus. If “true account” corresponds to Plato’s use of “logos”, it seems to me that the theory of documentality enjoys the status of “muthos” in at least the respects of being plausible, attractive and apparently explanatory. Perhaps something like the theory of documentality is also mythical in being at least half-believed or believed half-the-time by many people outside the philosophy room. It is mostly for this last reason that it seems to me to be worth setting out reasons for thinking that the theory in its systematic formulation does not provide a true account (logos) of social reality. The key claim of the documental theory is that every social object is constituted by an inscribed act: Ferraris presents the formula “(Social) Object = Inscribed Act” more than thirty times in his most systematic exposition. By an “inscribed act” he means a public execution of a piece of writing in such a form as to be registered and subsequently consulted by more than one person, and in such a way as to be able to ascertain the configuration of social facts arising out of the performance in question. Against this key claim, it may be well to begin by considering some underlying theses that Ferraris’ theory appears to share with the other major contender in contemporary theory of social reality, namely John

126

Chapter Nine

Searle’s theory that certain physical objects count as social objects in a certain context as a result of collective intentionality within that context. This theory was set out mainly in the first half of his 1995 book The Construction of Social Reality and revised so as to emphasise the notion of collective intentionality in Making the Social World (2010). Among the presuppositions that Ferraris’ and Searle’s theories share there is the thought that social objects stand in need of explanation in terms of something that is not a social object, indeed, that is not an object at all but, interestingly enough, in both cases, an act, whether of writing or of collective intending. Perhaps there are two facets to this shared presupposition that may be distinguished. One is that there are social objects. And the other is that such social objects as there are need to be explained in terms of something that can be identified even if we do not know what social objects there are. The first of these points seems to be almost banal: if there are such objects as—to cite some examples that recur both in Ferraris and in Searle—money, marriages, university degrees and national borders, they are paradigm cases of social objects. Given that we recognise money, marriages, university degrees and national borders, it seems that we recognise some social objects. But it is easy to feel that we are already in the territory of what Ryle had in mind when he was deliberately abusive about what he supposed was the Cartesian “myth of the ghost in the machine”. That is to say, the recognition that “money”, “marriages”, “university degrees” and “national borders” are all nouns does not of itself mean that what they denote are objects. Though the €10 note in my wallet is undoubtedly an object, the money that it is does not seem to me to be an object. Likewise, whether I am married or not at the moment of writing these words does not seem to me to depend on whether a certain object exists or not. That I am the holder of a certain number of degrees (of various degrees) does not derive from or reduce to there being some objects that are my degrees. And, if I have to present certain documents in passing from, say, Ventimiglia to Menton, that is neither because there are objects such as France and Italy, nor because there is an object that is the France-Italy frontier. It may well be that some of the things that a theory of social reality should take account of are objects, and hence, in some sense, social objects; yet it does not seem obvious that the things that are cited as the paradigmatic cases of what needs to be explained by a theory of social reality are really objects at all. To take the instances that we have borrowed from Ferraris and Searle, some first approximations to what is in play might sound rather as follows.

Why Society Is Not Made of Documents

127

My €10 note is a piece of paper that has, where I am now, certain powers, such as that of being recognised as exchangeable for other €10 notes, of being recognised as exchangeable for about three thousand Hungarian Forint, of being recognised as exchangeable for a packet of cigarettes-and-a-€5 note, of being recognised as exchangeable for about half of a paperback copy of Ferraris’ Documentality (but not in the sense of some set of two hundred pages of it). And so on. Though I do not subscribe to any of the more or less fancy “ontologies of powers” that have been coming onto the market of late, my impression is that the attribution to an object of a power, such as redness (the propensity to look red in normal conditions) or weight (the tendency to turn the scales to a certain degree) to a tomato, is less problematic than the conjuring of new objects (chromatic objects? gravitational objects?). Likewise, I recognise that being recognised as exchangeable is a power that has to be multiply relativized, but it is a power of which an account can be given in a theory of economic activity that does not ultimately require any objects over and above pieces of paper (some, but not all, of which are banknotes), commodities (whatever they might be) and economic agents (of probably more than one sort). Again, as regards marriage, the base or explanatory category here seems to be that of rights and duties that have been matured either by explicit promising or swearing, as in a wedding ceremony, or by certain persons’ acting towards each other in certain ways over a certain period of time. Of course, quite how this plays out will depend very much on what the ultimate nature of rights and duties turns out to be. But it would be very odd indeed if rights and duties were such as to be able to bring into existence new objects: even though denominations like “husband” and “wife” do denote objects, the objects they denote are persons with reciprocal rights and duties. It may well be that the unit that is a family may be regarded as a social object, in respect of not being reducible to its members; but I am inclined to think that families can subsist even without marriages. My being a Bachelor of Arts, a Master of Arts and a Doctor Philosophiæ does not depend on the existence of university degrees as social objects. Rather, these qualifications are the result, respectively: of my having written certain things on certain pieces of paper in a certain examination hall, which were regarded with leniency by my teachers; of my having continued in existence for three years from B.A. graduation without infringing certain rules of my university (I have been told— probably unreliably—that if, in the specified period, I had carried a bow and arrow on King’s Parade, Cambridge, I would have been excluded as a

128

Chapter Nine

Magister); and of my having written other things in a stack of paper, which were again regarded with leniency by my examiners. The result of all this scribbling and hanging around was that the university emitted certain pieces of paper that have, again, some powers. These are not quite those of my €10 banknote; for instance, I cannot transfer my qualifications to others in exchange for money or commodities. But I can report my possession of them on my curriculum vitæ; and, if need be, I can flourish them in the face of anyone who cares to take notice of them. If someone, who has not done things at least superficially similar to the sort of writing exercises I have gestured at, nor received the requisite approbation for them, claims to have a given degree, that person is guilty of fraud (and I am rather aware that the “M.A.” that I am entitled to put after my name is regarded as in odour of fraud by graduates of many universities). The fraud lies not in the non-existence of a social object but in the nonperformance of a certain task or due process. Thus, if someone buys a degree certificate—and such pieces of paper are indeed on sale—her attempt to use it as an academic qualification is very likely to fail, because the institutions that engage in this traffic are not universities, whatever they may call themselves. Naturally, there may appear to be some sort of circularity in the legitimacy of certain sorts of accreditation and the activities that take place in certain environments (lecture rooms, examination halls and so on), but it does not seem that that circularity is vicious nor that it would be made any more virtuous by the interposition of a social object. A similar sort of circularity may invest—and perhaps more viciously— the matter of national borders. While the universitiness of some institutions can be established by looking at what the people involved in them do (reading, writing, learning, teaching, examining, researching), the statehood of, say, France seems to amount to nothing more than the stipulation that, within a certain territory, a certain writ runs: France is French because the French say it is (and, at least for the time being, the Germans and others allow them to say so). As we shall see in a little more detail in discussing some specifics of Ferraris’ proposal, a theory of social reality that posits a series of social objects arising out of documents, such as treaties and the like, will have difficulty in saying what makes such documents valid. For the present purpose, however, the thing to notice is that the Frenchness of France does not need an extra object to account for the behaviour both of the French border police and of those who wish to cross the border. It is debatable to what extent and with what rigour a state is permitted to patrol its own borders against anything other than attempted armed invasion by another state (and hence whether or not there

Why Society Is Not Made of Documents

129

should be any function like that of the border police, which is not candidly military). Yet the custom (and I use the word advisedly) has grown up that even those who do have the right to pass from Ventimiglia to Menton have to be able to prove that they have that right every time they come a certain part of the road that leads from the one town to the other by being in a position to produce a passport or an identity card that the border police find convincing. And those to whom such a right has not—whether rightly or wrongly—been recognised by the French state find themselves faced with at least the threat of force against them to prevent their passage beyond a certain part of the road. Unlike my €10 note but like a degree certificate, a passport or an identity card is not transferable; and, like a marriage certificate, it is merely proof of a social fact—for instance EU citizenship—that seems to me to be capable of subsisting even in the absence of the document. For instance, a UK citizen who has not applied for a passport may have no way of proving that status: and, an instance within the instance, Queen Elizabeth II does not have a passport, but seems to be allowed to pass relatively unhindered on her foreign travels. The custom of passport control is in no way explained by the existence of social objects, such as states or borders; these latter are rather functions of an administrative system operating on a certain territory. And a function is not an object, even when it is backed by the threat of force. Even if these are but first approximations, it seems to me that they suggest grounds for hesitancy before accepting the existence of social objects as explanatory of social facts. This hesitancy is perhaps reinforced when we come to consider how the two leading theories of social objects seek to define the objects in question by appeal to acts, in Ferraris’ case to acts of writing and in Searle’s to acts of collective intending, and both are fairly emphatic that these acts make up, constitute or are identical with the objects that are the explananda of the theories (as well as being the explanantia of what money, marriages, degrees and borders are). In one sense, it seems as if these objects have been introduced only to be eliminated by acts: ontology becomes—what shall we call it?— praxiology. But, in another, it may be cause for puzzlement that any act can be identical with any object, any more than a state can be identical with a process or a change with a function. This latter may seem a finicky worry, but it points to a serious problem about what these theories are for. If someone thinks that social ontology is worth doing, then he is likely to think that he is called on—in the going jargon—to quantify over objects to whose existence he is thereby committed, and also that these objects stand in need of identity criteria, which can only be supplied by reference to acts. But this all seems a long

130

Chapter Nine

way round. For it is not at all clear that an ontology—a theory of objects— is going to get us very far in understanding social facts. If the first approximations suggested above are anywhere near right, then we need some economic theory to understand money, an account of interpersonal rights and duties to understand marriage, a description of how academic institutions work to understand degrees, and a genealogy of statehood to understand national borders. What is not at all clear is that these species of enquiry will, sooner or later, call on us to start identifying objects of sorts that are, in one way or another, of a wholly different sort from things like bits of paper and persons. And, even if we did find ourselves called on so to do, I think that, in terms of the Quinean distinction that is often blurred, we would be engaging in ideology. But that is another story. I have suggested that, despite their convergence on the theses that there are social objects and that these objects can be identified by acts of certain sorts, Ferraris’ and Searle’s theories present themselves as rivals for the role of being the true account (logos) of social reality. Because it seems to me that these shared theses are far from obviously true, I am inclined to think that the best they can give is a plausible, attractive and apparently explanatory account (muthos) of that reality. Ferraris has himself devoted some time and ingenuity to revealing the mythological status of some of the elements of Searle’s account, so we may be brief here. As already hinted, Searle’s basic formula is that a physical object X counts as a social object Y in a given context C in virtue of the imposition of a Y-function on X by an act of collective intentionality. And the equally basic objection to this is that it is explanation obscurum per obscurior. For instance, though I know about some of the powers of my €10 note, about some of my duties as a husband, about how little my degrees are worth, and about what documents to produce if I want to pass (relatively) unimpeded from Ventimiglia to Menton, I am deeply in the dark about what “collective intentionality” might be, and about when or where it was enacted in any of these cases. Even if I have some inklings about what intending is, and also about why it is not the sort of thing that social facts can be based on, such collective intendings as there may be found (for instance in the coordination of a football team or a group of musicians—Searle’s own examples) seem to be derivative of antecedent social facts about aims and projects shared by the participants. In this respect, collective intending is as mythical and as useless as the not entirely dissimilar notion of a social contract. Likewise with “counting as”, “context”, “imposing” and “Yfunction”: none of these notions looks elementary enough to help explain any social facts, still less to conjure into being any social objects.

Why Society Is Not Made of Documents

131

At first blush, then, Ferraris’ theory enjoys some distinct advantages over Searle’s. In its paradigm cases, the privileged objects of the documental theory are pretty readily identifiable: the presence or absence of some writing in a given language on a given physical support is a matter about which only cavillers could invent problems. Whether or not we want to reify such things as languages, there is at least nothing mythical about the difference between a text in English and a text in French; and there is nothing mythical about pieces of paper, tablets of stone or hard-disc drives that may serve as supports for such texts so as to constitute documents that are, at least in principle, consultable by more than one person and, hence, that are suitably public and social. Nevertheless, there are some underlying problems, of which we may consider three, two of which are damaging to Ferraris’ central contention about paradigmatic social objects as inscribed acts and the third of which concerns his approach to non-paradigmatic cases of documentality. I take it that these three problems are jointly fatal to the documental proposal as the logos of social reality, while they may leave it standing as a muthos of such things as money, marriage, university degrees and national borders, much as Ferraris’ criticisms of Searle’s theory leave us free, nevertheless, to talk of a certain physical object counting as a social object in a certain context, if we feel so inclined. That is to say, there is no real harm in talking as if social objects were brought into being by acts of inscribing, and indeed, many of the phenomena that Ferraris links together in his systematic exposition of the theory would be hard to see as connected without this way of talking, But it is well to be aware of the mythical status of these objects. The two problems I wish to raise for the characteristic and core thesis of the documental theory are intertwined and may be just two sides of the same coin. One, which I shall rather arbitrarily call the Regress Problem and which I shall consider first, concerns the priority both causal and conceptual of social facts over documents. The other, which I shall call the Validation Problem, concerns the difficulty for Ferraris of recognising that some apparent inscribed acts do not constitute social objects. What I am calling the Regress Problem arises in the following way. On Ferraris’ view, the explananda of a theory of social reality are objects like those we have already cited as the paradigm cases: money, marriages, degrees and borders. To explain the arising of these objects, Ferraris makes appeal to the acts by which they are instituted. “Act” here may waver slightly between an event or action of inscribing (the issuing of a note by the central bank, the signing of the marriage register, the award ceremony in a university or the stipulation of an international treaty) and

132

Chapter Nine

the enduring particular that issues from such an event or action (the banknote, the marriage certificate, the degree diploma or the text of the treaty). But this wavering is not a cause for particular concern, rather as talk of an “Act of Parliament” may refer to the completion of a voting procedure or to the statute that is then to be found in law books. That is to say, we are unlikely to be misled by this categorial slide because we can distinguish fairly easily between the event or action and the document. And, in turn, the documents may be considered either as tokens (as in the case of the banknote: reproduction is forgery and threatens debasement of the currency) or as types (as in the other three cases we have been considering: such registrations need to be copiable for the purposes for which they are kept). The move then is to say that the social object in each case (the money, the marriage, the degree or the border) can be picked out by appeal to the document that is, in some sense, its vehicle. As already noted, it is a salient virtue of such documents that they are readily identifiable and can be appealed to when we wish to determine the contours of social reality. If such an appeal fails and no document can be identified, then, on Ferraris’ view, there is no social object and no social fact. Where there are facts and objects that are not identified by some inscribed act, they must be brute facts or natural objects and, hence, not social in Ferraris’ terms. That is to say, on the documental theory, behind every social fact there must be an inscribed act on pain of an explanatory failure. Though he nowhere makes it explicit, Ferraris is thus committed to the view that in any explanatory sequence of arbitrary length of social facts or objects and inscribed acts, we have an explanation of the facts and objects only when the final term of the sequence is an inscribed act. To take a simple case, suppose we identify the social object that is a marriage by reference to the register that the spouses signed on their wedding day and we want an explanation of that. Ferraris might allow that, for sure, we can explain its being a marriage by reference to social facts such as the eligibility of the partners, the authority of the presiding officer and the regularity of the registration. But, then, he would say that these social facts must in turn be explained by the documents that constitute that eligibility (e.g. certifications that both partners are of a suitable age, unmarried, of sound mind and so on), that authority (e.g. a declaration that the registrar is authorised to officiate), and that regularity (e.g. a law determining the observance of certain procedures). At this point, we might wonder why a sequence (Doc) socialfactinscribedact-socialfact-inscribedact is more convincing or explanatory than a sequence (Soc) inscribedact-socialfact-inscribedact-socialfact, where

Why Society Is Not Made of Documents

133

any given term is explained by the term to its left, and the final term to the right is the ultimate explanation of the sequence. If a (Doc) sequence is the basic form of theses in the theory of social reality, then inscribed acts are always required at the end of any explanatory sequence, and every (Soc) sequence is partial or incomplete as an explanation of social reality. Let us note one feature of Ferraris’ commitment to (Doc) over (Soc). This is that, if there is no society without social facts, there is no society without inscribed acts. In the paradigm cases of the documental theory, the inscribed acts require writing. Hence, there is no society without writing. But there are, as a matter of undeniable anthropological fact, many societies without writing. Hence, the preference for (Doc) over (Soc) does not capture the logos of social reality. If some sequence (Soc) turns out to be an adequate or complete account of some social reality, as it must given that there are many societies without writing, then the documental theory is at best a muthos concerning some of the practices of some societies that do use writing. What the Regress Problem brings to light about the documental theory is that the preference for (Doc) over (Soc) leads to the thesis that there can be no social facts of the paradigm sorts without writing. This may be true as a matter of practicality: some complex social facts may be hard to police in absence of written documents. But practicality and policing are not the objective of a theory of social reality or of social ontology: the objective should be to discern what social facts there are. In at least three of the four paradigm cases we have been considering, relevant social facts can obviously obtain even in the absence of writing. Many societies without writing have developed means for obviating the problems of barter and have thus, to some extent, invented money and attribute to certain objects some subset of the powers that I attribute to my €10 banknote. Almost all societies, including those without writing, recognise arrangements that it is perfectly proper to call marriages because certain reciprocal rights and duties are recognised within family groups. And much social living—including nomadic living—involves the delimitation from time to time of spatial zones into which outsiders may enter only on certain conditions. The case of university degrees is, of course, one in which the social facts in question do make pretty essential reference to writing and are, so far forth, pretty unthinkable in a society without writing. But one might nevertheless see that the institutions in question are evolutions of the ways that, even in societies without writing, certain sorts of expertise confer rights and duties that may be exclusive to those who have matured them.

134

Chapter Nine

If the Regress Problem really is a problem for the documental theory, as I think it is, then one might reasonably infer that sequences (Soc) are better candidates for explaining some of the features of societies that use inscribed acts for the practical purposes of policing particularly complex or evolved social facts. In favour of preferring sequences (Soc), we may also cite the way that inscribing acts is itself a species of social fact, and hence cannot figure in the explanation of the genus of which it is a species. This last feature of sequences (Soc) provides a passage to the Validity Problem, which I have announced as something like the flip side of the Regress Problem. If the Regress Problem for the theory of documentality boils down to the observation that a sequence (Doc) has as its final term an inscribed act that is itself a social fact, and hence every sequence (Doc) is really a sequence (Soc) after all, then the Validity Problem boils down to the observation that social facts about inscribed acts are needed to distinguish between valid acts and invalid acts. Although there are some passages in Ferraris’ exposition of his theory in which he recognises the need to be able to distinguish between valid and invalid documents, the trend of his thought is, as we have seen, to say that the distinction can be made by appeal to some other inscribed act that validates valid documents and hence underpins the social objects that are identified by means of them. Of course, there will be cases in which some other piece of paper can be called in aid to establish the legitimacy of an inscribed act whose validity has been called into doubt. But, if the Regress Problem is a problem for valid documents, it should be clear how things stand with invalid ones: their invalidity derives from the lack of social facts apt to legitimate them. Rather than reconsider the paradigm cases of inscribed acts that, for Ferraris, constitute social objects, we may describe briefly how things stand with what I think may be a paradigm case of an invalid inscribed act: the so-called Constitutum Constantini. Though there is some room for uncertainty about where and when this document was drawn up, copies were in circulation, both in France and at Rome, in the second half of the eighth century C.E., being included in a collection of Papal decretals in the ninth century and in Gratian’s Concordia discordantium canonum of about 1150 (I, xcvi, 14). What the Constitutum inscribes is the donation by the Emperor Constantine of temporal power over the Western Empire to the then pontiff Sylvester and in perpetuity to the Roman Church. In this, it corroborates and is corroborated by fifth-century accounts attributing to St Sylvester a part in the conversion of Constantine by means, among other things, of curing him of leprosy. Though some voices, such as that of Dante (Inferno, XIX, 115-7), had been raised against the effects of what is

Why Society Is Not Made of Documents

135

inscribed in the Constitutum, its validity was taken for granted over five or six centuries. But I would be very surprised, however, if Maurizio Ferraris did not agree with me in saying that the Constitutum does not establish the temporal power of the Roman Church over Western Europe. While this is not a problem for me, for Maurizio Ferraris it would appear to be because we have what looked to everyone like an inscribed act, but no social fact. The point of the Validity Problem for the theory of documentality is that that theory seems to have difficulty accounting for what was wrong with the Constitutum. What seems to be the difficulty? In one sense, it ought to be obvious that, if the Constitutum was not written by Constantine, and so is a forgery, then the inscribed act that it purports to be is not valid. But how can the documental theory account for this? For Ferraris, the long temporal gap between the supposed donation in the fourth century and the earliest manuscripts in circulation might hold the key: if the documental tradition does not stretch back to Constantine, then there was no act of inscribing (in the sense of an event of writing) to set up the inscribed act (in the sense of an enduring particular). But a gap of this sort is not so very surprising with a technology of paper and ink in which a text could only survive (as a type) over centuries by being copied onto ever-new physical supports. From where we are, we cannot verify the absence of intermediates between the fourth century and the eighth. So how can we tell that the text we have before us is not an effect of an act of Constantine’s? Consider the position in the early fifteenth century. We have a powerful institution, the Roman Catholic Church asserting its temporal authority over large swathes of Western Europe. This assertion is backed up by a document that has been reproduced for hundreds of years and incorporated into canon law. This document says that the Emperor Constantine transferred his authority to the Pope. It would seem, then, that the document could be challenged only by another document. For instance, if it were in conflict with what we find in Eusebius of Cæsarea’s Vita Constantini, that might be grounds for doubt; but we have reason for thinking that Eusebius was not as well informed about Constantine’s doings as he presents himself as being. So given that the Constitutum presents itself as being an act inscribed by Constantine and has the support of so many institutions and authorities, the temporal power of the Church was for centuries a pretty solid social object (if you like to talk that way). So far documentality. But this seems back-to-front. Nevertheless, beginning with Nicholas of Cusa’s rather casual assertion in 1433 that the Constitutum was apocryphal (De concordantia catholica, III, 2) and Lorenzo Valla’s blistering attack De falso credita et

136

Chapter Nine

ementita Constantini donatione (1440), the document itself came to be discredited. In some crucial phases of his highly elaborate argumentation, Valla proceeds more or less as follows. The alleged social fact of Constantine’s donating the Western Empire to the Pope was impossible; the Constitutum inscribes his performing that donation; therefore the Constitutum is a forgery. The precise nature of the impossibilities that Valla exposes need not detain us here (though it is stirring reading that I highly recommend), but given that we know something about social facts, we can know something about a document that purports to represent one. If a certain social fact is impossible, then an inscribed act that would constitute it is invalid. If someone thinks that Valla’s procedure as I have caricatured it is in some way circular, for instance because it seeks to show that the Emperor could not have donated the Western Empire on the grounds that the Constitutum says that he did, I would repeat my recommendation to read the De falso donatione and recall that we can often tell that a text is not telling the truth because the things it tells are impossible (flying horses, invisible men, time travel and so on). As with the Regress Problem, the Validity Problem arises out of the way that documental theory seems to repose a certain faith in the genuineness of documents as such. Of course, without some such faith, it would be very hard to manage many of our more complex social affairs. But a document is not made valid only by other documents and is not made invalid by the absence of validating documentation. Rather, its standing depends on social facts about its being drawn up: if Constantine had donated the Western Empire to Sylvester, then that interaction would have been what would have made the Constitutum valid, just as being actually issued by a central bank is what makes my €10 note genuine, being conducted by a mayor is what makes some marriages binding, being earned in a reputable university is what makes a degree certificate worth hanging on the wall, and being established by inter-governmental agreement is what fixes where a border lies. If the Regress Problem concerns in the first instance the paradigm social objects that documental theory considers and the Validity Problem is best exemplified by paradigm non-objects, the third problem for the theory that I wish briskly to raise concerns the treatment of the less-thanparadigmatic social objects that Ferraris wishes his theory to cover. These are cases in which we have a social fact or social object but we have nothing written in any language on any physical support. Among these might be—Ferraris’ own examples—appointments, lunch invitations, bets and threats. These are clearly cases where there is something social insofar

Why Society Is Not Made of Documents

137

as they involve at least two persons: I cannot make an appointment with myself, invite myself to lunch, make a bet with myself or threaten myself. Though they may be written down, they need not be and are in force whether they are or not. Rigorous application of the documental theory would seem to dictate that here we do not, after all, have anything social because there is no inscribed act. If that were the position, the documental theory would collapse: if a theory dictates that an appointment, an invitation, a bet or a threat that is not written down is not a social arrangement, then the theory is not merely muthos, but simply false. And it is simply false because, even if these arrangements are not paradigmatic for the documental theory, they are paradigmatically social. Quite reasonably, Ferraris has to allow that uninscribed appointments, invitations, bets and threats are social, but has to accommodate them to the documental theory. Adopting—or perhaps adapting—some terminology invented by Jacques Derrida in his book On Grammatology (1967), Ferraris proposes that what we have in such cases are phenomena of “archiwriting”. Thus the social fact or object of an appointment, an invitation, a bet or a threat is constituted by there being traces in the participants to these events such that all of those involved recognise their bindingness. These traces are traces of archiwriting in way not entirely dissimilar to the way that the traces of the letters I am now typing are traces of writing, even though for the time being, those letters are just configurations of bits in the computer on whose screen certain configurations of pixels appear. Yet these letters have two characteristics that are worth attending to. One is that, in the combination I am seeking to impose on them, the text that they make up is in a certain language, which is my idiolect of English and may be more or less understandable by other readers of any of the many varieties of English that infest the globe. The other is that the support in which they are currently imprisoned may be accessed in various ways and the writing may be fully actualised, for instance by being printed on paper or disseminated to other computers. As we have noted, it is a strength of the documental theory that it emphasises the stability, shareability and verifiability of the documents that constitute social facts. But the appeal to Derridean archiwriting throws this strength away. For Derrida, archiwriting need not be in any language, such as French or English, and may be essentially private, residing only in the mind or brain of a single individual. I do not know how seriously Derrida himself took this proposal, or whether he meant the various things he said about it to be taken seriously by others. But, for myself, I find it

138

Chapter Nine

hard to take at all seriously because it is highly implausible to suppose that anything that answers to his description can be appealed to explain the social facts that are non-paradigmatic for the documental theory. If archiwriting is not in any language, then the content of a trace of it is not responsible to any rules of meaning. Indeed, it is quite mysterious how such a thing could have meaning at all. And if it can be private, then it cannot be used to explain anything essentially public or social. Indeed, it is quite mysterious what would difference its presence or absence would make. In these respects, archiwriting cannot give the logos of social facts. To avoid the falsity of the documental theory in the face of very many day-to-day social facts that do not involve paradigmatic documents, Ferraris appeals to the notion of archiwriting, and thus consigns his theory to the realm of muthos. If my friend—perhaps Ferraris himself—wants to say that I should be at a certain restaurant at a certain hour because our appointment for lunch was constituted by a document in archiwriting inscribed on his brain and mine (and mutatis mutandis in the other nonparadigmatic cases for the documental theory), I might think that his way of talking is a bit contorted and unenlightening; but it is only if he persists in talking this way in the philosophy room, or publishing books to promote this way of talking in others, that I might begin to protest. As I have.

References Derrida, J. 1967, De la Grammatologie, Paris: Les Éditions de Minuit Ferraris, M. 2009, Documentalità: perché è necessario lasciar tracce, Roma-Bari: Laterza —. 2013, Documentality: Why it is Necessary to Leave Traces, English translation by R. Davies, New York: Fordham University Press Ryle, G. 1949, The Concept of Mind, London: Hutchinson Searle, J. 1995, The Construction of Social Reality, Harmondsworth: Penguin —. 2010, Making the Social World: The Structure of Human Civilization, Oxford: Oxford University Press

CHAPTER TEN HOW TO INCORPORATE (WITH) DOCUMENTS PETAR BOJANIû

Regardless of the fact that I am using certain texts by Searle, Smith and Ferraris (including a criticism of a position of Searle’s, and the response thereof),1 my intention is not at all to reiterate someone else’s position in my own words, nor is it to question or modify some such position. My intention for now is to, using Smith’s theory of the document and Ferraris’ documentality, affirm the existence of several paradoxes—of which Searle rejected a few, but unconvincingly so, I think—regarding the institution (or the institutionalization of the institution) or corporation. In order to do that, it seems to me that I am forced to slightly disturb Searle’s conception, in attempting to offer yet another realist contribution to a new future theory of the institution. To begin with, I would connect two completely new elements in the social ontology of John Searle, which can be found in his brief sketch of a text on the institution from 2005. Both innovations appear thanks to Barry Smith, and Searle then repeats them once again in his latest book, without further developing them together. In “What is an Institution?”, Searle mentions the document as an addition to, but not condition or background, of any deontology. This is precisely the problem. But the deontic powers stop at the point where the larger society requires some official documentation, they lack full deontic powers. Collective recognition is not enough. There has to be official recognition by some agency, itself supported by collective recognition, and there have to be status indicators issued by the official agency (Searle 2005, 15).

The insufficiency of which Searle speaks refers to the difficult transformation or transfer of a social fact into an institutional one. To reinforce or stabilize collective recognition (it seems to me that institutionalization is precisely this process), it is necessary to de facto “draw” or “extract” (or activate) the document from some, in this case,

140

Chapter Ten

third place. This introduction of the document or its transfer from place to place, could be explained with a sort of specifying of the social, which in turn is the unconditional condition of any institutionalization. The second fragment, a most basic proof of Searle’s anti-realism, could be completely reconstructed and corrected only if the reality of the document and document acts were truly accepted. Of course, the reality of the document concerns the explanation of the nature of the power of documentality. He [Barry Smith] pointed out that there are some institutions that have what he calls “free-standing Y terms”, where you can have a status function, but without any physical object on which the status function is imposed. A fascinating case is corporations. The laws of incorporation in a state such as California enable a status function to be constructed, so to speak, out of thin air. Thus, by a kind of performative declaration, the corporation comes into existence, but there need be no physical object which is the corporation. The corporation has to have a mailing address and a list of officers and stock holders and so on, but it does not have to be a physical object. This is a case where following the appropriate procedures counts as the creation of a corporation and where the corporation, once created, continues to exist, but there is no person or physical object which becomes the corporation. New status functions are created among people—as officers of the corporation, stockholders, and so on. There is indeed a corporation as Y, but there is no person or physical object X that counts as Y (Searle 2005, 15-16).

“[...] but there is no person or physical object which becomes the corporation”. It seems to me unnecessary, easy and complicated at once, to show the shift in understanding of the phrase “physical object” in our day and age—is it necessary to prove the physical existence of an SMS, deposit, bank account, electronic signature, the displayed plaque of a company, the existence of telephone calls among the signatories of a charter, etc.? The physical object is never singular, just as the performative declaration is not singular, the way it is in Searle—on the contrary. Using two paradoxes or two models in which “documentality” appears—document acts refer exclusively to institutions, that is, corporations, meaning that there is no institution or corporation without documents, or else that the institution (or the corporation, as a specific form of the institution) is real thanks to document acts, and not speech acts—my intention is to explain how to incorporate (with) documents. I would like to transform the three best definitions of the document we currently have (two are Smith’s and the third is Ferraris’) into three essential characteristics that define the relation between the institution

How to Incorporate (with) Documents

141

(corporation) and documentational protocols: a) the document is something that is able to endure self-identically through time, b) that through documents things are done and the world is changed “by bringing into being new types of ownership relations, of legal accountability, of business organizations, and other creatures of modern economies, including mortgages, stocks, shares, insurance protection, and financial derivatives” (Smith 2008, 2012), and c) that the document is at once a representation of some fact and the inscription of an act (Ferraris 2013, 250).2 The first characteristic that protects documents from change and decay in time, and gives them an always distinct concreteness, is the plurality or series of documents tightly bound together, forming a chain and referring to one another. Documentality always supposes a series of documents, a multitude of permanent acts that refer to and succeed one another, complement and reaffirm each other. A marriage is composed neither of the marriage ceremony, nor the contract of the two sides, the promises and speech acts of the main actors, the signatures in registrars—their own or that of their witnesses—but rather the hundreds and hundreds of future documents (sometimes even documents that precede the marriage ceremony), which reassert again and again and supplement the identity of the marriage covenant between two persons and its beginning in time. In that sense, a series of documents that are in no way “accidentally” linked and cannot be “accidentally” linked represents and affirms an institution. Without the series of document acts, there is no institution of marriage, nor the memory that the marriage ever existed. The written record cannot be preserved “accidentally”, but rather only in connection with institutions (Groebner 2004, 124). The second characteristic of the document or documentational process refers to the documents’ authors, who are always plural (despite the fact that there is, for example, the corporation sole, comprised of one person only). The procedure, then, we designate with the words “to incorporate” or “to institutionalize” ought to be twofold or operate in two directions. I tried to express the difficulty of describing it properly with the title of this paper, where a group constitutes itself (as a group) into a legal entity (i.e. incorporating), while at the same time producing and making use of documents. The document is the product of a collective process, and at the same time, that same group is constituting itself “with” those same documents. This always double process of the same procedure represents the institution (or corporation). The third characteristic refers to the power produced by institutions or corporations, and which is essentially documental. The size, and then the

142

Chapter Ten

power of a group producing documents, and produced with those same documents, ought to be proportional to the “documental power” to further commit all the actors of the institutionalization or incorporation.

1. Paradox I: The Institution Thoroughly simplified, one of the basic paradoxes of Searle’s institutional ontology can be formulated as follows. “The creation of institutional facts by declaration presupposes the creation of other institutional facts. In order to avoid an infinite regress, there must be a way of creating institutional facts which does not require any special authority” (Prien & others 2010, 167)—which seems to be impossible. Clearly I would like to remove both the words “seems” and “impossible”, since I am not satisfied with the perspective offered by Searle’s response.3 My point, however, is not to claim that there is necessarily something outside the institution (something extra-linguistic) or some necessary authority (special or not), power or violence. That is to say, in thinking about performatives as declarations, Searle differentiates extra-linguistic declarations “such as adjourning the meeting, pronouncing somebody man and wife, declaring war, and so on—and linguistic declarations—such as promising, ordering and stating by way of declaration” (Searle 2002, 170). For Searle, these non-linguistic cases are prototypical of declarations, and their main characteristic is that they are not derived from semantics. In a well-known example which Searle quotes several times in different places and in different ways, a man can divorce his wife by uttering three times the sentence “I divorce you”. The divorce will in certain Muslim countries actually take place, says Searle, because speech acts in these cases are derived from legal or theological powers (Searle 2002, 170). Power or powers is a word which is used here perhaps for the first time in this way, whereas recently, as we know, many of Searle’s texts are organized around that word or words force, violence or constraint. It might be important that what Searle names as power or extralinguistic declaration, could also be named a document. War has been declared because the decision was reached and the declaration dispatched to the other side, the meeting was suspended because someone holds a warrant, a piece of paper, and exercises an authority, while the procedure of repetition of the sentence “I divorce you”, is in fact a quote found in codices and religious rules of some Muslim minorities and tribes. For war to be declared and of course begin, it necessarily needs to be written down somewhere. A “homo documentator” is necessary (Briet, 1951, 19). So, what Searle here calls “power(s)”, in fact comes from the document and

How to Incorporate (with) Documents

143

the consent that precedes the linguistic declaration, and in a sense even institutions in general. What will become a new problem, but remains outside the scope of this article, is that even so-called speech acts, which are derived from semantics (promising, ordering, declaring, etc.), could also follow from various collections of rules and laws, and have a documental origin. Contrary to Searle, then, an institutional fact is immediately preceded by the document, in the broadest sense of the term. In other words, a document is a special kind of institutional fact. And if I had to “document” this paradoxical moment now in Searlean language, I would take a sentence of his about the corporation in his latest book, Making the Social World: “So the Law is a Declaration that authorizes other Declarations”.4 In this case, “Law” is the document, and it is not at all surprising that it is precisely when we arrive at “A Complex Case: Creating a Corporation” that the “special role of writing” is problematized and that syntagmas like “writing language”, “written speech act”, “written constitutive rules”, or “written record” appear.5 To the extent to which the title and subtitle of this paper had to be imprecise, I will now assume that in between so-called “brute” and “social” facts, and then in between “social” and “institutional” facts, there is some sort of documental reality. Since we are dealing with the document, let me tangentially insist on the material (the paper, the ink, the body of the text, or if you will, the sound, the phoneme, the materiality of the signifier, the body of one making it). Namely, between, on the one hand, the virtual reality of the law or certain rules (a space in which a usually small group of people, certainly empowered or protected by some authority [or more precisely protected by weapons], imposes an institutional reality on others [or on all] by constructing [formulating, designing] the text of the law or rules) and, on the other hand, various statements often read (or uttered) by a rabbi, priest, lawyer, officer or air hostess, or else a money bill, property, marriage, or a declaration of the type “This is my house”—that is, between these two “realities” there is a so-called “written record”. This is the charter which creates a legal person or corporation, a decision of the governor to issue bills of this specific design, a record, a birth certificate, marriage license, lease, contract, proof of ownership, etc. I do not have to declare “this is my house”, nor say “I am married to Sonia”, to only then create the right to the house “because the right only exists by collective acceptance”. The possibility to document what I say when I say that “this is my house”, to show my papers, my ID, driver’s license, to demand that the lawyer show me the article of law that allows the creation of a corporation, or that the employer produce the decision based on which I am being let go—is paramount for

144

Chapter Ten

the creation of institutional reality. But not sufficient. The sentence “our marriage exists only on paper” (I am not sure whether this sentence works in English; this is what it would be in German. “Wir sind nur noch auf dem Papier verheiratet”) marks that our current relationship is not in harmony with the contract we signed, and that our marriage is not worth the paper it is written on. One of the main characteristics of the institution, which Hume differentiates from and opposes to the contract (law)—apart from that the institution, unlike the contract, supposes the existence of a third party (that is, the possibility that someone else join, a sense in which the institution implicitly counts everyone in, excludes no one, and ultimately assumes that there is nothing outside the institution), and that the institution is coercive because it is formed by various transformations of violence and force—is its artificial nature, as well as the possibility of incorporating law into a group or tying a collective together. The fact that the paper (the marriage certificate), as a document, is not a strong enough connective to keep a couple in love—after all, neither is the once upon a time performed ceremony in which we promised to have and to hold one another—does not exclude the institutional fact that our marriage still exists. The question then implied by the “normal little words ‘real’” (and which are not normal at all)6 in the subtitle of this paper (“real” and “realism”) regards the existence of the institution of marriage that exists only on paper, that is, the status of the document within the institution. Do we have a real marriage (or a factual marriage) because we are not divorced, or is the document the source of reality and then the institution? In other words, does the document institutionalize, or are we in fact, really separated, we could say organically separated, and together on paper only?7 An institution is real if and only if it can be documented. That is, to institutionalize (I choose the verb) means in fact to publish or further attach documents (a synonym for documentum or its prototype, is instrumentum [a statement made publicly, or in the presence of several witnesses]; in a different context, documentation is nothing other than argumentation). The institution becomes an institution if it is constantly in the process of institutionalizing or documenting. To defend this option, it seems necessary to me to show that a greater transfer or distribution of paper (documents) between partners, certainly makes the marriage not only on paper. The more invoices, bills, receipts, tax declarations, etc., put simply: the more papers or the more documents—the more love. Reality is thus ensured by the production and proliferation of documents, and their collecting.

How to Incorporate (with) Documents

145

2. Paradox II: The Corporation What is it that the following sentence is saying? “If you are planning to start your own business, you should know how to incorporate”. Does the second part, “how to incorporate”, have anything to do with the institution? More interesting, I think, than enumerating all the differences between the institution and corporation (or company), and the usual disdain towards the corporation as an emblem and essence of capitalism, would be to insist on a set of various operations or cooperations that go along with the claim made by person X about the necessity of possessing a certain kind of knowledge as the basis for, ultimately, producing profit and sustaining that profit or company.8 A company, a firm, a partnership, a société anonyme, a corporation, has to be registered (to be set in the register, to be incorporated) according to rules set by the state, generally simple, and which should be familiar (“you should know”).9 “X” is first a person (an “institutional person”, not just a subject or agent).10 “Institutional person” or “corporative person” implies the transfer or transformation from the singular to the plural, from first person singular to first person plural; “X” has certain interests and is surrounded by others with whom it wishes to work and who share that interest (right away we encounter the representative of this group, which will quickly lead us to Hegel and the problem of interest representation or identity representation; in that case, the unit or group will be incorporated in the action of a certain portion of the whole that represents—because considers itself identical to—that whole); further, this is neither a group of thieves, nor criminals, nor undocumented workers, since the group is ready to publicly declare its work (through “public declarations”; of course, it remains to be seen whether the institutional structure allows the group to make certain public declarations, or on the contrary, the public declarations institutionally design the group); “X” can also be single, in which case we are speaking of a “corporation sole; a legal entity consisting of a single (‘sole’) incorporated office, occupied by a single man or woman” (however, in that case, what has been abandoned is, in Searle’s words, one of the main inventions of contemporary capitalism— limited liability; thus corporation sole is characterized by unlimited liability, meaning that if you have business debts, personal assets would be used to pay them off); further still, “plan to start your own business” implies constituting a new, dynamic entity, capable of competing with other entities, fighting for profit, expanding and developing; finally, “to incorporate” (with all the theological and above all Christian connotations

146

Chapter Ten

of the term) de facto represents a certain documentation procedure, or writing act or acts. In these last two operations, certain real, living people,11 exist and “transform”, through a given procedure (one made up of documents, hence objective), into a completely new entity or new form in order to more simply, quickly or safely achieve certain goals (gather new people, money, property, reserves, funds, only to redistribute them again, since “to have” always also means to distribute, to change that which is). It is precisely these last two operations that introduce the word reality or realism. Paradoxically, the question of the real existence of this new, incorporated entity (or corporation) cannot be separated from the real existence of a group that has just been constituted and united by incorporating, and is now together “on paper”. Before I give my reasons to hold together the three words (that Realism is, or presupposes or assumes a connection between Institution and Corporation), I can say, from the off, that I am interested in money and profit (or more precisely, “amount” of money or profit, or even better, “amount” as such) as the integrative element or factor of a community, as well as one of the main attributes of Realism. As an example, let me offer a case that is as trivial as it is common, and that academics doubtless encounter often. A professor from the University of Irvine in California, expert in social ontology, was out to dinner in Milan with a few colleagues. Immediately before ordering food, the professor hesitates over the menu, and asks discreetly who is paying for the dinner. It is important to establish, before the food is ordered, that neither she nor anyone else at the table is paying “personally”, out of their own pockets, for dinner in Milan. Imagine that the university in Milan, which is in fact paying for the expensive meal, is a corporation—and not just an institution, but a corporation (whereby one of the simpler definitions of a corporation could be “a collection of many individuals united into one body, under special denomination and authorized by law to act as a unit”), that is, something insufficiently existent (someone without soul and body [“universitas non habet corpus nec animam, est res inanimata”, Pope Innocent IV], someone fictional, etc.). We ought heed this comment by Sinibaldus de Flisco, aka Innocentius IV, whose interpretation is usually considered relevant for so called fictitious theory, that is, for a potential explanation of a virtual entity or “as if” entity. We are dealing with something deficient, which exists but does not contain all the attributes of existence, which is a something or a thing, but a thing without either “body or soul”. Further, someone is paying for dinner (the restaurant will send the university a bill or one of the hosts will pay and be reimbursed at a later date)—without money it would be truly impossible to spend, say, two and

How to Incorporate (with) Documents

147

a half hours in a restaurant in Milan. Finally, the restaurant in Milan (an establishment registered as a partnership, for example), the university (an institution or corporation or “company” [the continental name for a corporation]), registered or incorporated Italian banks through which one pays, etc., are, for instance, three sites among which certain documents, receipts, copies, contracts, and charters flow. These documents, then, issued and controlled by certain departments or agencies of the state, ensure the reality and duration of these organizations. It seems to me that it is above all the number of people, money, documents, institutions, food, enjoyment, as well as the level of responsibility for the expenditure or crisis or bankruptcy, that determines the power or degree of realism here in play. I suggest that the concept of “amount” and number regulate the status of the “real” and admit that this is one of two reasons why I placed the word corporation, a specific kind of institution that immediately represents business, earnings, profit, and expansion, in the title. For something to be “real”, it must exist in time and in a process of realization. Among the many characteristics and themes initiated by the corporation (certainly one of the more significant human inventions), some of the most important are the protection of the rights of the individual, limited liability and corporate responsibility (the sentence “I didn’t do it; the company did” belongs to this protocol), as well as the relation between corporate and canonical law, the partnership and corporation, and not least the decades-long resistance and critical and hypocritical stances toward the corporation as a foundation of capitalism. However, the principal reason why I introduced the corporation is of course the status of the real (or the status of real), immediately brought into question when this “judicial hallucination” is mentioned. In other words, could the corporation or the company be a good instrument for the reconstruction of our concept of reality, and then all the alternative terms that have always stood side by side with the terms “real” or “realism” (effective, actual, concrete, evident, objective, material, physical, factual, etc.)? It is this link between the institution and realism (“institutional realism”)12 or reality and corporation that ought to affirm three circular moments which I would like to defend, and which are the key contribution to my little future theory of the institution. First, to register or to be incorporated (by way of charter or document) means to be “real” or, more generally, to “exist”. Second, even before the incorporation which assumes making a new person or new body (public and official) along with others (transforming oneself into something else along with others), there is a multitude of relations and they are social facts (which is why the

148

Chapter Ten

corporation or institution will be real).13 Third, that a “written document” or a “written act” (written Declaration or Charter) constitutes or institutionalizes a group or collective (and protects its existence in time). Thus, the institution is always preceded by certain, less developed, institutional forms. (Cautiously, we could here use Searle’s term “institutional background”). This idea of circularity protects me from certain other elements which play a role in Searle’s theory of the institution and corporation: the advantage of speech over writing, and the absence or hesitation in the introduction of the document (and a document is not an object); the importance of power and its ultimate example—the state (the institution above all institutions), which always has the first and last word in registering a corporation; the collective is secondary, and Searle displays carelessness when saying that acceptance of others is a necessary condition of creating deontic powers; and finally, Searle’s understanding of the physical object is too obscure and points latently to his anti-realism. The amount of documents, various transactions and contracts within the corporation and institution, truly surpasses the importance of the act of the founding of the corporation (“executing and filing articles of incorporation” or “filling a legislation document with a state officer”), and places the lesser importance on the function of the state, and in general the medieval institution of creating a corporation (fiat doctrine), something that especially interests John Searle. If we plot the role of the state in founding corporations (or institutions in general) through history, we can notice that only in certain very rare periods (of colonialism, for example) does the state found and control, meaning gives the privilege to incorporate to certain powerful groups. For that reason it seems to me that even a moderate institutional realism necessarily assumes an “ontological egalitarianism”, generally lacking in Searle.14 Homo documentator, as the inheritor and contributor to Paul Otlet’s great project Mundaneum, certainly surpasses “the government as ultimate institutional structure”,15 and puts in the place of the state—Europe and la Cité mondiale or République mondiale.16

References Austin, J. 1962, Sense and Sensibilia, Oxford: Oxford University Press, lecture VII, 62-78 Briet, S. 1951, Qu’est-ce que la documentation?, Paris: Édit Desqueyrat, A. 1933, L’institution, le droit objectif et la technique positive, Paris: Sirey Dejnozka, J. 2007, Corporate Entity, manuscript

How to Incorporate (with) Documents

149

Elon, M. (ed.) 1975, The Principles of Jewish Law, Jerusalem: Encyclopaedia Judaica Ferraris, M. 2009, Documentalità. Perché è necessario lasciar tracce, Roma-Bari: Laterza —. 2013, Documentality: Why It Is Necessary to Leave Traces, New York: Fordham University Press Franken, D. Karakus, A. and J.G. Michel (eds.) 2010, John R. Searle: Thinking about the real world, Frankfurt am Main: Ontos Verlag Le Goff, J. 1978, “Documento/Monumento”, Torino: Enciclopedia Einaudi, V: 38-48 French, P. 1979, “The Corporation as a Moral Person”, American Philosophical Quarterly 16, 207 Grafstein, R. 1992, Institutional Realism, New Haven & London: Yale University Press Groebner, V. 2004, Der Schein der Person. Steckbrief, Ausweis und Kontrolle im Mittelalter, Muenchen: Verlag C. H. Beck oHG Hodgson, G.M. 1999, Evolution and Institutions, Cheltenham: Edward Elgar Knight, J. 1992, Institution and Social Conflict, Cambridge: Cambridge University Press Mayer, C. 2013, Firm Commitment. Why the corporation is failing us and how to restore trust in it, Oxford: Oxford University Press Otlet, P. 1989, Traité de documentation: le livre sur le livre, théorie et pratique, Liège: C.L.P.C.F. Pettit, P. 1996, “Groups with Minds of Their Own”, in Schmitt F. (ed.), Socializing Metaphysics, New York: Rowan and Littlefield Prien, B., J. Skudlarek and S. Stolte 2010, “The Role of Declarations in the Constructions of Social Reality”, in Franken D., Karakus A. and Michel J.G. (eds.) John R. Searle: Thinking about the real world, Frankfurt am Main: Ontos Verlag, 163 Searle, J. 2002, Consciousness and Language, Cambridge: Cambridge University Press —. 2005, “What is an institution?”, Journal of Institutional Economics 1, 1 —. 2007, “Social Ontology and Political Power”, in Searle, J. Freedom & Neurobiology. Reflections on Free Will, Language and Political Power, New York: Columbia University Press —. 2010a, Making the Social Word, Oxford: Oxford University Press —. 2010b, “Reply to ‘The Role of Declarations in the Constructions of Social Reality’”, in Franken D., Karakus A. and Michel J.G. (eds.)

150

Chapter Ten

John R. Searle: Thinking about the real world, Frankfurt am Main: Ontos Verlag, 227 Smith, B. 2008, “Searle and De Soto: The New Ontology of the Social World”, B. Smith, D. Mark and I. Ehrlich (eds.), The Mystery of Capital and the Construction of Social Reality, Chicago: Open Court, 35 —. 2012, “How to Do Things with Documents”, Rivista di Estetica 50, 179 Sorokin, P. 1966, Sociological Theories of Today, New York: Joanna Cotler Books Tuomela, R. 2011, “Searle’s New Construction of Social Reality”, Analysis Reviews 71, 708

Notes 1

I am speaking of the text by Bernd Prien, Jan Skudlarek, Sebastian Stolte “The Role of Declarations in the Constructions of Social Reality” (D. Franken and others 2010), and John Searle’s “Reply” in the same volume. Both texts came about as the result of a conference held in Münster in December 2009. 2 Ferraris reiterates a long tradition beginning with Fustel de Coulanges, which is positivist in essence, that the document is above all a text (Le Goff 1978, 39). In that way Ferraris is faithful to a great project begun a hundred years ago, which he develops and complements, thus transforming it. I am speaking of Paul Otlet’s writings, which appear at the beginning of the century: Constitution mondiale (1917), La Cité mondiale (1927), and the book, Traité de Documentation, published in 1934. The “Palais Mondial” Mundaneum was created in Brussels in 1920 (having been conceived in 1895 when the “Institut International de Bibliographie” was founded). “Documentation” was envisaged “comme science globale”, and Suzanne Briet tried to introduce documentation as a discipline at the Sorbonne (and was rejected with peremptory explanation: “la documentation n’existe pas”). 3 Searle’s response to this problem is that it is not a problem. “A further point of disagreement between me and them is that they think you need a special authority to create institutional reality by (representations that have the same logical form as) Declarations. This is a mistake as several of my examples illustrate. You do not need special authority to create every type of institutional fact, otherwise it would not be possible for the system of institutional fact to ever get started. You have to begin somewhere, simply by creating and getting other people to accept institutional reality” (Searle 2010b, 229). I think it was sentences like this, in which someone who creates institutional reality ought to get others to accept it, that prompted Raimo Tuomela’s criticism about individualistic collective acceptance of what has been declared and of speech act theory as essentially an individualistic theory (Tuomela 2011, 708). In the book Making the Social Word,

How to Incorporate (with) Documents

151

the word “making” serves Searle in explaining the beginning of the institution or institutionalization. “God can create light by saying ‘Let there be light!’ Well, we cannot create light but we have a similar remarkable capacity. We can create boundaries, kings, and corporations by saying something equivalent to ‘Let this be a boundary!’ ‘Let the oldest son be the king!’ ‘Let there be a corporation!’” (Searle 2010a, 100). 4 Searle 2010a, 100. The corporation is the novelty in Searle’s latest book, even though he mentions it for the first time in 2004 at a conference at the University of Hertfordshire. 5 Searle 2010a, 98-100, 115. Law is law, or the document is the document, because it is alive and vocal, because the letter (the paper) “has a voice”. For example, Jewish political theory and Jewish Law Theory recognize a clear distinction between the Verbal and Written Obligation. A written obligation entitles the creditor to recover payment out of the debtor’s encumbered assets which are in the hands of a third party, a right unavailable in the case of a mere verbal obligation, since the obligation or debt has no kol (“voice”) and does not provide notice that will put prospective purchasers on their guard. In the case of a written obligation, a plea by the debtor that he has repaid the debt is not accepted without written proof, as would be the case with a verbal obligation. Thus, for example, an undertaking, even if in the debtor’s own handwriting but not signed by witnesses, will be treated as a verbal obligation, since only a properly written, witnessed, and signed obligation carries a “voice and constitutes notice”, Bava Batra 175b (Elon 1975, 244). 6 Austin 1962, 71. Let me offer a reminder that in his 1959 lectures, Austin designates the “little word real” first as a “normal word”, and then claims that “real is not a normal word at all”. 7 Hodgson’s “organicist ontology” (Hodgson 1999, 89) thematized the organic and biological analogies throughout the Western world. It seems to me that the introduction of the concept of “organization”, as well as the differentiating between an organization and institution in the last century, is analogous to contemporary differentiation between the real and the institutional. Organization refers to subjects, to collective actors, whereas institution refers to the relations between actors, to the objective (Knight 1992, 3); the institution refers to the rules of the game, the organization to the players. 8 In his thesis on institutionalism (1933), André Desqueyrat speaks about “institution corporative” which has three characteristics: the idea of a goal that ought to be realized in a social group; organizational power for the purpose of realizing that idea; and the manifestation of a community emerging from the social group for the sake of the idea and its realization. 9 “In America, for example, there are 6 million companies, employing 120 million people, which is two fifths of the entire American population. Around 3 million new companies are registered world-wide every year” (Mayer 2013, 22). 10 This is my paraphrasing of Pettit, that the personal point of view is the condition for institutional design: “The personal point of view must have this indexical, firstpersonal character” (Pettit 1996, 260).

152

Chapter Ten

11 “Corporations are at least as real as people”. “Corporation is real because it is an artifact whose constituents typically include real people as well as real assets, as well as intangibles such as credit and goodwill” (Dejnozka 2007, 71, 102). 12 “Institutions represent—in a sense, are—our connectedness [...] Institutions are physical wholes composed of human parts” (Grafstein 1992, 13, 22). Institutional realism is supposed to oppose new institutionalism, as yet another extreme form of conventionalism, as well as anti-realist philosophy of social science. “Realism assures us that those reciprocal influences which we are then in a position to attribute to institutions and participants are physically realized in ways that, in principle, are comprehensible to science. As aggregates of human beings, institutions are, unproblematically, entities determined by their participants. As distinct physical entities, institutions can be distinct social forces”. 13 In a sense, this is a reformulation of the so-called Reality Theory defended by Otto von Gierke. “Reality Theory recognizes corporations to be pre-legal existing sociological persons. [...] Law cannot create its subjects, it only determines which societal facts are in conformity with its requirements” (French 1979, 209). The Law only recognizes the corporation; it did not create it. Regardless of the fact that von Gierke differentiates the corporation from the institution (Anstalt), collective personality is, according to him, not a fictitious person, but rather a real existence (eine reale Existenz). Here is a passage explaining more clearly the corporation and institution, quite in line with von Gierke. Pitirim Sorokin writes: “Since the Roman law, two main forms of the juridical personality have been distinguished: (1) Corporations (universitas personarum or the medieval collegia personalia) where the union of the members as persons is stressed—such as most of various corporations, incorporated societies, firms etc. (2) Institutions (universitas bonorum or the medieval collegia realia) as a complex of property with a specific purpose, endowed by the law to act as a single person, such as various universities, asylums, etc.” (Sorokin 1966, 38). 14 “Relational descriptions tying objects together—son of, slave of, president of, votes for, interrogates—are not, according to ontological egalitarianism, ontological superior to other collections of objects” (Grafstein 1992, 24). 15 Searle 2007, 96; 2010a, 161. 16 Otlet 1989, 419. “The institution, says Paul Otlet, is a free grouping of forces of the will, a federation of organisms, a union of national and international associations” (Otlet 1989, 417). Ferraris’ book is structured in the same way and the chapters are marked like in Paul Otlet’s book.

CHAPTER ELEVEN TO BE REALIST WITH HISTORY: WHAT REALLY ARE HISTORICAL DOCUMENTS? STEFANO VASELLI

1. Premise: the oblivion of philosophy of history as social ontology The golden age of analytical philosophy of history is epitomized by three main works: Arthur C. Danto’s Analytical Philosophy of History (1965, 1985), William Dray’s Philosophy of History (1964, 1966), and Hayden White’s Metahistory (1973). Anglo-American philosophy of history was rather entrenched within the tradition of logical-empiricism, giving also important contributions to the Popperian debate on falsificationism. Authors in this tradition emphasized the analogy between science and historical methodology, and defended the principle of verifiability (as in the case of Hempel and Oppenheim 1948), refutability and nomological generalization (within Popper 1957, and Popper’s legacy, as in Donagan 1966). Along with (and besides) this tradition, other authors between the 1970s and 1980s were increasingly influenced by continental thinking, like hermeneutics, post-structuralism, the Frankfurt School, and post-modernism. Reference to this debate can be found in contributes like Rorty (1979) and Danto (1985), who believed that the richness and texture of the historical narrative was more salient than accounts based on causal explanations of historical outcomes (Little 2012). Frank Ankersmit (1995) fixed many of these themes in his notion of an historical narrative (Ankersmit and Kellner 1995). This new awareness in philosophy of history gradually led to a sort of relativistic subjectivism and hermeneutics rather than insisting on the idea of an objective truth (the so-called “Mr Gradgrind statement” of correspondence to the facts).1 Another fundamental feature of this approach is the preference for a historicist

154

Chapter Eleven

rather than a naturalistic approach to human nature and its ontological essence. This is the case of the Hegelian-Marxist perspective, which holds that human consciousness is itself a historical product, and that it is an important part of the historian’s work to mould together mentality and assumptions of individual agencies in the past (Pompa 1990). The result of this view was a strong and persistent de-ontologization of philosophy of history, perhaps with the only exceptions of Davidson (1963) about events and actions, and Hacking (2002). In what follows I claim that this deontologization is misleading and that it is time to re-establish a role for ontology in the analytical philosophy of history. Philosophy of history should be understood as an ontology of a special kind of social science: historiography. A proper methodology for historiography would then vindicate a genuine role for social ontology in a realist guise. I’ll call this version of philosophy of historiography, historiographical realism.

2. Why must philosophy of history focus on social ontology? There are several aspects of this problem. In fact, supposing that history is a social science with a methodology of investigation (whose settlement is still a work in progress), what does it mean making history from a realist point of view? First, realism, in historical sciences, stresses the existence of a reality that is independent from our capacity and abilities to conceive and know it. How is it possible that we can refer to a past which, by definition, is something that does not exist anymore? This is one of the most important metaphysical dilemmas raised by Michael Dummett in Truth and the Past (Dummett 2004). Second, if the idea that history is a work in progress has got some hope to result a valuable idea, and our previous question admits a positive answer, then shall we postulate the existence of historical “objects” or “entities”? In this case, what, exactly, are these entities? Third, if we answer in the positive way to the previous question, then what should historical realism be conceived in contemporary historiography? A tentative reply might be: (i) In order to investigate the historical past and present by means of a realist methodology, we need to adopt an objectivistic and anti-relativistic commitment, in agreement with the idea that there are informational structures, which are ontologically independent from interpretation.

To Be Realist with History: What Really Are Historical Documents?

155

(ii) So, those structures could be confirmed by discovering, observing, analysing, and classifying recordings (historical records) and inscriptions from historical past (from fossils to archive’s traces, evidence, and so on). (iii) This way, the entia of historical ontology are, in the first place, reproductions, prima facie inscriptions, monuments, and, more in general, historical documents, and in addition, past events and states of affairs. (iv) Yet these entities are neglected by the mainstream methodology of history, dominated by an antirealist and constructivist relativism, an attitude that closely resembles Nietzsche’s prospectivism (Nietzsche 18871988). Relevant examples of this attitude are given by some of the most important historians of twentieth century: The historians cannot work without testimonies and these testimonies become documents only thanks to the importance accorded to them by the historian himself, and within the works that this same historian makes with them. Properly speaking, there are no sources, or, more exactly, to make sources real it’s indispensable that the historian becomes a water-diviner; historical facts are not objective phenomena which exist without historians, but they are the results of the work and of the construction by the historian, who is the creator of the historical facts. To make the job of the historian (and to make the historical activity itself) work, let the historian give some, determined, questions to testimonies. [….] This is the well-known conception of “Annales” methodology (Le Goff in Bloch 1998, X-XI; our italics). The facts speak only when the historian calls on them: it is he who decides to which facts to give the door, and in what order or context. It was, I think, one of Pirandello’s characters who said that a fact is like a sack—it won’t stand up till you’ve put something in it. The only reason why we are interested to know that the battle was fought at Hastings in 1066 is that historians regard it as a major historical event. But, if historian could be the creator of historical facts, how he proceeds in his work of interpretation, choice, creation of historical narration events? Let us assume for present purposes that the fact that Caesar crossed the Rubicon and the fact there is a table in the middle of the room are fan of the same or of a comparable order, that both these facts enter our consciousness in the same or in a comparable manner, and that both have the same objective character in relation to the person who knows them. But, even on this bold and not very plausible assumption, our argument at mice runs into the difficulty that not all facts about the past are historical facts, or are treated as such by the historian. What is the criterion which distinguishes the facts of history from other fan about the past? (Carr 1961, 14-16).

In opposition to this view, Poincaré wrote, more than sixty years before Carr’s quote:

156

Chapter Eleven We cannot know all the facts, since they are practically infinite in number. We must make a selection […] There is no disputing the fact that a selection must be made: however great our activity, facts outstrip us, and we can never overtake them; while the scientist is discovering one fact, millions and millions are produced in every cubic inch of his body. Trying to make science contain nature is like trying to make the part contain the whole. But scientists believe that there is a hierarchy of facts, and that a judicious selection can be made. They are right, for otherwise there would be no science, and science does exist. […] This shows us how our selection should be made. The most interesting facts are those which can be used several times, those which have a chance of recurring (Poincaré 1997, 911, our italics).

Now, of course Poincaré’s answer is not satisfactory. That a fact must be repeatable may be a necessary (but surely not a sufficient) feature of its historical relevance, and this feature can help us to distinguish properly historical facts from episodic ones (apax historiae), and from the rhapsodic ones, e.g. from sporadically recurring facts. Reiterations can represent a sort of “constants” in human behaviour like “anthropological invariants”, i.e. invariants beside (physical) transformations. But the rationale for this distinction remains confused and arbitrary, unless a more fundamental criterion that helps us to define what recurrent, episodic, rhapsodic really means. As a borderline case, Poincaré’s insight fixes some limits in attempting to individuate archetypes of historical facts in its general repetitive occurrences. As analytical philosophy has often clarified, facts are not events (Varzi and Casati 2002). More precisely, facts are not individual, unique, unrepeatable, particular, monadic, changes in space and time, or in a “most fundamental” substances (i.e. like objects), but states of affairs (Ramsey 1927; Mulligan and Correia 2013); they are only potentially recurrent and, consequently, “archetypical”. But how to define the boundaries of a fact, its characteristic, its peculiarity, and its informational salience? However, the “Naive Realism” epitomized by Erodotus of Alicanarnassus’ dictum “To tell what happened and what is going on” (ĬĮ İȠȞIJĮ ȜİȖİȚȞ, ta eonta leghein) (Erodotus 1997, I, 1) does not seem to be more convincing. This view avoids any commitment to critical historiography, as invoked by a modern disciple of Erodotus, Leopold Von Ranke: You have reckoned that history ought to judge the past and to instruct the contemporary world as to the future. The present attempt does not yield to that high office. It will merely tell how it really was. To history has been assigned the office of judging the past, of instructing the present for the benefit of future ages. To such high offices this work does not aspire: It

To Be Realist with History: What Really Are Historical Documents?

157

wants only to show what actually happened (wie es eigentlich gewesen) (Von Ranke 1973, 57).

Indeed historiography rules out a not negotiable work of critical activity too. So, a realist option in philosophy of history can settle historical science in a concrete scientific array. The realist argument (broadly understood) follows as a conclusion of the following syllogism: i. Historical Science investigates its, proper, real research-entities. ii. These are, by definition, social objects. ‫ ׵‬The research-entities of Historical Science are, in primis, all the objects whose fields of existence and possibility are investigated by social ontologies.

Our thesis is that historical entities are both facts and events, and that documents are the relevant epistemic connections between historians and these entities. The historiographer analyses documents to grasp facts that are realized and causally instantiated in events. From this point of view we can individuate two basic realist options: Historiographical entities of the Searlean kind: X counts as Y in the context C (John Searle’s Thesis about Collective Intentionality). Ferraris’ theory of social ontology as Documentality.2 In what follows I’ll show that Searle’s thesis is not sufficient to support a relevant and exhaustive analysis of a phenomenon so complex as the emergence of a historical entity. Only a “strong” realist approach as in Ferraris’ theory of Documentality, can serve the purpose of true realist oriented methodology.

3. Three problems for Searle’s theory John Searle defines a status function (SF) as established by a constitutive rule of the form: (SF)

X counts as Y in C.

The proposition contained in (SF) states that a physical object X, e.g. a rectangle of water-marked paper, counts as—let’s suppose—a five euro banknote (Y) in the context of the European Union common currency system (C). According to Searle, the whole social reality derives from the repeatability of this simple basic constitutive rule. The argument provided by Searle (1995, 2012) can be thus summarized:

158

Chapter Eleven

(A) All social and institutional human reality exists and continues to function in virtue of a basic linguistic operation. (B) The operation in (A) is the (SF)-type. (C) All the differences and the complexities in human cultures and their social hierarchical structures are explained (though perhaps only in principle) by the fact that this operation is not restricted to one single field of application. The constitutive rule (SF) is a universal rationale in every human social context, that can be recursively repeated, such that (SF) is applicable to every output of a precedent application. Now, we would like to lay down some observations on the Searlian argument.

A. Problems with the Status Function. Causal reversibility Causal reversibility is the idea that, for a fact to count as “social”, it must be possible to trace back its causal path to the basic physical facts constituting it. As Ferraris (2012a) claimed, the Searlian “constitutive rule” can go against very difficult hurdles. According to Ferraris, one cannot take for granted that, from a reality of electrons and subatomic particles it may be possible to arrive to the immense “invisible reality” of a “social dimension”. If the physical structure of every “brute fact” is really the source of a social fact, then it is not clear how to distinguish between brute facts that can suitably count as social objects from those that don’t. For example, if the graphic designer of US Federal Reserve designed a fake bond-like document, we would surely not say that the fake bond counts as a US Bond (a categorical mistake that would cost us a federal prosecution act!). The best we could say is that he has falsified a token of a standard type, a categorical mistake that would cost us a prosecution. So, to make the transformation of X into Y a perspicuous sort of change, Searle is committed to the existence of Collective Intentionality (CI). Now, even assuming that this transformation is correct, how would it be possible to account for the reverse path from social facts to physical facts? Furthermore: how to behave with social entities that have “dynamical boundaries” as nation, states, terrorist’ attempts to people’s life, battles, technological discoveries? It is not clear how a manifold of entities, events, facts and documents can exit totally unscathed from the “ontic reversibility” of their respective Status Functions. Along with this, the Searlean theory seems to account only for synchronic facts. But what about those cases such as the “beginning of the Iron Age”, or the “Bubonic Plague”? How to define, within the Searlean theory, their spatial and chronological “boundaries” (Smith and Varzi 2000)? The Searlean way

To Be Realist with History: What Really Are Historical Documents?

159

reflects a more general difficulty in finding a plausible account for spacetime boundaries of any event of these sorts, e.g. the “Attempt to the life of Archi-Duke Franz Ferdinand”. This difficulty highlights a crucial problem in the passage from the level Y to the level X of the Searlean constitutive rule. The problem consists in the following: the “Attempt of 1914, June 28, in Sarajevo” cannot be reversed back to a fixed and genuine set of physical entities and events, for a combinatorial explosion of information data would overlap with the same informational data—the X of the (SF) thesis—that are supposed to ground the attribution of a status function. In conclusion, a problem of vagueness seems to emerge here (Evans 1978; see also Gaio 2009).

B. Second problem: recursivity of (SF) in the flow of time When we deal with the diachronic level the crux of the matter causes a real state of confusion. We can consider the example of the emblematic case-histories of some of the most famous archaeological relics which have sadly become victims of pseudoscientific speculations, very popular on Video and Web: the megalithic circle of Stonehenge, the enormous and enigmatic Moai sculptures of Rapa Nui and the sepulchral grave of the subterranean level of Palenque’s Pyramid. As commonly known, all these artefacts and complex of artefacts, since their discovery, have been made object of several explanatory hypotheses by historians, anthropologists, and archaeologists. A fact is irrefutable: whatever the “objectively given in the real beginning” origins of these works, and therefore, the original, remote, (SF) to be attributed to the physical “brute” substrates of their creations, any historical theory that does not describe this process of attribution in an absolutely perfect way would risk to transform these artefacts in real and proper unintentional forgeries of original historical objects. This happens because, pace Searle, every relics of a historical object or event transforms itself, with or without the contribute of a surrounding linguistic platform, in some structure of attributions and, in the notorious and more complex cases, in a document. For example, until today it has been a commonplace to believe that the megalithic circle of Stonehenge was a monumental complex with the double Status Function to be, at the same time, a religious area and an astronomic observatory for solstice and equinox days. Furthermore it has been common to guess that for these Status Functions, Stonehenge had been erected in the ancestral age of the Celtic colonization of Britain. Consequently, the collective intentionality of a great number of historical researchers has “elevated” the whole number of the gigantic granite blocks constituting Stonehenge to

160

Chapter Eleven

this historical (SF) “to be a sanctuary and to be an observatory of Sun”. In the last century, new discoveries made the falsification of this historiographical hypothesis possible, opening a way to the alternative hypothesis according to which the Stonehenge area, which was already existing since Neolithic age—and therefore, well-before the arrival of Celtic tribes from the near Gallia—had been destined by some pre-Celtic population to another particular “(SF)”: a cemetery. But such a discovery implied an enormous consequence on the intentional nature of the social object Stonehenge, because, suddenly, a “contemporary Status Function”, believed to be the same “original Status Function” of Stonehenge, was put aside as an obsolete supposition. Something similar has happened in recent times with the famous legend of sepulchral grave of Maya of Palenque Pyramid (that some lovers of science-fiction hypothesis have guessed to be the representation of a prehistoric astronaut), and with the Moai of Easter Island (extraterrestrial travellers from another world), and with the gigantic picture of Cuzco plain (extraterrestrial spacecraft). In all these cases we know now what all these things were really created for.3 Searle’s theory does not succeed in providing a valid and reliable heuristic apparatus able to solve this sort of problems. Searle’s hypothesis is able, at least, to “pass the practice” to other theoretical fields and other disciplines. Searle could justify his theory observing that it is not an intrinsic problem of Collective Intentionality if we are still unable to know with certainty what Men of Past had in their minds when they designed their artworks and monuments and when they constituted them of some (SF). Such ancient artefacts surely had some, univocal (SF) for their existence, and sooner or later historians and archaeologists will definitely discover it. But, this way, the troubles with this approach grow up increasingly. In fact, how can scientific historiography seriously act in this way? How can a serious discipline engaged in studying the institutional reality of past times be obliged to review and to use social objects—and, therefore, what for Searle is nothing else than a re-negotiable (SF)—with this overfunctional methodology? How Searle’s thesis can, without falling into a vicious circle, make the “definitive” discovery of past societies depend on researches employing, in their turn, other Status Function that, sooner or later, will be the Status Function of another past (since this present too is, just virtually, the past of our future posterities)?

To Be Realist with History: What Really Are Historical Documents?

161

C. Third problem: The repeatability and forgeries of Status Functions Let’s consider now two unforgettable exempla of historical forgery: the Constitutum Costantinii (VII Century BC, discovered by L. Valla in 1440) and the Protocoly Sionskix Mudrecov (The Protocol of Wise Old Men of Zion, late XIX century, Russia, with several, possible, attribution). With regard to the first example, the roman Pope could receive his political (temporal) power on Central Italy and the power to appoint the election of the Roman Emperor and his new land conquests from a donation (the socalled “Constantine the Great’s donation”) in the early IV Century BC. According to the second forgery, created by a squad of secret agents of Zar’s Secret Police Okrana, in the late XIX Century, especially in eastern and central Europe, Hebrews were preparing a formidable series of conspiracies against the legitimate sovereignty of Kings, States, Parliaments, under the appointment of a secret plan approved by a mysterious Council of Old Hebrew Wise Man (directed by the Austrian Socialist Party leader Th. Hertzl, the same Hertzl who was the author of very important writings in defence of the very idea of an independent Hebrew Country in Palestine). All these forgeries were written by professional counterfeiters, as a matter of fact not particularly professional to succeed in deceiving and misleading the culture of philologists such as L. Valla (in the case of Constitutum Constantinii), and the scientific eye of a platoon of historical proof-readers (that we cannot confuse with revisionist historians). Thus, Searle’s Collective Intentionality can be realized by attribution and by means of Status Functions based on false beliefs. As a further remarkable consequence, whole collective groups of several, different societies have believed in, and thus, realized Status Functions indicated by the two historical forgeries we’ve mentioned above. As a terrible historical consequence of this attribution of “faked and bogus (SF)”, in the Western History we have seen historical facts like the rise and the fall of the temporal and secular power of the Roman Catholic Church since the Middle Age (consider, for example, the Dictatus Papae by Pope Gregorius VII as the most important political consequence of the first forgery), or the rise (and the not definitive fall) of anti-Semitism, National-socialist ideology, with the consequent explosion of the Second World War and the diabolic persecution of the European Hebrew people that culminated in the tragedy of the Shoah. So, whole nations, political movements and parties, and collective societies have believed and, in virtue of their false beliefs, realized these (SF) in their collective and individual behaviour.

162

Chapter Eleven

Furthermore, (CI) can, by means of (SF) theory, not only iterate and re-iterate recursively false beliefs of the present time, transforming the same beliefs in real (SF)—which is a trivially true and innocuous consequence of Searle’s theory—but permits the recursive repeatability at the dimensional level of historical documents of “fictitious” (SF), which is not an excusable failure for an exhaustive project in order to find social reality in an ontological theory. Since nobody (no historical agent, no witness or chronicle-writer) creates the historical fact that he wants to leave to his posterities with all the final causal and effective properties of this same historical fact (this is the very relevant phenomenon that we can name praeter-intentionality, in the sense of “unintentional adjunctive creation of intentional acts and scopes”). Still, these observations bring us to consider that Searle’s theory is committed to postulate, more than the simple eventuality, the real necessity by the historical agent to transform a brute artefact into a historical document, and to provide ahead of time. That’s an absurdity, indeed.4 No historical semiophoron (using the philosophical lexicon introduced by K. Pomian 1999, e.g. an important building of the past considered a monument today, or a work of visual art stated as a historical document of History of Art today) in its first intentional aim is “elevated to the (SF)” of a historical document. Even if Ludwig Van Beethoven wanted to create perfect symphonic music for his time, and even if he believed he was to be a historical milestone of a certain musical tradition, this same Beethoven did not want to create “historical documents” as the 7th and 9th Symphony (even if we are, of course, entitled to consider all the Nine Symphonies historical documents of extraordinary importance). In conclusion, a critical limit of Searle’s “realism” is that no “assignment-maker” of (SF) really thinks how “to intend” his own social objects in sight of possible, eventual, forgeries or misunderstandings with the exception of some faker and falsifier. Accordingly, Searle’s theory, even considering the possibility, for what may concern such things as money, false artworks, forgeries, of these misleading use of “attribution of function”, doesn’t explain how it is possible that nevertheless a “mysterious” social dimension proper of “Past’s social objects” can become, thus, an ambiguous function of status (in fact: what are concerning and stating, beyond themselves, such objects whose contents/status’ function is, by definition, “falsifiable ad libitum and ad infinitum”?). The same object X, which counted as Y, in the historical context C, will “re-count” as Y' (with Y'  Y) in a future, different, time, and, therefore, in a historical context C' (with C'  C) and so on. Therefore, all CI’s subjects in every different historical period will risk to believe Y', or Y'', or Y''', as the only, unique,

To Be Realist with History: What Really Are Historical Documents?

163

possible (SF), and with the variation of the historical contexts C, C', C'', and thanks to the well known repeatability of every (SF), this operation risks to be repeated, and, in principle, to be repeated, in theory, ad libitum. We conclude that the Searlean definition of social object cannot work in the field of a (even moderate) realist metaphysics for the ontological foundation of historical sciences, because the Collective Intentionality of historical agents in time Tn can be semiotically wholly different from the (CI) of historiographical agents at the time Tm>n.

4. Documentality approach After the previous excursus, I think that Ferraris’ Theory of Documentality furnishes the only ontology available in today’s debate, which is a reliable candidate to be an efficient philosophical platform for a realist methodology of history. According to this theory, exposed by Ferraris (2012a), social objects can be divided into strict or hard sense-documents, as “acts–inscriptions”, and into soft-sense documents, namely facts-recording. Ferraris had developed theoretical statutes of Documentality in eleven axioms, whereby we report five, which are the most relevant and pertinent to the aims of our research. i. Ontology is distinguished by epistemology. ii. Constitutive Rule for social objects are “Object = scripted Act”. iii. Society doesn’t settle on communication but upon recordings, namely on the creations of scripted acts that record, creating it, social objects’ existence. iv. The mind is a table that gathers inscriptions, it's a “tabula rasa”. The mind is viewed by Ferraris as a recording process, the opposite of the Cartesian sense of “subjective gathering of data”, and more precisely in Derrida’s sense of acquiring and memorizing, and already at the level of animal imitation, namely in the narrow sense of Derrida’s archiscripture. v. Documents, in a strong, narrow, sense are inscriptions of acts. Ferraris’ theory combines perfectly with a realist analysis of historical documents, almost for two reasons. The first reason to endorse Ferraris’ realist theory is that according to Ferraris’ proposal, the stance historical scientist should have regarding historical documents must obey to the axiom that the Nobel Prize J. Monod (1970) named objectivity postulate: the idea that scientific observations could chance upon and bump into a world given in itself and provided for its own, effective, but inertial existence, in autonomy from our capacity and will to know it—which is, namely, the property of not

164

Chapter Eleven

alterability, or in general to be unamendable, that is the ground for the distinction between epistemology and ontology stated by Ferraris in (i). So, we can discover or not discover the historical documents and facts of the past (in the same manner as the historical agents and their chronicle writers can decide to leave them to the posterities), and we can succeed in understanding or deciphering them; but their (past) existence rests, as real facts. This is also the surprise or unpredictability—a factor that is the ground of the realist nature of science, different from what Ferraris calls experience (Ferraris 2012a). The second reason is that, if Ferraris is right in maintaining that a soft sense-document cannot have the same causal weight on the (past or present) reality than a strong-sense document, then the historian, when he is doing his job, can never be in his turn a historical agent, but he could be only an analytical and heuristic descriptor of other people, other times and ages, historical facts and events, and a detective of historical problems linked to the same events. A historical document cannot have any pretension to “act” on the reality at the time of its discovery in the very same way as a historical fact or event could do, because a fact or a historical event is a matter, another matter is the recording of a fact or of a historical event in a historical document, and—finally—another matter is, too, the institutional acknowledgement (which is a further recording) of the same document as circumstantial or legal evidence, also of a political nature. Historical events can be changed thanks to historical changes that are other historical events, and it is impossible sic et simpliciter to change them on the basis of the discovery of historical documents only. The same thing, for our “documental realism”, is to maintain that institution makes history, or, which is namely the same fact, that they (institution and history) are institutional facts, whereas their discovery is not an “institutional fact”. Historical documents, to count as historical facts (versus Searle’s idea) must be institutionalized as legal, political, civil acts, such as, namely, The Memory Day (January 27th in Europe), or the Italian Liberation Day (April 25th). For this reason, the notification to Pope Pius IX of the falsity of Constitutum Costantinii has never been sufficient to persuade Pope Pius X of the anti-historical nature of Doctrine of Temporal Power. It is well-known that the breach open near Porta Pia in September 20th 1870 played a causal, much more incisive role (but not even decisive, maybe, in Italy). Vice versa, Constitutions and Declarations could be acknowledged and recognized as historical documents in their act of birth, because their Author wanted them properly and namely as written acts (if they are serious people) (Fuhrman 1968; De Leo 1974).

To Be Realist with History: What Really Are Historical Documents?

165

We can summarize this point stating that the legal-institutional value of a document is never intrinsic, but it’s always the result of a (further) social and institutional acknowledgement act.

5. Conclusions: two types of historical documents, two axioms for Historical Realism The most basic claim for a “realist” historiographer is, basically, the possibility to make (and to consider this act of making) genuine discoveries able to bring real, objective events, facts, objects (such as past documents) from the not known dimension of “Ontological Unknown” to the well known level of a discovered reality. This way, a historiographer can not only find some issues for typical, epistemological, problems giving a historiographical explanans to one (or more) historical explanandum (or explananda), but he can also report to the light of the present such things as social objects of past—from remote past or recent past. Archaeologists who make ancient buried towns come back to daylight don’t define the artefacts they discover as “documents of the past”, but they simply discover them, giving attention to something that was socially recorded, by means of an “inscription of an act”, in the same past eras in which those artefacts were realized and provided with their social values. Therefore, the sense by which an artefact of the past provides documentary evidence, is quite different from the sense by which (by means of a signature, a stamp or a seal) it provides documental recording, giving certification and authentication to an individual’s identity or to the juridical value of a document. So, we must underline a new urgency in our analysis: the need to classify the type of documents depending on the strength of the degree of intentionality, “general intentionality” and unintentionality of references and documents, or on the strength of the degree of intentionality. 1. Autoylemorphical documents. They are those actual documents that are not “materially born” as historical documents. Nevertheless, they acquire the informational structure or form (morphè) of historical documents only at a later stage, in an evolutionary way, by means of an exaptation (Gould and Vrba 1982, 4-15) with a not-completely-intentional or praeterintentional process of transformation of the same matter (hyle) whereby they are structured and created. They are recognizable as “autoylemorphical” in virtue of an extrinsically scriptural-credit intentionality, i.e. by means of an exogenous convention whereby a semiotic community established to attribute specific meanings and references to them.

166

Chapter Eleven

2. Eteroylemorphical documents. They are all those documents which are created ad hoc by an almost material, intentional, author (historiographer, chronicle-writers, publicists, narrators, writers, artistic descriptor or critics, etc.) and they are provided for a documental intentionality that, varying from a maximum level of intentionality (e.g. the voluntary way of historical texts of historiography) to a minimum level of such eteroylemorphical intentionality (the involuntary way of a ready made type of objects as monuments). They are also expressions or structured clues of one or more communicative and recording aims, which are deliberately founded in an intention of reference to, at least, an event, a fact, an action, an (historical) agent. A very good example of eteroylemorphism is a wide, very large, subset of Internet, World Wide Web, with the so-called Internet 2.0. The latter provides the privileged recording platform where all these forms of documental eteroylemorphism can be kept together and included. In fact, whenever some user links (in or out of a hypertext) an autoylemorphical document inside the platform, in every case this same user acts this way to “record” his presence by attribution of documental type. This one will be imposed and fixed from the external world to the same matter of the document. (The most evident example is the Time Line of notorious social networks such as Facebook or Twitter). Finally, we can sum up the pivotal claims of ontological documental realism for historical science in two ontological and methodological principles: 1. Causal Inertia between historical facts and historiographical documents. By this, I mean that nothing else but a historical fact or a historical event can interact with, respectively, a historical fact or a historical event; documental descriptions, intrinsically considered, have not this causal power and the causal inertia of historical facts and events results to be impenetrable to them, unless a document, a testimony, or a historical report or relic has been recognized (at a higher social and institutional level) a superior “de-ontic power”, as only a legal or political argument can be. 2. Ontological humility (it follows as a corollary of the previous claim). So, the historical meaning of a document is always the result of a (further) social and institutional act of acknowledgement of the particular content of that document. 1. and 2. form the formal architecture to which a realist theory of historical facts and documents must conform.

To Be Realist with History: What Really Are Historical Documents?

167

References Ankersmit, F.R. 1995, Language and historical experience, Bielefeld: ZiF Ankersmit, F.R. and H. Kellner (eds.) 1995, A new philosophy of history, Chicago: University of Chicago Press Bloch, M. 1993, Apologie pour l’histoire ou metier d’historien. Paris: A. Colé Editeur Carr, E.H. 1961, What is History?, London: MacMillan Danto, A.C. 1968-85, Narration and Knowledge. Analytical Philosophy of History, New York: Columbia University Press Davidson, D. 1963, “Actions, Reasons, and Causes”, Journal of Philosophy 60 (23), 685 De Leo, P. 1974, Ricerche sui falsi medioevali: I-Il Constitutum Constantini, compilazione agiografica del sec. VII. R. Calabria Donagan, A. 1966, “The Popper-Hempel Theory Reconsidered”, in Dray, W.H. (ed.), Philosophical Analysis and History, New York: Harper & Row, 127 Dray, W. (ed.) 1966, Philosophical analysis and history (Sources in contemporary philosophy), New York: Harper & Row Dray, W. 1964, Philosophy of history, Englewood Cliffs, NJ: PrenticeHall Dummett, M. 2004, Truth and the Past, New York: Columbia University Press Erodotus 1997, Storiae, (ǿıIJȠȡȚĮȚ), Rome: Newton & Compton Evans, G. 1978, “Can There Be Vague Objects”, Analysis 38, 208 Ferraris, M. 2012a, Documentality. Why is necessary to leave traces, New York: Fordham University Press —. 2012b, Manifesto del nuovo realismo, Roma-Bari: Laterza (German transl. Frankfurt a.M.: Klosterman 2014) Fuhrman, H. 1968, Das Constitutum Constantini (Konstantinische Schenkung), Hannover: Monumenta Germaniae Historica, Fontes iuris Germanici antiqui, 10 Gaio, S. 2009, Vaghezza, in the webpage Temi, of APhEx; Analytical and Philosophical Explanations, www.aphex.it Gould S.J. and E.S. Vrba 1982, “Exaptation. A missing term in the science of form”, Paleobiology 8, 4 Hacking, I. 2002, Historical ontology, Cambridge, MA, London: Harvard University Press Hempel, C.G. and C. Oppenheim, 1948, “Studies in the Logic of Explanation”, Philosophy of Science 15, 135

168

Chapter Eleven

Kremer, M. 1993, “Population Growth and Technological Change; One Million b.C. to 1990”, Quarterly Journal of Economics 108 (3), 681 Little, D. 2010, New Contributions to the Philosophy of History, Dordrecht: Springer Science Monod, J. 1970, L’Hazard et la Necessitè, Paris: Gallimard Mulligan, K. and F. Correia 2013, “Facts”, The Stanford Encyclopedia of Philosophy (Spring 2013 Edition), Edward N. Zalta (ed.), URL = http://plato.stanford.edu/archives/spr2013/entries/facts Neale, S. 2001, Facing Facts, New York: Oxford University Press Nietzsche, F. 1887-1888, 7 [60] (1967) Posthumus Fragments I-IV in Colli, G. and M. Montinari, Friedrich Nietzsche Digital critical edition of the complete works and letters, based on the critical text, Berlin/New York: de Gruyter, (ed.) by P. D’Iorio in http://www.nietzschesource.org/#eKGWB/NF-1887,8 Pirandello, L. 1921, Six Characters in Search for an Author, English version by Edward Storer, New York: E. P. Dutton, 1922, with a preface by the author Pomian, K., 1999, Sur l’histoire, Paris: Gallimard Pompa, L. 1990, Human nature and historical knowledge: Hume, Hegel, and Vico, Cambridge, New York: Cambridge University Press Poincarè, J.H. 1909, “The Choice of Facts” in The Monist, XIX, 232 Repr. in The Value of Science, 1958, Eng. tr. by G.B. Halsted, New York: Dover Popper, K.R. 1957, The Poverty of Historicism, London: Routledge Ramsey, F.P. 1927, “Facts and Propositions”, Proceedings of the Aristotelian Society, Supp. 7, 153 Rorty, R. 1979, Philosophy and the mirror of nature, Princeton: Princeton University Press Searle, J. 1995, The Construction of Social Reality, New York: Free Press —. 2010, Making the Social World The Structure of Human Civilization, Oxford: Oxford University Press Smith, B. 2002, “Fiat Objects”, Topoi, 131 Smith, B. and A.C. Varzi 2000, “Fiat and Bona Fide Boundaries”, Philosophy and Phenomenological Research 60, 401 Valla, L. (ed. W. Setz) De falso credita et ementita Constantini donatione (Weimar, 1976: Monumenta Germaniae Historica, Quellen zur Geistesgeschichte des Mittelalters 10) Varzi, A.C. and R. Casati 2002, “Un altro mondo?”, Rivista di Estetica 19 (1), 131

To Be Realist with History: What Really Are Historical Documents?

169

Vaselli, S. 2012, “Due ontologie della realtà storica. Documentalità e intenzionalità collettiva alla prova della storicizzazione”, Rivista di Estetica, 50 Von Ranke L. 1973, “Preface: Histories of the Latin and Germanic Nations from 1494-1514”, in Stern, The Varieties of History, 57 —. 1881, The Theory and Practice of History, in W. Humboldt (ed.) 1973, The European Historiography Series, Indianapolis: Bobbs-Merrill White, H. 1973, Metahistory, Baltimore, MD: Johns Hopkins University Press

Notes 1

Thomas Gradgrind is a character from Charles Dickens’s novel Hard Times, notorious for his repetitive mantra: “Facts, facts, beat the facts”; see also Carr, 1961, Chap. I. 2 Truly speaking, it’s possible to insert another theoretical solution: Barry Smith’s Thesis about the boundaries between natural and artificial entities as bona fide/fiat boundary. For this option social entities as historical entities are Fiat Object. 3 With the only exception of Cuzco Plain, which is still object of several, plausible, possibilities. 4 In fact, every time it happens that men or women in-the-events-of-history want to elevate a brute artefact to the level of a social object as “historical document”, the institutional ground of what those men or those women are actually creating is, in turn, an intrinsically historic object.



ABSTRACTS

Mario Alai, Why Antirealists Can’t Explain Success Scientific realism explains the success of novel predictions by assuming that theories are at least partially true accounts of the unobservable reality. Here “novel” means not used essentially in constructing the theory, heterogeneous to previous observations and a priori improbable. I argue that no alternative antirealist explanation is possible: van Fraassen’s evolutionary analogy explains why we don’t have unsuccessful theories, not why we have some successful ones. His empirical adequacy and Fine’s surrealism (i.e., that theories are compatible with all the observable data) do not entail that they predict anything, let alone new phenomena. Hacking’s deflationist explanation, that success is already explained by theories themselves, presupposes the truth of theories. Saying that theories have true novel consequences because they are instrumentally reliable (Fine, Lyons), or predictively similar to the true theory (Stanford) is like saying that they have lots of true consequences, including many novel ones. But this repeats the explanandum rather than explaining it. Moreover, realists have a plausible account of how partially true theories can be found, but antirealists cannot explain how theories with true novel consequences can be found (except by looking for partially true theories): by inductive extrapolation one can only predict observations homogeneous to the old ones.

Petar Bojaniü, How to Incorporate (with) Documents In the paper I address two questions concerning John Searle's account of institutions and corporations. I therefore focus on the relationship between institutions and documents, and on the corporation as a prototypical case of institution. Finally, I outline a theory of the relationship between institutions and documents, which I will call “institutional realism”.



Metaphysics and Ontology Without Myths

171

Andrea Bottani, The Myth of the Distinction between Ontology and Metaphysics It is a widely shared idea that no accurate account can be given of what was originally called “first philosophy” that does not distinguish the theory of what there is, sometimes called “ontology”, from the theory of what it is, sometimes called “metaphysics”. Is the distinction well grounded? I answer in the negative by arguing that 1) ontology is a systematic taxonomy of what exists; 2) no apparent disagreement about the nature of the entities belonging to some taxonomical category counts as a real disagreement about the ultimate nature of something, provided there is total agreement about the nature of what belongs to the highest categories, those that restrict no other category; 3) every disagreement about the nature of what belongs to the highest categories is eo ipso a disagreement about what there is. As a result, metaphysics and ontology are no more distinct than the morning star and the evening star. The distinction is flawed and should be given up. It is just a myth. I conclude by saying something on the bearing that the refutation of the distinction can have on the very notion of a “first philosophy”.

Roberto Casati and Giuliano Torrengo, The Identity of Indiscernibles and the Principle of No Co-location In this paper, we propose a revised version of Black’s original argument against the principle of identity of indiscernibles. Our aim is to examine a puzzle regarding the intuitiveness of arguments, by showing that the revised version is clearly less intuitive than Black’s original one, and appears to be unjustified by our ordinary means of assessment of intuitions.

Richard Davies, Why Society Is Not Made of Documents The central thesis of Maurizio Ferraris’ Documentality is that every social object is constituted by an inscribed act. One of the main strengths of the theory of documentality is that its privileged objects are readily identifiable, paradigmatically as pieces of writing in some language on some physical support. In this respect, Ferraris’ theory has an advantage over its main rival as a theory of social objects, namely Searle’s appeal to physical objects “counting as” social objects on the basis of “collective intentionality”. Ferraris makes at least tacit appeal to the idea that, in any explanatory sequence involving social reality, the final term must be an



172

Abstracts

inscribed act. But this leaves it unclear both what it is for an inscribed act to be valid and how its content is to be determined. For, both validity and (linguistic) content appear to derive from social practices that are not themselves inscribed acts. Moreover, Ferraris makes use of the Derridean notion of “archiwriting” to cover the cases of social objects for which no writing in any actual language can be found on any identifiable physical support. Given that this notion is a muddle, invocation of it undermines the theory that it is meant to help.

Maurizio Ferraris, New Realism, Documentality and the Emergence of Normativity This paper sets out three fundamental theses. The first (New realism) is the thesis that there is a fundamental layer of reality that not only exists independently of the mind, but has causal powers on the mind. The second (Documentality) proposes an ontology that accounts for the construction of social reality with a bottom-up process whose very principle is not mind, but recording. The third (Emergence) asserts that intentionality and normativity are a product of documentality.

Matteo Plebani, The Content Challenge to Nominalism About Mathematics Nominalists think that there are no mathematical objects. According to the received wisdom, our best scientific theories contain claims (like “The atomic weight of beryllium is 9.012182”; or “there is a function mapping space time points into the real numbers” etc.) that entail that there are mathematical objects. It seems therefore that a nominalist cannot accept the whole content of our best scientific theories. The problem for her is to specify which part of these theories she retains (the so called “nominalistic content” of the theory). Some of the “easy road” nominalists presented by Colyvan (2010) think they can express the nominalistic content by just subtracting the unwanted implications of the original theory. This strategy has recently been charged of rendering the nominalistic content obscure (Azzouni 2009, Colyvan 2010, Raley 2012). Call this objection “the content challenge to easy road nominalism”. After presenting the content challenge in general terms, I discuss the way Yvonne Raley (2012) uses it to argue that Jospeh Melia’s nominalistic account of mathematics is untenable. I present then some replies to Raley’s objection inspired by the work of Stephen Yablo (2012, MS).



Metaphysics and Ontology Without Myths

173

Antonio Rainone, Realism and Truth from a Quinean Point of View This paper is intended to reconstruct W.V. Quine’s views on realism and truth. Ontological and naturalistic Quinean conceptions appear the key topics for approaching these views. The central thesis here defended is that Quine’s philosophical perspective is not a form of instrumentalism, as it has sometimes been considered, but a form of internal realism based on his naturalism. According to Quine, realism and truth are not at all metaphysical concepts, as are the correspondence theory of truth and the belief in a reality an sich that our theories should mirror; they are simply natural concepts, intrinsic to our way of considering the external world through our most deeply rooted ontological beliefs of common sense and the best confirmed scientific theories. According to Quine, there is no “cosmic exile” independent of our epistemic categories and procedures from which we could judge the truth of our beliefs as a correspondence with an independent reality. Truth and the reality are necessarily mediated by our conceptual scheme. Nevertheless, we cannot do other than call ourselves realists.

Claudine Tiercelin, Is There Such a Thing as Metaphysical Knowledge? The aim of the paper is to examine some arguments in favor of the possibility of genuine metaphysical knowledge. Starting from some arguments in favor of humility, it proceeds to show how contemporary metaphysicians have often pursued a similar approach and that some Humean and Kantian arguments are not that far from D. Lewis’s arguments for humility. After showing the epistemological and metaphysical drawbacks of such an attitude, the paper presents another option, which consists in following a deliberately metaphysical stance, along dispositional realistic lines. After presenting some of its characteristics, it shows why dispositional realism is a better account than some variants of ontic and causal structuralism and why it allows us to entertain a reasoned form of boldness, and claims a genuine metaphysical knowledge of things, of their real nature and of their properties.





174

Abstracts

Achille Varzi, Realism in the Desert Quine’s desert is generally contrasted with Meinong’s jungle, as a sober ontological alternative to the exuberant luxuriance that comes with the latter. Here I focus instead on the desert as a sober metaphysical alternative to the Aristotelian garden, with its tidily organized varieties of flora and fauna neatly governed by fundamental laws that reflect the essence of things and the way they can be, or the way they must be. In the desert there are no “natural joints”; all the boundaries we find are lines we have drawn, artificial fencings that merely reflect of our own demarcations, our classifications, our desperate need to ward off the flux and meet an excusable but ungrounded demand for order and stability. The desert returns a picture of reality that is radically anti-realist. And yet the picture does not amount to a form of irrealism. The desert is out there and is what it is regardless of how we feel. And it is not completely structureless. It’s just that the structure it has is very thin and does not correspond to the sort of structure that so-called metaphysical realists— and scientific realists alike—tend to attribute to it.

Stefano Vaselli, To Be Realist With History: What Really Are Historical Documents? Realist philosophers about history claim that entities such as documents, their sources, and events are objectively given in the reality of social history, over and above our capacity to grasp them. How is it possible to provide an account of the ontological independence of historical findings? If there is any (mis)interpretation of those findings, where does it come from? A realist of this sort is committed to provide a suitable theory to solve the problem of the reliability of historical findings. In this paper I analyse how such realist views, like Searle’s Collective Intentionality, cannot provide the solution. A different form of realism is needed. One proposal is the Documentality Theory defined by Maurizio Ferraris. This theory satisfies the test of historical realism, because its definition of a social object as an inscription of an act is able to distinguish between what a fact/event/object X of the past is, and which features X must possess in order to count as a document. Furthermore, this approach is able to discern, this way, voluntary from unintentional effects of an object/event/fact, how a social fact is instantiated as social fact indeed, which are the causes of historical events, and their different degrees of historicity.





CONTRIBUTORS

Mario Alai, Ph.D., is Associate Professor of Theoretical Philosophy at the University of Urbino (Italy). After graduating at the University of Bologna, he received postgraduate degrees from the Universities of Urbino, Helsinki, Maryland and Florence. He taught education, psychology, philosophy and history in high schools, and was Researcher in Philosophy of Language at the University of Urbino. He has written A Critique of Putnam’s Antirealism (U.M.I., 1989), Modi di conoscere il mondo (Angeli, 1994), Filosofia della Scienza del Novecento (Armando, 1998). Petar Bojaniü is a director of the Institute of Philosophy and Social Theory (Belgrade) as well as the Centre for Advances Studies (Rijeka). He is the author of several books and co-editor of World Governance (Cambridge Scholars Publishing, 2010). Andrea Bottani is Full Professor of Theoretical Philosophy at the University of Bergamo (Italy). He is author of Il riferimento imperscrutabile (Angeli, 1996) and editor of Individuals. Essence and Identity (Kluwer, 2003, with Carrara and Giaretta) and Modes of Existence (Ontos Verlag, 2006, with Davies). He published in a number of journals, among which Dialectica and The Monist. Roberto Casati, philosopher, is a CNRS Directeur de Recherche at the Institut Nicod, Ecole normale supérieure, Paris. He has published Holes and other Superficialities (MIT Press, 1994) and Parts and Places (MIT Press, 1999), both with Achille Varzi. His Shadows (Knopf, repr. Vintage, 2002) has been translated into eight languages. His latest work is Contre le colonialisme numérique (Albin Michel, 2013). Richard Davies is Researcher in Theoretical Philosophy at the University of Bergamo (Italy). He has written books on Descartes and survey of symbolic logic as well as numerous articles on logic, metaphysics, ethics and the history of philosophy. He is currently preparing a text on “paraphilosophy”.



176

Contributors

Maurizio Ferraris is Full Professor of Philosophy at the University of Turin, where he is also the director of the LabOnt (Laboratory for Ontology), and fellow of Käte Hamburger Kolleg “Recht als Kultur” (Bonn). He wrote more than forty books that were translated into several languages, among which History of Hermeneutics (Humanities Press, 1996), Documentality or Why it is Necessary to Leave Traces (Fordham UP, 2012) and Goodbye Kant! (SUNY UP, 2013). His latest books to be published in English are Where Are You? An Ontology of the Cell Phone (Fordham UP, 2014) and Manifesto of New Realism (SUNY UP, 2014). Matteo Plebani received his Ph.D. in Philosophy from Ca’ Foscari University of Venice (Italy) with the thesis Reconsidering Wittgenstein’s Philosophy of Mathematics. He is mainly working on nominalist and fictionalist strategies in the philosophy of mathematics and on deviant approaches to meta-ontology. He is the author of Introduzione alla filosofia della matematica (Carocci, 2011). Antonio Rainone is Associate Professor of Logic and Philosophy of Science at the Università di Napoli “L’Orientale” (Italy). His research and publications deal with issues in analytic philosophy (action theory, rationality, and epistemology). His most recent works are Quine (Carocci, 2010) and Quale realismo, quale verità: saggio su W.V. Quine (Quodlibet, 2012). Claudine Tiercelin is a French philosopher. Since 2010 she is Professor at the College of France, where she holds the Chair of Metaphysics and Philosophy of Knowledge. She is also a member of the Institut Jean Nicod. The works of Claudine Tiercelin build on the heritage of European and American analytic philosophy, American pragmatism, and the tradition of French rationalism. She defends a scientific metaphysics, both realistic and rational, in which real dispositions, viewed as causal powers, play a central role. Her most recent books in metaphysics are: La Connaissance métaphysique (her inaugural lecture, delivered in Paris in 2010 and published at Fayard Publ., 2012) and Le Ciment des choses: petit traité de métaphysique scientifique réaliste (Éditions d’Ithaque, 2011). Giuliano Torrengo is Researcher of philosophy at the University of Milan (Italy). He has published two books: I viaggi nel tempo. Una guida filosofica (Laterza, 2011) and Time and Cross-temporal Relations (Mimesis, 2008). His publications in journals include articles in Synthese, Analysis, and Metaphysica.



Metaphysics and Ontology Without Myths

177

Achille C. Varzi is Professor of Philosophy at Columbia University, New York. Among his books: Il mondo messo a fuoco (Laterza, 2010), Ontologia (Laterza, 2005), Parole, Oggetti, Eventi (Carocci, 2001), An Essay in Universal Semantics (Kluwer, 1999), Parts and Places (with R. Casati, MIT Press, 1999), and Holes and Other Superficialities (with R. Casati, MIT Press, 1994). Stefano Vaselli, Logic and Epistemology Ph.D, Teaching Assistant at University of Rome “La Sapienza” (2009-2011). He teaches philosophy and history at MIUR (Italian Department for Research and Education). Research Areas: Realism in Philosophy of Social Sciences, Social Ontology, Economics of Cognition. Memberships: APhEx–Portale di filosofia analitica, www.aphex.it, Res Cogitans, www.rescogitans.it.





INDEX OF NAMES

Abbot F., 91 Adams R.M., 103, 106 Alai M., 48-49, 61-62 Ankersmit F.R., 153 Aristotle, 1, 17, 30, 34, 84, 89, 96 Avicenna, 89 Azzouni J., 68, 72-73, 75, 77, 81 Barnes E.C., 57, 59, 66 Beebee H., 86 Beuchot M., 123 Bird A., 88, 92 Black M., 101-106, 108-109 Bloch M., 155 Borges J.L., 5, 10 Bradley F.H., 26 Brandom R., 124 Briet S., 142 Broad C.D., 17, 30 Calosi C., 63 Calvino I., 73 Carr E.H., 155-156, 169 Casati R., 156 Chakravartty A., 93, 95 Clark M.J., 77 Colyvan M., 67-69, 73-74, 76 Caruso G., 47 Cordero C., 63 Correia F., 156 Cross C., 106 Curzon G.N., 20, 31 Da Costa N., 93 Dante, 134 Danto A., 153 Darwin C., 23, 34 Davidson D., 31, 154 Dawkins R., 123



De Leo P., 164 Della Rocca M., 105 Dennett D., 117 Derrida J., 137, 163 Descartes R., 30 Dewey J., 32, 44 Dickens C., 169 Donagan A., 153 Doppelt G., 59 Dorr C., 73, 76-77 Dray W., 153 Drewery A., 88 Dummett M., 154 Duns Scotus, 30, 89-90 Einstein A., 34 Elgin C., 23 Ellis B., 88, 92, 99 Erodotus of Alicanarnassus, 156 Esfeld M., 93-94, 96 Eusebius of Cæsarea, 135 Evans G., 159 Fano V., 63 Ferraris M., 27, 123-135, 137-139, 141, 150, 157-158, 163-164 Feyerabend P., 31 Fichte G., 123 Field H., 67, 73, 76 Fine A., 40-41, 50, 52, 55, 58, 62, 81 Frege G., 15-16, 24 French S., 93-94, 103 Fuhrman H., 164 Fustel de Coulanges N.D., 150 Gabriel M., 123 Gaio S., 159 Garavaso P., 63

Metaphysics and Ontology Without Myths Gibson J.J., 123 Goodman N., 24-26, 31, 123 Gould S.J., 165 Groebner V., 141 Hacking I., 55, 103, 154 Hawley K., 109 Hawthorne J., 89, 94 Heidegger M., 124 Heil J., 94 Hempel C.G., 153 Hertzl T., 161 Hoefer C., 47, 63 Hume D., 18-19, 28, 83-84, 144

179

Lyons T.D., 51, 66 Macchia G., 63 Maher P., 57, 59 McLuhan M., 119 Meillassoux Q., 123 Meinong A., 16-17 Melia J., 67-73, 76-78 Mendeleev D., 49 Molnar G., 88 Monod J., 163 Morganti M., 63 Mulligan K., 156 Mumford S., 88, 92, 95 Musgrave A., 52

Ingarden R., 2 Jackson F., 85 Jerez J.L., 123 Jünger E., 120, 124 Kant I., 84, 110, 124 Kellner H., 153 Kelly K., 124 Kepler J., 34 Kitcher P., 63 Kittler F., 119, 124 Kneale W.C., 30 Kripke S., 3 Kuhn T., 62 Ladyman J., 93 Lam V., 93-94, 96 Langton R., 84-85 Laudan L., 61 Le Goff J., 150, 155 Leibniz G.W., 84 Leplin J., 51 Le Poidevin R., 1-3 LeĞniewski S., 110 Lewin K., 123 Lewis D., 8, 13, 79, 84-86 Liggins D., 70, 72-73, 77 Little D., 153 Locke J., 83, 96 Lowe E.J., 96



Nerlich G., 102 Neurath O., 35, 44 Newton I., 34-35 Nicholas of Cusa, 135 Nietzsche F., 124, 155 Niiniluoto I., 63, 66 Occam W. of, 16, 18 O’Connor D.J., 105 O’Leary-Hawthorne J., 109 Oppenheim C., 153 Otlet P., 148, 150 Panza M., 63 Peacocke C., 83 Peirce C.S., 42-43, 90, 92, 96, 100 Perry J., 16 Plato, viii, 16, 28, 113, 125 Poincaré H., 33, 156 Pomian K., 162 Pompa L., 154 Pope Gregorius VII, 161 Pope Innocent IV, 146 Pope Pius IX, 164 Popper K.R., 53, 153 Prien B., 142, 150 Psillos S., 52, 66, 93-95 Ptolemy, 34 Putnam H., 25, 31, 42-43, 47-48, 56, 123

180

Index of Names

Quine W.V.O., 2-3, 16-19, 26, 28, 30, 32-45 Raley Y., 69-72, 78-79 Ramsey F.P., 156 Raposa M., 90-91 Rees P., 50, 61-62 Rescher N., 13 Rickert H., 123 Rodriguez-Pereyra G., 109 Rorty R., 24, 30, 38, 153 Rosenberg A., 47 Ross D., 93 Ryle G., 126 Scarpa R., 123 Schmitt F., 120 Schurz G., 53, 66 Searle J., 126-127, 129-131, 139140, 142-143, 146, 148-150, 157-162, 164 Sellars W., 87 Shoemaker S., 92 Sider T., 8 Skudlarek J., 150 Smart J.J.C., 48 Smith B., 123, 139-141, 158, 169 Sokal R.R., 30 Spengler O., 124



Stamos D.N., 23, 30 Stanford K.J., 51, 62, 66 Stolte S., 150 Tichy P., 53 Tiercelin C., 83, 87, 92 Tolkien J.R.R., 75 Valla L., 135-136, 161 van Fraassen B., 50, 62-63 van Inwagen P., 1-2, 13, 16 Varzi A.C., 1-3, 6, 10, 30-31, 156, 158 von Neumann J., 6 Von Ranke L., 156-157 Vrba E.S., 165 Weingartner P., 53, 66 White H., 153 White R., 50, 57-58, 66 Whittle A., 84-87 Wittgenstein L., 117, 123 Wright C., 43 Wright J., 49, 61 Yablo S., 67, 72-75, 77-79 Zalta E.N., 6 Zermelo E., 6