Facets of Sociality 9783110326932, 9783110326598

Citation preview

Nikos Psarros, Katinka Schulte-Ostermann (Eds.) Facets of Sociality

Philosophische Analyse Philosophical Analysis Herausgegeben von / Edited by Herbert Hochberg • Rafael Hüntelmann • Christian Kanzian Richard Schantz • Erwin Tegtmeier Band 15 / Volume 15

Nikos Psarros Katinka Schulte-Ostermann (Eds.)

Facets of Sociality

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

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

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

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

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

2007 ontos verlag P.O. Box 15 41, D-63133 Heusenstamm www.ontosverlag.com ISBN 10: 3-938793-39-2 ISBN 13: 978-3-938793-39-8

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

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

Index Introduction

PART I: SOCIAL ONTOLOGY Barry Smith, Saarbrücken/Buffalo The Foundations of Social Coordination: John Searle and Hernando de Soto

I

1 3

Nikos Psarros, Leipzig The Constitution of Social Objects by Common Actions

23

Arto Siitonen, Helsinki Kosmos Noetos and Carnap’s Constitution of Cultural Objects

33

Wolfgang Detel, Frankfurt Mental Causation and the Notion of Action

51

Pirmin Stekeler-Weithofer, Leipzig From Individual Mind to Forms of Human Practice

85

Jakob Lindgaard, Warwick The Logic of Mind-Talk – Comment on Stekeler-Weithofer’s From Individual Mind to Forms of Human Practice

117

Ingvar Johansson, Saarbrücken How do non-Joint Commitments come into Being? An Attempt at Cultural Naturalism

135

PART II: COLLECTIVE ACTION

151

Margaret Gilbert, Storrs/CT Acting together, Joint Commitment, and Obligation

153

Raimo Tuomela, Helsinki Joint Action

169

Frank Kannetzky, Leipzig Levels of Collectivity

209

Boris Hennig, Saarbrücken Social Facts Explained and Presupposed

243

Richard Raatzsch, Potsdam “We intend …”

265

Hans Bernhard Schmid, St. Gallen/Basel On Not Doing One’s Part

287

PART III: EPISTEMIC HOLISM

307

Deborah Tollefsen, Memphis/TN Collective Epistemic Agency and the Need for Collective Epistemology

309

Lars Lundsten, Helsinki Epistemology of Holes

331

Bernt Österman, Helsinki Interpretation, Understanding, and Application

355

ABOUT THE AUTHORS

369

Introduction This volume comprises talks given to the workshop Holistic Epistemology and Theory of Action that was organized by Professor Pirmin-StekelerWeithofer, the Forum Analytische Philosophie of the German Philosophical Society (DGPhil), and the editors in June 2004 at the University of Leipzig, and papers that were prepared afterwards by participants of this workshop as a reaction or in response to some of the speakers. The aim of the workshop was to explore new approaches to the problem of the constitution of the various aspects of sociality and to confront these with received ideas. Therefore many of the contributions to this volume are devoted to a rather holistic and antireductionist conception of social objects, groups, joint actions or collective knowledge. Holistic approaches are especially concerned with the question of whether and how groups can be subjects of actions and bearers of knowledge arguing that the actions and knowledge of groups cannot be reduced to individual actions or knowledge, while antireductionist approaches aim at showing that the mental and cultural vocabulary is not reducible to the vocabulary of the natural sciences so that the ontological claim that only the objects of natural science exist is either false or at least impossible to prove. These approaches mark the difference to a recently published volume on Social facts and Collective Intentionality edited by Georg Meggle.1 However, even if several contributors to this volume are proponents of the thesis of the irreducibility of the social fundament of personal life, the volume also includes authors defending the received individualist or reductionist view, namely Raimo Tuomela, Wolfgang Detel, Jakob Lindgaard and in some aspects Margaret Gilbert. The first section is devoted to the question of the ontological status of social objects and to their relation to physical objects. Barry Smith takes up his discussion with John Searle and claims that social objects do not need to be bound or even in principle to be reducible to physical objects. Nikos Psarros agrees in general with Smith’s approach but criticizes that both Smith and Searle are not paying enough attention to the constitution of so1

G. Meggle, Social Facts and Collective Intentionality, Frankfurt 2002.

II

cial objects as being depending on common actions and therefore on intelligent agents, who live and act in and as a social community. Arto Siitonen’s paper takes up the question of the relation between the cultural, the mental and the physical as inquired by Rudolf Carnap. Siitonen shows that Carnap tries to develop an intermediate account of this relationship that is neither reductionist in the sense, that cultural or mental entities are in principle identical with physical ones, nor defends the thesis that the cultural, mental and physical realms are from the ontological point of view total independent from each other. The three following papers are dealing with the mind-body-problem. While Wolfang Detel defends an emergence-theoretical approach with respect to the mental that is founded in the idea of a teleological functionalism, argues Pirmin Stekeler-Weithofer in favour of the fundamental irreducibility of the mentalist vocabulary, claiming that an approach to the understanding of the human life form depends on the irreducible notions of, for example, consciousness, ego, rationality and apperception. Jakob Lindgaard attacks Stekeler-Weithofer’s account for leaving two important issues open namely the question about a sufficient criterion supporting the claim that one can infer from the existence of the mental vocabulary the ontological independence of norms, and the possibility of interaction between the physical and mental. In the last paper of this section Ingvar Johansson claims in contrast to Psarros and Stekeler-Weithofer that plural subjects sharing we-intentions and we-beliefs can exist even without a shared language. He calls them “purely traditionalistic plural subjects” regarding the concept of this kind of social groups as capable to give an understanding of how the ability of modern humans to join groups and be part of joint activity could have evolved from the language free interaction of non-speaking primates. The second section of this volume is devoted to the topic of collective activity. It starts with the quite well known more individualistic accounts of Margaret Gilbert and Raimo Tuomela. While for Gilbert the notion of joint commitment as being the glue between the individual intentions and actions is central, Tuomela’s account is focused on the content of the intentions and beliefs of the participants of a joint action. Both give in their papers a summary of their so far developed theories and try to refute several

III

objections. Some of these objections are taken up by Frank Kannetzky and Boris Hennig. Kannetzky directly criticizes the individualist approach for neglecting the social background of collective action as a necessary condition for the evaluation of individual behaviour. The contents of intentions, he argues, are always action types and action types are standards or norms, on which action tokens are evaluated. Thus the success conditions of individual or collective actions cannot be mere private, they are always bound to the third person perspective. Therefore a purely individualistic account either of individual or of collective actions is not possible, but always presupposes the existence of common practice. Hennig, on the other hand, claims that a non-circular approach to collective action is possible. He is sharing Kannetzky’s thesis that the content of intentions are action types or, in his terminology, action forms. Because action forms are themselves not actions, a non-circular account of joint action referring to we-intentions, or joint intention, and joint intention action forms is possible. But in contrast to Kannetzky he argues that for the existence of action forms of joint action the existence of a common practice has not be presupposed necessarily. Hans Bernard Schmid criticizes Gilbert’s, Tuomela’s and Searle’s accounts by examining the phenomenon of dissent. He argues that most of the given theories of joint or collective action are neglecting dissent as an important element of collective agency. Dissent, so his claim, involves normativity because the notion of dissent is meaningful only in the realm of normative constrains. The last contribution in this section is Richard Raatzsch’s discussion of the dilemma between individualistic accounts of joint action that seem unable to explain the “jointness” of joint actions, and holistic accounts that seem to introduce metaphysically suspicious entities like group souls, an irreducible “We” etc. Raatzsch shows that this dilemma, if it is a real dilemma, is not only a dilemma of any account of joint action, but of individual action as well. However, because the notion of “we-intent” and the notion of “I-intent” have parallel structures it is possible to explain the one notion with the help of the other and vice versa even if there can be no reduction. The third section of this volume takes up a rather new topic in the discussion of the social aspects of personal life. Under the label Epistemic Ho-

IV

lism two different questions are addressed, namely whether there can be shared knowledge and shared beliefs – discussed by Deborah Tollefsen and Lars Lundsten –, and the relation between conceptual analysis and interpretation. Bernt Österman gives here a coherentist approach to the question of the justification of interpretation that is not only depending from the interpreted text, but also from the context, in which the interpretation takes place and from the guiding purpose of the aim of the interpretation. Regarding shared knowledge and beliefs, Tollefsen defends the thesis that there should be a collective epistemology as a sub-theory of the theory of collective agency, focusing thereby on the subject of knowledge. Lundsten explores the epistemic operations of justifying knowledge of and about social phenomena. The editors are deeply indebted to the Philosophical Institute of the University of Leipzig for providing the institutional frame, and to the Deutsche Forschungsgemeinschaft and to the Gesellschaft für Analytische Philosophie (GAP) for their generous financial support that rendered the workshop possible. Leipzig, September 2006

Katinka Schulte-Ostermann Nikos Psarros

Part I: Social Ontology

The Foundations of Social Coordination: John Searle and Hernando de Soto Barry Smith, Saarbrücken/Buffalo1

1. A Game of Chess What is a game of chess? This simple question has, in the simplest case, a simple answer: A game of chess is a sequence of deliberate moves of certain distinctively shaped pieces across a distinctively patterned board made by two opposing players who alternate in making their moves in accordance with certain well-defined rules of which the players are aware. In short: a game of chess is a sequence of events of a certain patterned sort. Each move in the game is associated with a certain intention on the part of the responsible player, and these intentions – above all the intention to win, and not just to move pieces in accordance with the rules – are indispensable to the game. But they are not parts of the game, any more than the thoughts in the minds of staff officers behind the lines are parts of the battle raging at the front. We can write down the moves made by each player and so keep a record of the game. But this record, too, is not a part of the game, and indeed it may come into existence only long after the game has been concluded. Like the thoughts in the minds of the players, and like published histories of military engagements, it belongs rather to what we shall call the domain of records and representations. Chess reality then consists of two complementary dimensions: the dimension of the game itself, which consists of moves of pieces on a board; and

1

This paper was written under the auspices of the Wolfgang Paul Program of the Alexander von Humboldt Foundation. Thanks are due to Leo Zaibert for valuable comments. N. Psarros and K. Schulte-Ostermann, Facets of Sociality, Frankfurt 2007: 3-22; ontos verlag

4

Barry Smith

the dimension of thoughts, ideas, intentions and deliberations. The latter is an indispensable accompaniment of the game, but it is not a part thereof. The above, however, describes only the simple case. Suppose, instead, that two people play a game of blind chess. Here there are no pieces, no board, and no moves; rather there are just players, their intentions, and the announcements of these intentions in an alternating rhythm that enables each intention to be registered by the opposing player. What is the game itself, in this case? Is it the passage of messages back and forth? I will argue that it is not, but rather that these messages, like the intentions, are again a part of the domain of records and representations. To see why this is so, consider a game of normal chess that is being played on a giant chessboard in which the players send messages to surrogates who are called upon to move the pieces. These messages, again, are not a part of the game, any more than the messages sent from headquarters to the troops on the battlefield are part of the battle. Certainly in both cases the messages, like the intentions, which underlie them, have an important causal connection to the game itself. But the events and processes, which cause an endurant entity to exist, is not a part of that entity – at least not on standard views of the relation between cause and effect. And the same applies also in the case of blind chess: only here, the giant chessboard is absent. What we have instead are images of a chess board in the minds of the players. But these images, too, cannot constitute the game of chess – for they are present also when board and pieces truly do exist. A parallel problem arises in the case of a game of Internet chess. Here the player’s intentions are conveyed by movements of electrons along wires, with resulting changes in computer memory and monitor displays. Here again, the signs conveyed, and the associated blips inside computers, belong to the domain of records and representations. But if not a sequence of messages, or images, then what, in the case of blind chess or Internet chess, is the game itself ? Some might be tempted to suggest that the game itself in such cases is some sort of conceptual entity. But concepts, too, belong on the side of representations; entities such as chess games belong to the side of what concepts represent. (And the fact that the dichotomy between representations and objects is not absolute –

The New Ontology of the Social World

5

there are, for example, paintings which fall under both headings – does not affect our argument here). A thesis to the effect that a game of blind chess is a conceptual entity faces the further problem that concepts as normally conceived are timeless entities. Chess games, however, including blind chess games, are tied essentially to time. In fact, they have a double temporal structure, in that they occur in a specific time interval and they unfold themselves within this interval in a specific order of before-and-after of successive moves (the same order which is captured in the spatial form of above-and-below in the written record).2 An alternative answer to our question might consist in the claim that, when we play blind chess, then there is no game at all. It is merely as if such an entity exists – as in the cinema it is merely as if the represented events were actually taking place in the theater. The players in a game of blind chess are, according to this account, just pretending to play chess, as a pianist may pretend to play the piano by touching the keys but without actually depressing them. This amounts to a doctrine of fictionalism: it asserts that talk about entities of given sorts is only putatively about such entities. When talking about a game of blind chess, just as when we talk about the absence of a pulse, or about the average Spaniard, we are using the corresponding words as mere façons de parler about something else. But this fictionalist alternative, too, is to be rejected. For we can indeed imagine that two people do in fact pretend to play a game of blind chess; but then, on the reading in question, we would have to say that they were in such circumstances in fact pretending to be pretending. The correct answer to our question is rather the following. A game of blind chess is what we shall call a quasi-abstract pattern, something that is: (i)

2

In this respect a game of chess is analogous to a reading of a work of literature, where we can in fact distinguish three levels of temporal order: the level of the reading itself, as a succession of events in real time; the level of the sentences succeeding each other in an abstract temporal order that is reflected in the spatial order of the corresponding printed marks on the page; and finally the level that is constituted by the plot of the work itself, that is to say, by the fictional events which these sentences depict. See Ingarden (1973).

6

Barry Smith

like abstract entities such as numbers or forms, in that it is both nonphysical and non-psychological; but at the same time, (ii) through its association with specific players and a specific occasion, tied to time and history. A quasi-abstract pattern thus has two properties that are normally assumed to be incompatible. On the one hand it has no physical parts, and is not able to stand in physical relations of cause and effect. But on the other hand it is an historical entity, which means that its existence is tied to a certain interval of time and to certain actions of specific players. Already Plato would have regarded such a combination of properties as something impossible. For Plato the forms are essentially non-historical, indeed atemporal; the objects participating in these forms are essentially bound to time and change. To do justice to phenomena like the blind chess game, we need to recognize that there are entities of a third sort, entities which are both abstract (non-physical) but yet historical (they are tied to time). A normal game of chess includes the movements of the pieces as its parts. It is part of physical reality. Interestingly however, normal chess too has a certain abstract character, since it contains these movements as granular parts. This means that certain parts of these movements, for example the interactions of the molecules inside the pieces, are not themselves parts of the game. Rather, they are traced over, in the same way, in which, when we look at an oil painting, we trace over the fine-grained structure of the molecules of which the pigment is made.3 A game of blind chess, in contrast, is a wholly abstract entity. It has no physical parts of any sort. The messages communicated by the players in the course of the game are, like the game itself, ephemeral. They can be transformed, however, into representations that have a lasting existence by being written down. And we note in passing that on the basis of such records a new dimension of chess reality can come into existence: the dimension of status. Chess masters enjoy a special status not least because there exist records of the games they have played. In virtue of the existence of such records, the game has the chance to shape the lives of those involved in new and lasting ways.

3

Cf. Bittner and Smith (2003) and Smith and Brogaard (2002).

The New Ontology of the Social World

7

2. Two Sorts of Social Reality The ontology advanced by Searle in his The Construction of Social Reality 4 focuses primarily on the physical domain of the social world – on dollar bills, presidents and driving licenses, on promisings, marryings and buyings of beer. The formula at the heart of this ontology is: X counts as Y in context C. This formula, which lies at the center of Searle’s thinking all the way from his book Speech Acts to Construction, is satisfied first of all by objects – by husbands, cathedrals, and the listes des prix posted in Paris bistros. In each case there is some physical X term (a human being, a building, a piece of printed cardboard), which counts as a social object of a certain kind in a corresponding context. It is satisfied also by events – by votings and goal-scorings and launchings of ships. In all such courses we have certain distinctively patterned parts of physical reality, which in certain specific kinds of contexts fall under certain specific kinds of descriptions. The corresponding objects and events, correspondingly, come to be ascribed certain properties or powers of non-physical sorts. As falling under such descriptions the X term counts as a Y term of a certain sort. Unfortunately, however, there are entities in social reality – debts, rights, obligations, bond derivatives (and games of blind chess), which do not fit well with Searle’s formula. For here there is no physical X term to which the corresponding properties or powers could be ascribed. They are, rather, in the terminology introduced above, quasi-abstract patterns tied to time and to specific bearers by the speech acts and associated thoughts and intentions that brought them into being. A debt, for example, is in this respect like a game of blind chess. It differs only in its what we might think of as its inner temporal structure and also in its possession of a deontic element: if you have incurred a debt, then this means that you are subject to a certain obligation to repay in the future. A debt is tied to a specific initiating event and to specific initiating partners, but it is able thereafter to 4

Searle (1995).

8

Barry Smith

float free and to enjoy an existence of its own. This existence is however in normal cases an entirely humdrum affair which involves merely enduring through time in a changeless fashion until, through one or other terminating event (such as being paid off or waived), it comes to an end. Debts depend for their existence on representations, which may enjoy a merely ephemeral existence in the form of memory traces, or which may be transformed into enduring representations by being written down. Note that on the basis of such records a new dimension of economic reality can come into existence: the dimension of formal debts. The latter enjoy a special status as a result of the fact that they are registered and recorded according to official procedures laid down in advance. As a result, such debts can be bought and sold, bundled and unbundled, inherited, bartered, negotiated away. And as we shall see, they thereby also have a chance to shape the lives of those involved in new and lasting ways.

3. Constitutive Rules As the rules of chess create the very possibility of our engaging in the type of activity we call playing chess, so, Searle holds, constitutive rules in general, rules of the form X counts as Y in C, create and allow the forms of behavior we call electing, promising, marrying and buying beer. Examples of the formula at work are: X = moving an arm; Y = commanding an infantry troop to stop advancing; knocking over one’s king; refusing an offer; waving to a friend; C = war; chess; business; everyday life. As we can see, the movement of an arm can mean different things in different contexts. This variety reflects the many different sorts of things we do together. We participate in meetings, attend concerts, compete in football games, sell stock, pay taxes, and engage in a huge variety of other types of cooperative behavior, which involve the bringing into existence of what Searle calls social facts through the application of constitutive rules.

The New Ontology of the Social World

9

Human beings enjoy the capacity for what Searle calls collective intentionality. They are able to engage with others in cooperative behavior such a way, which involves sui generis types of beliefs, desires and intentions. Often these involve human beings collectively awarding status functions to physical parts of reality – which means: functions those parts of reality could not perform in virtue of their physical properties alone: the function of a traffic signal in compelling drivers to turn left or the function of a railway ticket in allowing its bearer to travel on a certain train are in this respect to be contrasted with the function of a screwdriver to insert and extract screws; only in the case of the screwdriver does the exercise of a function depend on specific physical properties of the object in question. Note, though, that functions of all types share many ontological features in common with debts and other quasi-abstract entities of the social world (including those entities which Searle calls ‘social facts’). The function of my heart (to pump blood) begins to exist at a certain point in time and continues to exist unchangingly until the terminating event which is my death. The function of the screwdriver, similarly, begins to exist at a certain point in time (the point of first assembly) and continues to exist unchangingly until some terminating event when the screwdriver is broken or destroyed. And the function of the dollar bill, similarly, begins to exist from the point in time when it is issued to the later point in time when it is destroyed or withdrawn from circulation. In each case we can tell a complicated story about the functionings of these functions (the processes, of pumping, inserting and extracting of screws, of being used as medium of exchange) – the functions themselves, however, endure invariantly throughout such changes.5 It is presumably this character of quasi-abstractness which explains Searle’s view that functions are in every case socially constructed (they are a matter of imposition against the background of values that we take for granted.6 Note, however, that since Searle insists at the same time that he is a realist about functions – admitting that we can ‘discover’ functions in nature (p. 15) – this means that he is to this extent already at this point lending ontological credence to the re-

5 6

Cf. Smith et al. (2004). Searle (1995: 13-23).

10

Barry Smith

ality of the quasi-abstract. (And we note that similar remarks could be addressed also to both constitutive and regulative rules; these, too, satisfy all the conditions of quasi-abstract entities as these have been described in the foregoing. The game of chess itself – as type rather than as the tokens we have been considering elsewhere in this essay – might then be subjected to a similar treatment.) In the case of those functions which exist as a result of constitutive rules, they characteristically mark the potentiality for consequences of a specific sort, for example in the form of rewards, penalties, obligations, reasons to act. When, in the right context, I make an utterance of the form “I promise to pay you a hundred dollars tomorrow”, then my utterance counts as the making of a promise. This means that it has highly specific consequences, which include a mutually correlated claim and obligation together with a certain tendency to act. These are deontic consequences, which go far beyond the realm of purely physical causality. A certain entity (in this case, an utterance) has what Searle calls deontic powers in virtue of the fact that the participants involved (for example as speaker and as hearer) have imposed those powers on the entity in question. Such an imposition must rest always on a foundation of ‘brute facts’, by which Searle means facts of natural science, facts which obtain independently of all human institutions, including language. In the final part of Construction Searle rightly attacks those who hold that reality consists of social facts all the way down, so that the facts of the natural sciences would be no different in this respect than facts concerning politics or styles in footwear. Certainly the sentences of natural science are parts of social reality – but, as Searle shows, the same cannot be said, on pain of absurdity, of the facts which make these sentences true. Unfortunately, however, Searle misinterprets the implications of his own insight when he takes it to imply that social reality must in every case be made up of physical parts. On almost every page of Construction Searle either assumes, or states explicitly, or employs examples and arguments which reinforce, the thesis that the X term in his formula must be a part of physical reality. This is so even in those cases where there is some iteration involved, so that the Y term resulting from the imposition of deontic powers on an initial X term itself serves as the X term in a new application of

The New Ontology of the Social World

11

the formula. For the X and Y (and Z …) terms within the scope of a given instance of the formula are in any case identical: they differ only as to the descriptions under which they fall in different contexts. All human institutions, from money and marriage to government, property, and inheritance are, Searle repeatedly suggests, to be understood in terms of a reading of the formula in which the ultimate X term is physical in nature. This insistence that the X term must be part of physical reality (of the realm of what, in Searle’s idiolect, are called ‘brute facts’) derives from Searle’s standpoint as a naturalist, which is to say, as a defender of the view according to which everything in reality is governed by the laws of physics (and thus also by the laws of chemistry, biology, neurology, and so forth). The challenge, which Searle embraces in Construction, is precisely that of building an ontology that is both realist about social reality and naturalist in just this sense. ‘Realism,’ here, means the opposite of fictionalism. It consists in the doctrine that social reality exists, that entities such as claims, prices, financial transactions, elections, trials and weddings are not mere fictions and that our talk of such entities is not a mere collection of roundabout ways of talking about other things.

4. The Ontology of Social Reality Naturalism asserts that everything in reality is constituted by physical particles or fields of force or by the patterns of movement of such entities. For a naturalist like Searle to be convinced of the existence of God would be for him to have some physical evidence of this existence. In fact for a naturalist like Searle, not only the X term but also the Y term in every application of the counts as formula must be physical through and through – the X and Y terms are after all in each case one and the same entity, merely viewed as falling under two distinct descriptions. George W. is still George W. even when he counts as President. Miss Anscombe is still Miss Anscombe even when she counts as Mrs. Geach. But what of those values of Y terms where no candidate X term drawn from the realm of physical reality is available? How can Searle’s natural-

12

Barry Smith

ism allow a realist ontology of those parts of social reality which are constituted by prices, licenses, debts, and taxes? The assumption that X and Y terms are identical works well when the Y term exists simultaneously with the X term, for example, when the issuing of sounds from John’s mouth counts as an utterance in English: the two events are here quite reasonably conceived as identical parts of physical reality, merely: conceived under different descriptions. But an event of promising might last several seconds while the deontic powers to which it gives rise – the claims and obligations – might exist for several months. An event in which Jane gives her watch to Joan might exist for only two seconds while the new relation of ownership that is founded in this event might go on existing for many years thereafter. There is here no piece of paper, no organism, no building, no movement of molecules to serve as physical X term in the future. The watch itself cannot serve this purpose (the watch itself does not count as the relation of ownership by which it becomes tied to Joan) and the same applies, too, to other physical phenomena such as the relevant memory traces in Jane’s brain. The relation of ownership is, rather, what we shall henceforth call a free-standing Y term – it is a sui generis social object of a quasi-abstract sort. Certainly it depends on physical bearers – in this case Joan and the watch. But these physical bearers do not overlap with the relation of ownership itself, as is seen in the fact that the latter has no physical parts. Only in one or two isolated passages does Searle recognize the existence of entities of this sort. He points out that when I promise something on Tuesday, the obligation continues to exist over Wednesday, Thursday, Friday. This “is not just an odd feature of speech acts, it is characteristic of the deontic structure of institutional reality. … think for example, of creating a corporation. Once the act of creation of the corporation is completed, the corporation exists. … It need have no physical realization, it may be just a set of status functions.”7 What Searle does not recognize is that such a set of status functions, even though it depends on physical reality, is not itself a part of physical reality. It is, precisely, a quasi-abstract pattern that is tied to history and time in virtue of its relation to certain persons and events. 7

Italics added. Cf. Smith and Searle (2003).

The New Ontology of the Social World

13

In the following passage, too, Searle accepts that free-standing Y terms exist: The whole point of institutional facts is that once created they continue to exist as long as they are recognized. [...] You do not need the X term once you have created the Y status function. [...] At least you do not need it for such abstract entities as obligations, responsibilities, rights, duties, and other deontic phenomena, and these are, or so I maintain, the heart of the ontology of institutional reality.8 With this, Searle effectively abandons the naturalist horn of the dilemma upon which he has thus far been impaled. In his official stance, however (reproduced also in the comments from Searle appended below), he continues to insist on the correctness of the naturalistic doctrine.

5. Towards Documents Searle correctly emphasizes that the world cannot consist of social facts all the way down with no brute reality to serve as their foundation. But he is in error when he takes this to mean that social reality must be furnished through and through by Y terms which coincide with parts of physical reality. Certainly Y terms cannot float entirely free of all phenomena whose existence is not a matter of human agreement. But this anchorage need not take the same (X counts as Y) form in every case. For there is a second and no less important kind of anchorage, an anchorage in the realm of records and representations. Searle comes close to recognizing the importance of this second kind of anchorage in a post-Construction passage in which he corrects his earlier view according to which credit cards and blips in bank’s computers can count as money.9 Rather, as Searle now recognizes, they are both more properly speaking different representations of money (or more precisely, in the case of credit cards, they are representations of a commitment on the

8 9

Smith and Searle (2003). Smith and Searle (2003).

14

Barry Smith

part of a bank to meet liabilities incurred by the card owner). Similarly, title deeds are not themselves property rights, but rather representations of property rights. An IOU note merely records the existence of a debt; it does not count as the debt, and its destruction need not in and of itself cause the debt to cease to exist. When Juan and Hank need to fly together from Lima to Oakland, Juan lends escudos to Hank at the beginning of the trip, which Hank then repays in dollars on arrival. But no physical money changes hands until they reach their final destination. Rather, in keeping track in their minds of who paid for what in the course of the journey Juan and Hank move quasi-abstract money around in a quasi-abstract space in a way that very much resembles the quasi-abstract movements of quasi-abstract pieces that is a game of blind chess. And when Hank uses his credit card to guarantee his hotel bill upon arrival in Lima, then he and the owners of his hotel are playing what is very like a game of internet chess with their respective banks’ computers.

6. The Mystery of Capital In The Mystery of Capital,10 Hernando de Soto expounds an ontology of social reality in which not physics but rather precisely the realm of records and representations is awarded a central role. The sub-title of de Soto’s book is: Why Capitalism Triumphs in the West and Fails Everywhere Else and its thesis is summarized in the following sentence: “It is the ‘invisible infrastructure of asset management’ upon which the astonishing fecundity of Western capitalism rests.” By “invisible infrastructure”, de Soto means precisely the realm of those quasi-abstract structures which exist not as parts of physical reality but rather in virtue of an anchorage in the domain of records and representations. Mystery covers, though in a different terminology, much of the ground explored by Searle in his theory of collective intentionality and deontic powers. Searle, too, as we have seen, accepts that there is a non-physical side to the ontology of social reality. But he does this only reluctantly. And in focusing on the realm of property records and titles, de Soto shows us how 10

de Soto (2000).

The New Ontology of the Social World

15

what I have been calling free-standing Y terms work to add a new dimension of economic powers in addition to the deontic powers recognized by Searle. It is property, and formal property records, which lie at the heart of de Soto’s analysis. Such records do more than sustain the corresponding property relations in existence; they also bring into being a new phenomenon, called capital. They do this by capturing in concentrated form the economically significant facts about the corresponding physical assets – their economic powers – in ways which allow the latter to be parcelled out and manipulated in new sorts of ways. “The formal property system that breaks down assets into capital is,” de Soto tells us, “extremely difficult to visualize”. The nature of free-standing Y terms allows us to explain why this is so: the system consists of quasi-abstract entities not carved out within the realm of physics.

7. The Construction of Economic Reality When Searle, in Construction, describes how we are able to impose special rights, duties and obligations on our fellow human beings by acting in accordance with constitutive rules, he confesses that this seems to involve “a kind of magic” (p. 45). He then attempts to dispel the air of magic with his notion of collective intentionality. De Soto, similarly, recognizes that there is an air of mystery attached to the way in which capital is born out of physical assets. He tackles the same problem with his account of the role of records and representations. As de Soto shows, “Capital is born by representing in writing – in a title, a security, a contract, and other such records – the most economically and socially useful qualities” of assets. “The moment you focus your attention on the title of a house, for example, and not on the house itself, you have automatically stepped from the material world into the … universe where capital lives.”11

11

de Soto (2000: 50).

16

Barry Smith

This is, be it noted, a non-physical universe, a universe populated by free standing Y terms, where we can take advantage of the quasi-abstract status of its denizens in order to manipulate them in quasi-mathematical ways. We can create ever new types of such entities by composition, division and derivation. We can pool and collateralize assets; we can securitize loans; we can consolidate debts. Shareholders can buy and sell property rights in a factory without affecting the integrity of the physical asset. Individuals and institutions in different countries can trade unlimited quantities of these entities without the need for any physical items to be shifted from one place to another and without the need to build any special storage facilities to house them. Pension funds can exploit the mathematical divisibility of capital to bring about a state of affairs in which the ownership of capital is no longer the privilege of the few. Most importantly, for de Soto, the quasi-abstract nature of capital allows it to serve as security in credit transactions by being moved about, virtually, between different owners and lien- and mortgage-holders. It is not land or buildings, but rather the associated equity – something represented in a legal record or title – which provides security to lenders for liens, mortgages, easements, and other covenants. We add a codicil to a title deed thereby certifying who has access to the property and under what conditions. We present the title deed to a bank and thereby allow the equity associated with the underlying asset to be set free for purposes of investment in other things. It is in this way that the records and representations constituting the formal property system bring a new domain of quasi-abstract reality into existence, whose growth is intimately associated with those advances in human welfare which are associated with economic development. Title deeds, stock certificates, mortgage contracts and their computerized counterparts are the reliable means to discover, with great facility and on an ongoing basis, the most potentially productive qualities of resources, and “As Aristotle discovered 2,300 years ago, what you can do with things increases infinitely when you focus your thinking on their potential.”12 By unleashing the potential of physical assets in the form of credit, thereby allowing new sorts of ventures and new sorts of risk, and new sorts of shar-

12

de Soto (2000: 51).

The New Ontology of the Social World

17

ing of risk, the development of the formal property system gave rise to that quantum leap in human welfare which we associate with the success of Western capitalism. The formal property system also fosters accountability; and thereby promotes higher levels of trustworthiness among those who participate in its development. For accountability means that those who abuse the system, for example those who do not pay back their loans, are diminished in their ability to draw on its benefits in the future. Calling people to account for their actions in this way has positive effects of a range of familiar sorts13 and we can compare the dissemination of institutions of credit checking, debt collecting, payment insurance and the like with the development of the formal accreditation systems for chess masters administered by the Fédération Internationale des Échecs. Both have spurred those involved on to new heights of achievement. From de Soto’s perspective, the modern world can be defined as a common system of enforceable formal property registrations. These registrations make knowledge functional by securing all the information and rules governing accumulated wealth and its potentialities in one knowledge base that makes people accountable across the entire property jurisdiction. This single property system, through trade and the concomitant division of labor, makes possible the astonishing economic development that has been the privilege above all of Western societies in the age of industrial civilization. To be part of the system means to be represented therein with the help of such proxies as one’s name or social security number. With these are associated in turn formal records (of domicile, creditworthiness, ownership) together with those informal estimations which exist in potential investors’ and customers’ minds of the skills and reliability and resourcefulness of those with whom they have to deal. Because the bank knows your address, and has the title to your property in its vaults, the bank trusts you with resources to invest in new ways. You then have the means to try out new ideas. And because you know that failure will bring real loss, you have a real incentive to succeed with these ideas, and thus to acquire a reputation for reliability, honesty and integrity. All these things contribute to your own wealth and to the wealth of those around you, and 13

Klein (1977).

18

Barry Smith

own wealth and to the wealth of those around you, and the reason that you can use credit cards when you travel from Oakland to Lima and back is because of the records that tell the card-issuing authorities who should and who should not be given credit.

8. The Realm of the Quasi-Abstract De Soto errs, however, when, as in the following passage, he talks of freestanding Y terms as if they were mere concepts: “The proof that property is pure concept comes when a house changes hands; nothing physically changes.”14 For concepts, as we have already noted, belong with ideas and intentions to the realm of representations. Property itself, by contrast, belongs to the realm of that which is represented. More precisely, property relations belong to the realm of the quasi-abstract and they are in this respect comparable to symphonies, laws and other quasi-abstract denizens of the social world. That they exist on the side of the objects and not on the side of the concepts in people’s heads can be seen from the fact that concepts can exist even where there are no corresponding objects. To be sure, concepts are important. Without concepts, and without associated thoughts and intentions, the corresponding free-standing Y entities would not have been brought into existence. When we buy and sell, however, then we are interested not in concepts but in the objects themselves: in equity and capital, and in all that goes together therewith – starting with the simple trading, offering and splitting of stock and moving on to the unimaginably complex edifices of contemporary derivatives markets. Formal property requires the existence of two distinct sorts of entities. It requires on the one hand quasi-abstract free-standing Y terms, and on the other hand records upon which the existence of the former depends. Note – in case this is still not clear – that the latter are not X terms according to the letter of Searle’s formula. The pieces of paper in the bank vaults do not count as the ownership rights that are documented therein. Rather, they represent them. But the pieces of paper are like X terms at least in this: 14

de Soto (2000: 50). In the passage already quoted above, de Soto talks of stepping “from the material world into the conceptual universe where capital lives”.

The New Ontology of the Social World

19

they are physical entities which serve, through the workings of collective intentionality, to provide the basis for the corresponding Y terms. The pieces of paper can also serve to represent the associated objects and relations in another sense: they serve as their proxies – so that control over paper dees or titles implies a form of control, too, over the quasi-abstract entities for which they stand. We have seen that Plato would have rejected the existence of quasiabstract entities of the sorts, which populate those rapidly growing suburbs of the social world which are so important for the sorts of economic development. The same applies, too, though for different reasons, to Marxist economists, who cast aspersions on the ‘speculators’ and others who tend the realms of the quasi-abstract – because of their conviction that all that is of value must flow from physical labor. In this they are, like the defenders of legal positivism, and like naturalists of various other stripes, manifesting a prejudice in favor of what you can touch and see. De Soto with his analysis of the workings of the formal property system, and Searle with his doctrine of the ‘huge invisible ontology’ of social reality,15 have taught us that we need to slough off this prejudice in favor of a more adequate system of categories. Searle, now, should have the confidence of his convictions and recognize that the social world contains more, much more, than appendings of non-physical descriptions to physical objects and events.

9. Coda: Searle vs. Smith Searle: I agree with most of what Barry has said, but I think that he is being needlessly paradoxical when he suggests that there is some challenge to naturalism here; that somehow or other, in addition to physical particles and fields of force, there are all these abstract entities running around between the molecules. That’s a misleading picture, which comes from treating the object as the unit of analysis. We’re not interested in the object, we’re interested in the processes or, as I like to put it, we’re interested in the facts. It isn’t the obligation as an object that is the topic of our investigation, rather it is our undertaking an obligation, our recognizing a pre15

Searle (1995: 3).

20

Barry Smith

existing obligation, our fulfilling an obligation. And when you realize this the threat to naturalism disappears. In fact I like the example of blind chess: here representations of the pieces take the place of the pieces. So when blind chess players play a game, then they keep a record and say PK4 will be the first entry on the sheet, which means white moves pawn to king 4, and then they fill in the rest and they can always go back and look at the record and see what the position on the board is. Well, of course, there’s an abstract character to all this; but the record is part of the real world and I think part of the difficulty with his presentation is that Barry is really attached to the old notion of the physical and to this dumb Cartesian vocabulary we’ve inherited. But I think we shouldn’t be misled by it. The world contains everything it contains; we’re used to calling it ‘physical’ because we think that physics is somehow or other the basic science, which it is. But if you describe what Barry said without using the ontological categories that he seems to be committed to, then it contains no threat to naturalism at all. I think Barry made a valuable contribution in recognizing that in many cases the representation is all the reality we need to make the entity function. Interestingly, my very first examples were cases of that: paper money was originally a representation; it was a note that said “I promise to pay the bearer on demand.” Then the representation of money became money. In the other cases you have representations of chess pieces that now function the way that chess pieces do. So I agree with the general thrust of the argument, but I think it’s needlessly paradoxical to suggest that we’ve somehow got to alter our whole metaphysics, we don’t. Smith: I agree with John that we should get rid of these old Cartesian dualist notions; I disagree with him when he thinks that there are no problems here for naturalism. He thinks that we can solve the problem by turning away from social objects and by looking at facts. Do you own stock John? Searle: Well, in the aspect that I … Smith: Just say ‘yes’, John. Searle: Alright, I’ll say yes ... Smith: And when you’re lying in bed at night are you thinking about the facts and processes that pertain to when you bought them and the transac-

The New Ontology of the Social World

21

tions that you made? Or are you thinking about the stocks themselves, the objects? Searle: I don’t think about the stock an sich. No, what I think about are not the stocks themselves, but rather their current market value and how it is declining. Because, you see, the subject-predicate structure of language makes it look as if there’s this pre-existing set of objects, my stock, but in fact what we’re talking about is a process: the stocks go up and down, the stocks split, and when they split this doesn’t mean that they physically split, it means that there’s a different entry in the system of representations. Smith: I think that if we are going to understand the wondrous ways of capital, then we need to think very precisely about this process called ‘splitting stock’. But that means also that we need to take seriously the fact that such processes involve objects – and that this is so even when the stock itself exists only in virtue of certain representations which themselves exist in the form of blips on computers. Do you agree with that? Searle: No I don’t, I think that that’s the wrong picture. The real picture is this: we have a set of processes and we have a set of representations which enable these processes to function, and in the case of the stocks splitting the corporation makes certain entries into their databases, into the system of records and representations whereby you are now represented as having twice as many shares of a stock as you did before and that is the reality, that representation is constitutive of your having twice as many shares as you did before. Smith: I agree with all of that, I just think that, in the spirit of the First Axiom of Realism for Social Reality, you should take equally seriously every single word in this description you just gave.

22

Barry Smith

References Bittner and Smith (2003): Bittner, Th. and Smith, B. A Theory of Granular Partitions, in: Duckham, M. Goodchild, M.F., and Worboys, M.F. (eds.), Foundations of Geographic Information Science, London 2003: 117-151. de Soto (2000): de Soto, H. The Mystery of Capital, New York 2000. Ingarden (1973): Ingarden, R. The Literary Work of Art: An Investigation on the Borderlines of Ontology, Logic, and Theory of Literature, Evanston/IL 1973. Klein (1977): Daniel B. Klein, D.B. (ed.), Reputation: Studies in the Voluntary Elicitation of Good Conduct, Ann Arbor: University of Michigan Press, 1997. Searle (1995): Searle, J. The Construction of Social Reality, New York 1995. Smith and Brogaard (2002): Smith, B., and Brogaard, B. Quantum Mereotopology, Annals of Mathematics and Artificial Intelligence, (2002) 35/1-2: 153-175. Smith and Searle (2003): Barry Smith, B. and John Searle, J. “The Construction of Social Reality: An Exchange”, American Journal of Economics and Sociology, (2003) 62:2 (2003), 285-309, also published in: Laurence Moss, L. and David Koepsell, D. (eds.), Searle on the Institutions of Social Reality, Oxford: Blackwell, 2003. Smith et al. (2003): Smith, B., Papakin, I., and Munn, K. Bodily Systems and the Spatial-Functional Structure of the Human Body, in: Pisanelli, D.M. (ed.), Ontologies in Medicine: Proceedings of the Workshop on Medical Ontologies – Rome, October 2003, Amsterdam 2004.

The Constitution of Social Objects by Common Actions Nikos Psarros, Leipzig As shown by Barry Smith,1 the manifestation of social objects does not rely on the existence of physical objects as proposed by John Searle.2 It is rather so that in order for a physical object to play a certain social function, e.g. for pieces of paper to count as money, the existence of the corresponding social object is a conceptual prerequisite. In other words, the relationship between a physical and a social object is not by any means a constitutional, but a representational one. Social objects can exist without being attached to any physical substrate whatsoever. This circumstance, however, leads to a problem regarding the validity of an ontology of social objects: Physical object ontologies drain their validities from the natural order of their corresponding objects. Being nonpalpable and non-sensible, social objects cannot guarantee the validity of a corresponding ontology in the same manner as physical objects do. Does this mean that social object ontologies have always to be related to physical object ontologies, rendering thus the independent existence of social objects obsolete? The problem appears because social objects are regarded by Barry Smith as self-sufficiently existing individual entities (like physical objects), either as individualized states, or as “potentialities” that can be reified in form of individual physical objects – B. Smith gives the example of recording a debt, or registering a property. A self-sufficiently existing entity neither relies on the existence of other similar entities, nor on entities on the existence of which it could depend, either logically or conceptually, or because it needs them for becoming effective (as in the case of money), or because it needs them for example as sources of food or energy. Typical self-sufficient enti-

1 2

Smith (2007). Searle (1995).

N. Psarros and K. Schulte-Ostermann, Facets of Sociality, Frankfurt 2007: 23-31; ontos verlag

24

Nikos Psarros

ties in this sense are material things, elementary particles, Empedokleian elements, genes and also “propositional” entities like axioms or principles. Regarding social objects as self-sufficient entities would mean that they exist independently of the social interactions – in Searle’s terms: the context C – they are involved. This contradicts, however, Searle’s very definition of a social object “X counts as Y in context C”3, and also the idea that Y is attached to a physical object X. Y as a self-sufficient entity exists independently both of the physical and the social world. A point for Searle’s conventionalist approach is that the idea of the selfsufficient existence of social objects opposes our intuition that such entities should not exist or persist without our acceptance, recognition or consent. A person, for example, cannot claim ownership upon North America even if she produces credible evidence that she is a direct descendant of the 10th century Viking Chieftain Olaf the Blue who set foot ashore on the continent long before Christophorus Columbus and took possession of the land by virtue of his royal authority, because this authority is no more recognized by anyone. An additional argument against the self-sufficiency of social objects is that it confronts us also with an epistemic problem: How can we have knowledge of objects that do not produce any sense data and do not interact physically with the rest of the world? On the other hand, as Barry Smith argues, the existence of social objects cannot depend, at least completely, on the existence of physical objects or spatiotemporally defined social contexts. Human rights, for example, do not cease to exist only because in a given social context they are not accepted, or because a government thinks that they do not apply to some classes of persons like “unlawful combatants” and the sort. Similarly property rights and legal or moral obligations can in certain cases be enforced upon people who do not even have a grasp of them creating thus a new social context as in the case of the prohibition of cannibalism among the indigenous tribes of Papua-New Guinea imposed by the Dutch and the British colonial authorities, or of the prohibition of ritual homicide practices by the Aztecs that was enforced by Hernán Cortes, or of the abolition of the

3

Searle (1995: 28).

The Constitution of Social Objects by Common Actions

25

death penalty that is demanded from every state wishing to join the European Union. It seems that both aspects are somehow necessary for our life-worldly concept of social object, so that it does not make any sense to abandon the one in favour of the other. Thus it is worth to look for a way to reconcile the conventional with the “absolute” aspect. This way takes us, however, away from the idea that social objects are a sort of individual entities, either selfsufficiently existing or not, by stressing the peculiar relationship between physical and social objects in Searle’s definition. It is the relationship of “counting” that can provide us with a clue about the very nature of social objects: “X counts as Y in the context C”. “Counting as” means here that a given physical object X has the value of being (representing, manifesting, realising) a given social object Y in C. With other words the given social object is realised in C by means of X. X is thus a sort of a tool for achieving the purposes connected with Y in C. Tools are normally embedded in actions. In Searle’s definition the only candidate for being an action is the context C itself, since X is the tool and Y, the social object, is connected with the goals that should be achieved by using X. However, the social object Y cannot be the goal, or the final result of the context-action C, since it is not for Y’s sake that C takes place. On the other hand Y comes into being by performing C; it is a product of C despite the fact that it is not its final result or goal. Money for example is a product (but not the only product) of the context-action of “trading” by means of which the goal of exchange of goods is achieved. In individual good exchanges the immaterial tool money is realised by means of material tools like coins, bills, cheque issues or credit card transactions. There are, however, also context-actions like “promising” or “arguing” that do not involve any material tools, although they always employ immaterial ones like promises or arguments. Context-actions involve always more than one agent. They also require that the participating agents co-operate and co-ordinate their individual actions. Context-actions do not require, however, that the participating agents perform individual actions of the same type, or that their actions are sequentially arranged or that they are rigidly organized. Context-actions like trading, promising, or arguing are performed not by “adding” or superimposing iterated individual actions of the same type, but by the interference

26

Nikos Psarros

of individual actions, which, besides of their individual goals, aim at the same instance also at the goal of the context-action in question. I call actions of this kind, common actions. I use here the term “common” in the same sense as in expressions like “the Boston common”, or “common property” etc. Such a property is something that can be used by a lot of individuals for their individual purposes, but each co-proprietor is obliged to take care for the common property so that it remains in an operational state. The use of a common property does neither require that it can be used only when every entitled co-proprietor is present, nor that everyone uses it for the same individual purpose, although not every individual purpose may be allowed when using a given common property. Another important feature of common properties is that single co-proprietors cannot dispose of the common property without the consent of the others, although they may be allowed to carry out certain modifications on it. Analogously the aims and the success conditions of a common action are conceptually independent from the aims and the success conditions of the individual actions, by which it is realised. A common action has to be learnt and should be described without recurring to individual actions. On the other hand, a given common action can be misidentified despite the fact that the realising individual actions are identified correctly. For example, a person that is not familiar with the European sportive culture could think that the crowd moving on a Saturday evening through the city shouting, waving flags and fighting with the police is a sort of political demonstration not recognizing that it is a mob of drunk and violent hooligans after a soccer game. The problem of the foreign spectator is not that he cannot identify correctly the individual actions, but rather that he doesn’t know that under the given circumstances they realise a common action that is not a political demonstration. I regard “joint” actions as a special case of common actions: Joint actions, like make a walk together or going to the cinema, require the physical presence of all participants realising them at the location where the action is realised, and they require also that the realising actions are of the same type: walking, watching the same movie etc. Common actions require neither the physical presence of all participants at one spot nor the typesameness of the realising actions. The staff of an international company is

The Constitution of Social Objects by Common Actions

27

scattered around the world performing quite different tasks, nevertheless it realises a common action. Common actions are also not identical with “public” actions, although some common actions are public in the sense that everybody is allowed to partake. The success or the failure of the individual actions realising a common action is a necessary but not a sufficient condition for the success or the failure of the realised common action. A vote, for example, can fail in spite of the fact that almost every voter has voted correctly, because none of the candidates has gained an absolute majority or because a given participation quorum has not been reached, or for other formal reasons that affect the vote. If and to which extent an individual action was sufficient for the achievement of the common goal is a matter that can be settled only post hoc as part of the explanation of the success or the failure of the common action in question. The success or the failure of the common action itself depends, however, only on the fact that the goal of the common action is achieved or not. Between a common action and the realising individual actions exists a conceptual hierarchical relationship. The description of the structure and the success conditions of the common action are the measure for the evaluation and the recognition of an individual action as realising part of the common action in question, so that in order to recognize an individual action as realising part of a given common action, the common action must be in progress. This does not mean that common actions cannot have a starting and an ending point in time, although there are also perennial common actions. However, the commencing of a common action cannot be connected to the actualisation of a single individual action. Cases where common actions are initiated by a single individual action, as for example the declaration of the opening of the Olympic Games by the head of the state, do not count as counter evidence because every individual action that initiates a common action unfolds is authority as realising part of a higher order common action that has the structure of a state, a family, a tribe or an institution. Thus the Olympic Games can be declared open only by persons who are entitled to do so in a given social framework. The conceptual hierarchical order between common and individual actions holds even in cases where it is not possible to determine exactly the start-

28

Nikos Psarros

ing and ending point of the common action. The storming of the Bastille, for example, was neither explicitly arranged, nor ordered by any law, nor initiated by the appeal of a person of recognized authority. Obviously, the Bastille was a symbol of the Ancien Régime and it was generally known that it served as a post for the military control of Paris. Thus it might be that among the insurgents of the year 1789 there were people who planned or had taken into consideration attacking the Bastille. There is, however, no historical evidence for a concrete individual or a group who called for or initiated the storming. Nevertheless, an individual attacking act was part of the events only if it occurred during the riots. That means that even if the temporal frame of a common action cannot be determined exactly, the action has to be in progress for an individual action to be recognized as a realising part of it. Any individual attack at the Bastille before and after the storming does not count as a realising part of it. The conceptual hierarchy between common and individual actions is only a necessary and not a sufficient condition for determining the parthood of an individual action to a common one: Not every individual action that was performed in Paris on July 14th, 1789 was part of the storming of the Bastille. This underlines the fact that in order to participate in a common action the individuals have to orientate themselves at the goal of the common action. Common actions are not the result of the superposition, but rather the result of the interference of the individual actions realising them. This means that they are not realised by iterating similar individual actions, as it is the case for example with a path produced by individuals who repeatedly cross the lawn at a certain point, but by the mutual interaction of individual actions that are dedicated, organized and co-ordinated with respect to the common goal. Necessary conditions for the success of such an enterprise are the communication among the participating individuals and the circumstance that they are aware of the common goal and of their relative positions and tasks in the structure of the common action. This means that the participating individuals have at least a rough idea of what they should expect from the others and of what they have to do themselves. People involved in a commercial enterprise for example, should know the kind of enterprise they are working for, their particular roles therein, the defini-

The Constitution of Social Objects by Common Actions

29

tions of their jobs, and the hierarchical structure of the company. They should also know at least in outlines what kind of assistance they can request from their colleagues. At the level of the action types – in case of an enterprise would this be the level of the description of the company type, to which a particular commercial enterprise belongs – constitute the legitimate expectations and commitments of the individual participants in a common action the rights and duties this action type entails. Rights and duties are products of common action in the same sense as any other social object like money, money bills, exchange values, prices, ranks, social roles, or institutions because by their means a common action is given a structure that renders possible the achievement of the common goal. At the level of action types these objects have generic forms like money, exchange value, property, social roles, and constitutive explications of ranks or institutions. At the level of actions in progress realised forms of these objects are involved like bills and coins, prices, property titles, concrete ranks or individual institutions.4 Some of them appear in form or material objects as in the case of coins, bills and military rank signs and uniforms, other remain immaterial and their awareness relies on the situational knowledge of the participants, as it is the case with directors, mothers, fathers, professors, promises and so forth. The concrete appearance of social objects, namely as material objects or as immaterial entities is mostly a historical and cultural issue: in the middle ages the social status of women and men was clearly indicated by their clothing, today the outfit of a person is a quite unreliable indicator of her social status. Nevertheless, among contemporary Hindu women it is still customary to indicate their status as non-widows by carrying a red dot on their brows. As products of common actions, realised forms of social objects exist as long as the common actions that bring them into existence are not interrupted. They cease to exist when these actions come to an end. The generic forms of social objects exist as long as the corresponding common action type does not fall into oblivion.

4

I skip here the quite interesting case of “intermediate” social objects like currencies that are located conceptually between generic form social objects like money and realised forms like coins and bills.

30

Nikos Psarros

The controversy between Searle and Smith rises from the circumstance that both neglect the distinction between the generic and the realised forms of social objects outlined here. John Searle’s considerations concern thus mainly realised forms of social objects. He examines the conditions of their coming into being and disappearing within the framework of common actions in progress. From this point of view the conventional aspect of social objects predominates, with “conventional” referring here to the fact that people must be reforming a common action of a given type for the corresponding realised forms of social objects to come into existence. This is the reason why particular property rights, for example the ownership claims upon North America from descendants of Olaf the Blue, are not recognized notwithstanding the fact that property as an institution resp. as a generic form of a social object remains intact. Barry Smith, on the other hand, focuses his attention on the absolute aspect of social objects by pointing to the existence of generic forms of social objects. His controversy with Searle arises from the circumstance that Smith thinks that some of those generic forms are in fact tokens or individual realisations of general concepts, arriving thus to the conclusion that “free floating quasi-abstract Ys” exist in the same sense as Searle’s coins, paper titles etc. However, he also stresses the fact that these “free floating quasiabstract Ys” need always to be represented in order to have any effect – either by material objects or by individual actions. Thus his quasi-abstract social objects are ideas in a platonic sense, despite the fact that they contain references to individual persons. This is so because individual persons are not tokens of a generic type ‘person’, but something like ‘perceptible ideas’. Being a person means that I realize my life being led by the idea of myself as a person by participating in common actions of the above mentioned structure. It is on that basis that the ontology of social objects can be established since generic forms of social objects exist without the existence of corresponding realised forms. Namely, as long as a common action type is realised successfully without problems, or as long as it is not realised at all without having fallen into oblivion, or as long as we do not become aware that we actually do not realise a common action properly missing thus the real common goal, we are not necessarily aware of all the social objects,

The Constitution of Social Objects by Common Actions

31

generic or realised, involved in that particular action although ontologically speaking they exist independently of our knowledge of them. This is the reason why simple historical unawareness of key concepts of human life like human rights does not automatically permit individuals or even cultural groups to practice torture, ritual homicide, cannibalism, or the death penalty. An ontology of social objects, either in their generic or in their realised forms, drains its validity from the structure and the order of the common actions (more exactly: common action types) in a given community. Actions in general are not a sort of mere physical events, but constitute a conceptually and ontologically independent category of human life. Thus the existence of social objects relies on the existence of individuals with full fledged theoretical and practical capabilities that live in a community providing the social, moral and material framework for the realisation of those common action types that can produce the particular social objects. Since the nature of physical objects is grasped only by integrating them into common enterprises, like technical, artistic or scientific practices, the validity of physical object ontologies relies partially on the same conditions that sustain the validity of social object ontologies: common actions in a theoretically and practically structured community.

References Searle (1995): Searle, J. The Construction of Social Reality, New York 1995. Smith (2007): Smith, The Foundations of Social Coordination: John Searle and Hernando de Soto, in: Psarros, N., and Schulte-Ostermann, K. (eds.), Facets of Sociality, Frankfurt 2007: 3-22.

Kosmos Noetos and Carnap’s Constitution of Cultural Objects Arto Siitonen, Helsinki

1. Introduction Rudolf Carnap introduces “die geistigen Gegenstände” in § 23 of his work Der logische Aufbau der Welt.1 Let us call these objects the noetic ones (from the Greek word ‘noetos’ = capable of being grasped by the intellect, the intelligible). There is, according to Carnap, a manifold of autonomous kinds of objects.2 The different categories of objects are the mental (psychisch), the physical (physisch) and the geistig – i.e., the noetic (Carnap does not use this word). In his constructive epistemology, Carnap starts from elementary impressions (Elementarerlebnisse) which occur in one’s own mind (eigenpsychisch) – that is to say, in Carnap’s mind – and builds up the physical world and the world of the cultural objects, i.e., the noetic world. The “geistige”, noetic, and the cultural objects (kulturelle Gegenstände) rely on a physical basis. According to Carnap, the physical foundation of, for instance a work of art, is to be called the documentation of that work. Documentation is based on the coordination between the noetic and the physical levels of reality. A spectator’s mental apprehension of that work is called by Carnap its manifestation. Manifestation is based on the coordination between the noetic and the mental levels. According to Carnap cultural objects can be traced back to mental objects. However,3 he stresses that the constitution of the cultural from the mental does not mean that the former would be “psychologized”.4 His view can be understood to be as follows: mental activity is required to produce, and to

1 2 3 4

Carnap (1928). Carnap (1928: §160). Carnap (1928: §55). Carnap (1928: §151).

N. Psarros and K. Schulte-Ostermann, Facets of Sociality, Frankfurt 2007: 33-49; ontos verlag

34

Arto Siitonen

maintain, noetic objects, but noetic objects are something different than mental objects. Noetic objects are studied by cultural sciences (Geisteswissenschaften).5 The basic problem of these sciences is concerned with the coordination of objects in accordance to the documentation relation and the manifestation relation.6 In the present article, the levels of physical, mental and noetic are first inspected, and then Carnap’s basic coordination relations are studied. After that, the relations of manifestation and documentation, as well as their converse relations (sign; expression) are analyzed. The order of Carnap’s constitution is then examined. The third section concerns Carnap’s view on understanding and interpretation. The fourth section contains an inspection of the notion of reducibility and of Carnap’s claims concerning reduction. The fifth section is devoted to critical questions and comments concerning Carnap’s undertaking. In Plato’s philosophy, ideas make up the uppermost or true reality, kosmos noetos. Bernard Bolzano’s theory of sentences and truths in themselves is concerned with the occupants of the noetic sphere. Nicolai Hartmann in his “new ontology” studies the problem of noetic being, “geistiges Sein”. Karl R. Popper divides reality into the three worlds of physical, mental and conceptual objects. It is advisable to study Carnap’s theory of the “geistige Gegenstände” in the context of these, related views. This will be done in the final section of the article.

2. On Carnap’s Program: Constitution and its Levels Carnap’s “Der logische Aufbau der Welt” is not a treatise on ontology, and due to his critical, even polemic relation to metaphysics he may be thought to shun ontology as well. However, the very name of the book suggests an ontological approach. According to §3 the method to be followed is 5 6

Carnap (1928: §23). Carnap (1928: §24).

Kosmos Noetos and Carnap’s Constitution of Cultural Objects

35

analysis of reality with the help of the logical theory of relations.7 “Analysis of reality” is clearly an ontological undertaking. One may also speak of the ontological commitments made by Carnap. A basic role among these commitments plays his distinction between various spheres of objects, and the claim that these spheres are autonomous. In traditional metaphysics, the adequate expressions are “Spheres of Being” and “Levels of Being”. Carnap distinguishes at the outset three such levels: the physical, the mental and the cultural one. However, the mental sphere is divided by him into the egocentric area and the outward area. The latter of these is called by him “fremdpsychisch” – i.e., it is supposed to be the region of the other minds.8 LEVELS noetic mental physical

ν ψ φ

Figure 1. The mental level has two aspects: “eigenpsychisch” and “fremdpsychisch”. Carnap introduces the concepts ‘coordination problem’ and ‘essence problem’.9 According to him, it is a task of science to treat the former ones, and that of metaphysics to treat the latter ones. From this he infers the claim that the coordination problems do not pose any fundamental difficulties.10 The basic coordinations are that between the mental and the physical (the psychophysical connection), that between an expression and what it expresses, and that between a sign and what it signifies.11 The expression 7 8 9 10 11

Carnap (1928: §3). Carnap (1928: §58). Carnap (1928: §20). Carnap (1928: §21). For the sake of clarity, let us give the following list: documented: ν in φ manifested: ν in ψ expressed: ψ in ν ; or ψ through φ

36

Arto Siitonen

relation is converse to the relation of manifestation; the sign relation has as its converse relation that of documentation. ν

expression

manifestation documentation

ψ sign

φ

mind-body

Figure 2. The arrows are indicating the directions of the relevant relations. Carnap calls his method ‘constitution’ and says that he employs a “neutral language” in respect to the alternatives of realism and idealism.12 A system of constitution gives an order of objects according to the receipt that the objects of a higher sphere are “constituted” from objects of a lower level.13 Carnap compares his method to Leibniz’ ‘ars combinatoria’.14 Bertrand Russell speaks of logical constructions in the motto of Der logische Aufbau der Welt: “The supreme maxim in scientific philosophising is this: Wherever possible, logical constructions are to be substituted for inferred entities.”15 – one may wonder why Carnap uses the word ‘constitution’, which has a more static connotation than Russell’s term. The main question of the book is how various entities can be constituted logically when we begin from our immediate sensations. The expression ‘entity’ is hereby to be understood to contain everything on which it is

12 13 14 15

signed: φ for ν Carnap (1928: §5). Cf. Carnap (1928: §2). Carnap (1928: §3). Carnap (1928: 1); cf. Russell (1954: 148). Russell actually writes “philosophizing“, whereas in Carnap’s motto, ‘z’ has for some reason been replaced by the ‘s’

Kosmos Noetos and Carnap’s Constitution of Cultural Objects

37

possible to form a judgment. The main question is a strongly Humean one, and the program based on the question is indebted to Russell. Carnap calls the goal of his approach “a constitution system” of entities and concepts; this system is to be built according to the order of knowledge (erkenntnismäßig).16 Carnap notes that there are two possible forms of a constitution system:17 its basis may lie either in physical or in mental occurrences. This is due to the reciprocal reducibility between the physical and the mental. The decision in favour of the chosen system form leans on the criterion of immediacy. It is a characteristic feature of knowledge that we have an immediate access to our own mental occurrences. On the other hand, we may know of the mental occurrences of other persons only through perceiving their bodily movements – and in the future possibly by studying the occurrences in their brains. Thus the constitution has to start from one’s own psyche, then build up the external physical world from this basis, and finally proceed on to other minds. On this psychophysical basis the world of cultural objects will then be constituted. The following figure shows, how the Carnapian constitution proceeds: ν 3 ψ 2

1 φ

Figure 3. The arrows show what is constituted on the basis of what, and the numbers indicate the order of constitution. 16 17

Cf. Carnap (1928: §§54, 58). Carnap (1928: §58).

38

Arto Siitonen

As one may see, the starting point lies in the mental, and the constitution returns to the mental after the physical has been reached. But it returns to it with a new view on the mental, not reflected in the figure. Carnap calls the result of this constitution step “das Fremdpsychische” – i.e. the world of other minds.18 Cultural objects are thus constituted on the basis of an individual mind, after the meanwhile constituted level of physical objects and the aspect of other minds are accounted for.

3. On the Relations of Manifestation and Documentation: The Sign Relation and the Expression Relation. Carnap stresses in §23 the autonomy of the cultural sphere.19 According to him, this autonomy was not properly considered in the philosophy of the 19th century, but the new philosophy of history since Wilhelm Dilthey has given it its due significance. Cultural objects are not built up from mental objects, let alone from physical objects, although they are constituted on the basis of the mental and the physical. Mental occurrences are always bound to one specific carrier, whereas the carriers of a cultural object (a state, a habit, etc.) may change without the object disappearing by that. The cultural sphere manifests itself in the sphere of the mental and is documented on the physical level. Manifestation is defined by Carnap as those mental occurrences in which a cultural object appears, i.e. becomes manifest.20 He adds here the following qualification: also a physical object can be a manifestation of a cultural object (for instance, when a disposition is accompanied by some bodily movements). However, a closer inspection shows that the mental manifestation relation remains basic. – In fact, if physical object per se would be accepted as the other relatum of manifestation, this would blur the distinction between manifestation and documentation. The documentation of a cultural object is defined by 18 19 20

Carnap (1928: §58). Carnap (1928: §23). Carnap (1928: §24).

Kosmos Noetos and Carnap’s Constitution of Cultural Objects

39

Carnap as enduring physical objects on which the cultural life has been perpetuated.21 ν

φ1

φ2 φ3 …

Figure 4. The arrows indicate the documentation relation, ν the cultural content in question, and the φ’s the relevant physical objects. (e.g. Carnap’s work Der logische Aufbau der Welt and its printed copies).

4. Carnap on Interpretation and Understanding. Accordingly, manifestation is fastened to something mental, and documentation to something physical. Documentations are physical carriers of conceptual contents, which manifest themselves in the minds of persons. Such a person may be one who has created the content in question, or one who interprets that content. The creator has given to his or her experience an expression by way of the conceptual content in question, and has documented this content by means of a certain set of signs. The interpreter has the task of showing what the signs in question signify and how they do this. In giving the interpretation, the interpreter proceeds, on the one hand, from the manifestation which the content has impressed to his/her mind, and, on the other hand, from the outward, documented sign of that content. The interpretation, then, analyses what and how the sign signifies.22

21 22

Carnap (1928: §24). Let us add to the previous note the following: apprehended: ν by φ created: ν by ψ interpreted: ν by ψ

40

Arto Siitonen

The above conclusions are not all of them Carnap’s, but they can be drawn from his text. In spite of the robust conceptual machinery developed by Carnap for evaluating interpretations, he considers that interpretation is a metaphysical theme.23 In an interesting way, Carnap says that the relation between the sign and what it signifies always contains a conventional ingredient, and this means that it is established by volitional decisions (irgendwie willensmäßig gestiftet).24 He adds somewhat cryptically that only seldom is here supposed a special essential relation of symbolizing.25 Carnap also says that the relation of signs to what they signify is manifold and almost inexhaustible in its extension. There are written signs, signals and indices. The coordination problem concerning the sign relation can be resolved, but hardly to be answered by only one theoretical system.26 This semiological consideration comes close to the ideas of Charles S. Peirce, but his name is lacking in the register of the book.

5. On the Notion of Reducibility and Carnap’s Claims concerning Reduction. Carnap presents in the title of §55 of “Der logische Aufbau der Welt” the claim that cultural objects are reducible to mental objects. What this means he explicates in the text. The relations of manifestation and documentation mediate our knowledge of cultural objects. Cultural sciences know their objects, whether these belong to morals, language, state, economy or art etc., through understanding (Verstehen). All understanding of cultural objects starts from their manifestations and documentations. Carnap concludes that all cultural objects are reducible to their manifestations and documentations. Documentation of a cultural object necessarily takes place

23 24 25 26

Carnap (1928: §169). Carnap (1928: §21). „Nur selten wird hier eine besondere Wesensbeziehung des ‘Symbolisierens’ angenommen.“ Carnap (1928: §21). Carnap (1928: §21).

Kosmos Noetos and Carnap’s Constitution of Cultural Objects

41

with the help of manifestation, because the act of creation by one or several persons contains mental occurrences in which the cultural object in question becomes alive. Carnap claims moreover that all documentations are reducible to manifestations.27 His theory of constitution contains the idea that cultural objects are reducible to mental objects. Finally, it is not only the case that physical objects are reducible to mental ones.28 Reducibility holds also in the reverse direction: all mental phenomena go back to physical objects. This possibility of reduction is based on psychophysical parallelism and the relation of expression. To every statement on mental occurrences it is possible to coordinate a statement on nervous systems. A person may express him/herself by speaking, writing or by other signs. These let others know what is going on in that person’s mind. Accordingly, every statement on a mental occurrence can be transformed into a statement concerning those physical indicators which tell what is happening. The following figure describes Carnap’s view: ν

ψ

φ

Figure 5. The arrows indicate what is reducible to what. It is easy to misunderstand Carnap’s view on reduction in Der logische Aufbau der Welt. This happens when reduction is understood to imply reductionism, or when the pages concerning it are read in the light of physicalism to which Carnap was later to subscribe. However, as was said

27 28

Carnap (1928: §56). Carnap (1928: §57).

42

Arto Siitonen

above,29 Carnap thinks that the cultural sphere is autonomous. Cultural objects are not built up from mental ones, let alone out of physical ones. This is shown by the fact that a cultural object – say a state – a political organization – or habit can change its carriers (persons) without losing its identity but mental occurrences are always bound to a certain given person. This means that no statement concerning cultural objects can be translated into a statement concerning a physical or a mental object.30 Carnap repeats this point in §56.31 There he says that natural science may lead one to think that a state, a custom or a religion consists of mental occurrences, correspondingly as a piece of iron consists of molecules. This is countered by the attitude typical to the representatives of human sciences who reject that reductionistic view. According to the conventional picture of Carnap as a sensualist, positivist, and physicalist, he would be expected to choose the view inspired by the natural sciences. However, the Carnap of Der logische Aufbau der Welt rejects that view and shows a predilection for cultural studies. As he says, he adopts “die genannte Auffassung der Geisteswissenschaften”.32 Carnap’s point is reinforced by his criticism of an error of thinking which he calls “Sphärenvermengung” – the mistake of confusing the spheres of knowledge. This mistake is analysed by him in §30.33 In order to avoid the confusion, one has to keep the spheres apart and be careful in the application of concepts. Carnap compares this remedy to Russell’s theory of types and says that his spheres of objects are Russell’s types, as applied to non-logical concepts.

29 30

Sec. 2. Carnap’s expression is not especially clear: Das besagt, daß kein geistiger Gegenstand in eine Aussage über einen physischen oder einen psychischen Gegenstand mit Sinn eingesetzt werden kann. Carnap (1928: §23).

31 32 33

Carnap (1928: §56). Carnap (1928: §56). Carnap (1928: §30).

Kosmos Noetos and Carnap’s Constitution of Cultural Objects

43

The relation of reducibility in Der logische Aufbau der Welt is meant by Carnap to mirror the method of scientific research in considering mental objects as the constitutive bases of cultural objects. According to the order of knowledge, mental objects are primary and cultural objects are secondary.34 In the system of constitution presented in Der logische Aufbau der Welt, the sphere of mental objects functions as the basis. Carnap considers also the possibility of choosing the physical sphere as the basis.35 This may help one to understand why Carnap was later to adopt the position of physicalism.

6. Critical Comments Carnap’s Der logische Aufbau der Welt is a valuable and thought provoking work. He admits that his system is not a ready one but partly a sketch – especially concerning the higher levels. That these have not been extensively elaborated by him, is according to Carnap due to the state of the psychology and phenomenology of cultural knowledge which have not been as much cultivated as the psychology and phenomenology of perception.36 The more painstaking and pedantic are Carnap’s analyses of the physical level. What Carnap says of confusing the spheres (Sphärenvermengung) as mistaken thinking, is relevant and could be considered in works concerning fallacies. As the bibliography of Der logische Aufbau der Welt indicates, Carnap is wide-hearted in the choice of sources of his research. He has consulted, among others, works by Henri Bergson, Franz Brentano, Ernst Cassirer, Wilhelm Dilthey, Nicolai Hartmann, Edmund Husserl, Alexius Meinong, Friedrich Nietzsche, Max Scheler, Max Wertheimer and Wilhelm Wundt.

34 35 36

Carnap (1928: §56). Cf. Carnap (1928: §§ 59, 62). Carnap (1928: §150).

44

Arto Siitonen

It was only later that he became more exclusive and strenuous in his orientation. Carnap’s view on the themes of interpretation and understanding raises questions. The concepts of documentation and manifestation fruitfully make these themes more concrete and understandable, and Carnap’s semiotic considerations strengthen his approach. His view of the multifaceted theoretical background of the use and interpretation of signs is realistic, as is the point that sign relation is always somewhat conventional. This all seems to deliver objective criteria for the evaluation of the validity of interpretations. When he nevertheless claims that interpretation is a metaphysical theme because it belongs to the so-called problems of essence,37 he seems to bypass problems which are highly relevant. – It appears that metaphysics is for Carnap a kind of “wastebasket” into which it is convenient to throw not only various speculative theories but also problems. There is a certain indefiniteness in Carnap’s notion of autonomous kinds of objects. At the outset, the spheres of objects are: the physical, the mental, and the cultural. Then he says that there are further autonomous kinds of objects.38 Examples of these are: logical objects, mathematical objects, spatial shapes, colours, sounds, biological objects and ethical objects. What is one to say to this ramification? It appears counterproductive, because it unnecessarily complicates the issue of spheres. Is not this multiplication of kinds of entities against the principles of empiricism? Does it not confuse the original clear-cut tripartite division? Colours, for instance, are something physical and, as experienced, something mental, whereas colour concepts are cultural. In general, one may consider various objects from divergent points of view, without committing the error of confusing the spheres. Thus, a work of art can be thought as a physical object and as a cultural object, and one may switch from one viewpoint to another one. This happens when a restorateur repairs a damaged painting, for instance. One may wonder which role social relations and institutions play in Carnap’s world. In fact, Der logische Aufbau der Welt is somewhat 37 38

Carnap (1928: §169; cf. also §§ 20, 21). Carnap (1928: §25).

Kosmos Noetos and Carnap’s Constitution of Cultural Objects

45

indeterminate concerning the social. The question of spheres rises here again. To clarify the issue, let us take a look at Figure 3 above. The mental sphere plays two roles: as the basis of constitution and as constituted. To begin with, it is as it was purely mental, and after the second step it is turned into a psychophysical whole. It has an outward and an inward aspect. The former is connected to how I appear to myself through perception in my bodily constitution, and furthermore, how other persons appear to me. The inward aspect is in question, when I reflect on my mental states. The outward aspect is concerned, when I attribute mental states to other persons and interpret their gestures, mimics and movements as indicators of their mental processes. They are supposed to resonate correspondingly in respect to me. The question is: does this guarantee that we really share something – that the combination of the aspects would open for us a genuinely intersubjective area, a social reality? According to Carnap, “my world” and “the world of another person” are coordinated in relation to each other, and this lays the foundation for the intersubjective world.39 He does not use the word ‘social’ at all, but he speaks of cultural, historical and sociological objects,40 which belong to the highest level.41 Is there a difference between ‘social’ and ‘sociological’? If not, then social relations belong to the cultural level. If there is a difference, then social relations are supposed to be established at the level of psychophysical objects. But this is questionable under Carnap’s premises. How can I have an access to other minds as minds? Carnap speaks of intuitive understanding of other persons.42 According to,43 intuitive understanding is not knowledge. This seems to exclude sympathy and empathy as genuine forms of knowledge and thus to block the way to true intersubjectivity. – Concerning the issue of social relations was David Hume, with his idea of “fellow feeling“, a more comprehensive empiricist? 39 40 41 42 43

Carnap (1928: §§139, 146). Carnap (1928: §23). Cf. also Carnap (1928: §151): “sociological groups” (such as family, tribe, or state). Carnap (1928: §21). Carnap (1928: §181).

46

Arto Siitonen

7. On Related Theories of Cultural Objects In Plato’s philosophy, kosmos noetos is the intelligible universe which contains the eide or the ideas, and the visible image of that universe is kosmos aisthetos. These universes are – to use a Carnapian expression – coordinated to each other. The intelligible universe contains the first principles of being and knowing. Plato analyses in his dialogue The Republic44 a reasoning procedure which is supposed to reach these principles. It begins from a hypothesis and “leads to a beginning that has no hypothesis...”45; “...it comes as far as that which is no hypothesis, to the first principle of everything.”46 This upward reasoning has its converse in a descent which proceeds from a hypothesis and derives its consequences, until it arrives at the original object of inquiry which is something sensible.47 One may see that in Plato’s kosmos, the relation of dependence is reverse to that in Carnap’s world. The sensible physical world has conceptual entities – ideas – as its conditions of being. The connecting link between Plato and Carnap is their use of certain intellectual operations which complement each other. Bernard Bolzano distinguishes between a spoken or written – physical – sentence, a sentence which is entertained in one’s mind, and a sentence as a conceptual entity. Sentences may be true; a true sentence has, correspondingly, a physical, a mental, and a conceptual appearance. The objective conceptual content of a sentence is called by Bolzano “sentence in itself”. A special class of sentences in themselves embody truths in themselves.48 Correspondingly as sentences consist of words, sentences in

44 45 46 47 48

Quoted as Pol. Platon (Pol: 510b) Platon (Pol: 511b). Platon (Pol: 510d). Bolzano (1981: §§ 19 – 23; 25 – 27).

Kosmos Noetos and Carnap’s Constitution of Cultural Objects

47

themselves consist of ideas-in-themselves.49 According to Jan Berg,50 a “proposition in Bolzano’s sense is a structure of abstract ideas.” Friedrich Kambartel51 compares Bolzano’s and Plato’s views. He admits that Plato’s concept of idea can with certain qualifications be coordinated to the concept of idea-in-itself but notes that there is no Platonic equivalent to sentence in itself. – As to Carnap’s Der logische Aufbau der Welt, there is no mention of Bolzano in it. Nicolai Hartmann’s book Das Problem des geistigen Seins appeared originally in 1933;52 thus, it could not have been mentioned by Carnap in Der logische Aufbau der Welt. But one of Carnap’s sources was Hartmann’s work Metaphysik der Erkenntnis from 1921, and there is a critical remark on that work in §163 of Der logische Aufbau der Welt. Hartmann meant his 1933 book to contribute to the foundations of philosophy of history and of human sciences. To the object of the latter belong literature, art, language, knowledge, religion, morality, law etc. Hartmann stresses Dilthey’s significance as a pioneer of a philosophical study of these fields. According to Hartmann, “geistiges Sein” is a dependent but at the same time autonomous sphere of reality.53 It is superimposed on the mental sphere; Hartmann’s expression is “Überbauungsverhältnis”.54 The other spheres of reality distinguished by Hartmann are that of the physical and that of the organic. It characterizes the relation between the mental and the geistig that no individual consciousness can adequately grasp the various contents of the latter.55 Moreover, consciousness is something that isolates persons from each other – nobody can share consciousness with another person – whereas Geist is a common property of persons and unites them.56 There is a recent study on Hartmann’s ontology by Ingvar Johansson. He calls the 49 50 51 52 53 54 55 56

Bolzano (1981: §§ 48 – 56). Berg (1962: 70). Kambartel (1978: XV). Hartmann (1962). Hartmann (1962: 62). Hartmann (1962: 66). Hartmann (1962: 302 ff). Hartmann (1962: 303 ff).

48

Arto Siitonen

Hartmannian strata of reality inorganic, organic, psychic and spiritual.57 He relates Hartmann’s views to recent analytical philosophy and to new scientific discoveries. Karl R. Popper in his book Objective Knowledge58 makes the same division of strata of reality as Carnap in Der logische Aufbau der Welt. Popper calls these spheres world 1, world 2 and world 3. Surprisingly – unforgivably – he does not mention Carnap’s contribution at all, despite the same basic division. Popper considers Bolzano’s concepts of sentences and ideas on themselves; even more than the Bolzanian ideas, he stresses Gottlob Frege’s distinction between subjective acts of thinking and objective thought contents. “My third world resembles most closely the universe of Frege’s objective contents of thought”59. Similarly as Carnap, Popper considers the mental level “as the mediator between the first and the third”; “...the mind may be linked with objects of both the first world and the third world”60. Like Carnap, he thinks that world 3 is “autonomous in spite of the fact that it is our product”61. However, there is one important difference between Popper’s and Carnap’s views. According to Popper, world 3 “has a strong feedback effect upon us; that is to say, upon us qua inmates of the second and even of the first world.”62. This effect could help to solve the problem discussed in the previous section: social institutions into which we belong and may join can be expected to make us sociable and thus build up our social relations.

57 58 59 60 61 62

Johannson (2002: 196). Popper (1981). Popper (1981: 106, 156). Popper (1981: 155). Popper (1981: 112). Popper (1981: 112, 117, 119, 122).

Kosmos Noetos and Carnap’s Constitution of Cultural Objects

49

References Berg (1962): Berg, J. Bolzano’s Logic, Stockholm 1962. Bolzano (1981): Bolzano, B. Wissenschaftslehre 1, 2. Neudruck der 2. Auflage Leipzig 1929. (1837), Aalen 1981. Carnap (1928): Carnap, R. Der logische Aufbau der Welt, Berlin 1928. Hartmann, Nicolai (1962): Das Problem des geistigen Seins. Walter de Gruyter, Berlin (1933). Johansson (2002): Johansson, I. Hartmann’s Nonreductive Materialism, Superimposition, and Supervenience, Axiomathes (2002) 12: 195-215. Kambartel (1978): Kambartel, F. Einleitung des Herausgebers, in: Kambartel, F. (ed.) Bernard Bolzano: Grundlegung der Logik, Hamburg 1978. Platon (Pol): Plato, The Republic, transl. by A. D. Lindsay, London 1964. Popper (1981) Popper, K. R. Objective Knowledge, Oxford 1981. Russell (1954): Russell, B. The Relation of Sense-data to Physics, in: Rusell, B. Mysticism and Logic, Harmondsworth 1914, pp. 139-170.

Mental Causation and the Notion of Action Wolfgang Detel, Frankfurt

1. Introduction The standard theory of action distinguishes actions from mere behaviour by claiming that actions are behaviour that is describable under, and caused by, contentful mental states like intentions and beliefs. Are we supposed to conclude, then, that collective actions are not only describable under, but also caused by collective mental states? It seems that we are1, although this seems a little mysterious. This is small wonder, for the distinction between action and behaviour implies, among other things, a notion of mental causation, which remains puzzling. Indeed, one of the strongest reasons for asserting the naturalization of mental properties is the assumption that the problem of mental causation can only be solved if they can indeed be naturalized. But the most powerful programs of the naturalization of the human brain's mental properties seem to threaten the distinction between action and behaviour and thus the autonomy of human agents. Basically, the problem of mental causation is that it is difficult to see how the claim that specific mental properties (like being conscious or having semantic content) have autonomous causal powers can be consistent with the worldview offered by the natural sciences. Recently there has been some stimulating research devoted to the matter that looks quite promis-

1

See, for instance, Tuomela (2007) who says that people perform x together partly because of their joint intention to perform x, and that this purposive generation of collective actions always involves a causal notion (he therefore calls his account a purposive-causal theory of joint action). In Tuomela’s view joint actions are an important sub-class of collective actions.

N. Psarros and K. Schulte-Ostermann, Facets of Sociality, Frankfurt 2007: 51-83; ontos verlag

52

Wolfgang Detel

ing.2 In this paper I shall try to sketch an approach that works with, and develops further, a number of ideas articulated in this research – above all the idea of essentialist naturalism, the idea of causal powers of determinables, and the idea of a downward causation and of causal filters. I think that these enable us to make a clear distinction between behaviour and action, which has a number of consequences for our understanding of collective actions. Current discussions on mental causation are in the grip of the “exclusion problem”. This is normally expressed as follows: The physical realm is complete; the physical and the mental are distinct; and the mental is multiply realized in the physical. It follows that there are sufficient physical causes for every physical fact, and in particular for every behaviour, and thus that there is nothing left for the mental to do – the causal powers of the mental seem to drain away. Note that the same notion of causation is used to characterize both physical and mental causation. Furthermore, the exclusion problem would arise even if we had a good account of mental causation at hand; namely in this case the completeness assumption implies an overdetermination claim, which again threatens to drain away the causal powers of the mental. Thus, the challenge is to provide an account of mental causation that permits us to avoid overdetermination assumptions,3 and this is what I will attempt to do in the first sections of this paper.

1. Essentialist Naturalism In the past the basic idea of a modest naturalism of the mental was usually described in terms of a “supervenience relation”. The claim was that mental properties supervene on physical properties. It turned out, however, that an unqualified supervenience relation is too weak to save even the most

2

3

It is true, though, that some of the most influential recent works on mental causation, such as Beckermann et al. (1992) or Kim (1998), seem to be rather sceptical about our ability to solve the problem of mental causation. In her interesting paper Why the Exclusion Problem Seems Intractable, and How, Just Maybe, to Tract It Karen Bennett expands on this thought. Bennett (2004)

Mental Causation and the Notion of Action

53

basic physicalist intuitions,4 and that it does not provide any clues about the nature of the relation between the physical and the mental. A more promising approach, and one that is widely accepted today, is to say that mental properties (like being conscious in some way or having some specific semantic content) are realized in corresponding physical properties. One way of defining the idea of realization is to say that if a system S is in a mental state F at time t, then F can be said to be realized at t by the physical state G if and only if S is in G at t, and G has in S all the features (monadic, relational etc.) that are characteristic for states of kind F.5 However, the claim that a property G has all the (monadic, relational etc.) features that are characteristic for states of kind F seems to imply that all Fs are necessarily Gs. So we can propose what has been called the strong supervenience thesis: (1) Necessarily, for every x and every mental property M of x, x has some physical property P such that necessarily all Ps are M.6 But we must also capture the widely accepted idea that mental properties are (necessarily) multiply realized in physical properties: (2) Necessarily, for every mental property M, and every physical property P which necessitates M, possibly something possesses M but not P. As Yablo has correctly argued7, from (1) and (2) it follows (3) Necessarily, something has a mental property M iff it has some physical property P such that M is asymmetrically necessitated by P. Since this approach proposes a necessary metaphysical relation between mental and physical properties, it can be called essential naturalism. One well-known version of essentialist naturalism works with the idea that the necessitation of mental properties by physical properties is due to natural laws, and is therefore a nomological realization. This idea certainly offers a clear conception of how exactly the mental is necessarily realized in 4 5 6 7

For instance, this claim does not by itself exclude the possibility that some physical properties supervene on mental properties. See Beckermann et al. (1992), Introduction. See Kim (1984). See Yablo (1992).

54

Wolfgang Detel

the physical.8 However, nomological realization presupposes that there are psychophysical laws, and this assumption seems doubtful at least. In particular it would have to be possible to analyze the mental property in terms of those notions that are used to describe the natural laws that make the realization possible, but often this cannot be done.9 This objection has recently led to a revival of identity theories claiming that mental properties are in some sense identical with, and so necessarily related to, physical properties, without natural laws holding between the mental and the physical realm.10 One particularly attractive way of putting an identity theory has been offered by Frank Jackson. His basic idea is: (4) A system S having a property Q is necessarily realized in S having one of the properties Pi if (a) every phenomenon in our world that makes it true that S has a Pi also makes it true in our world that S has Q, and if (b) in every possible world that is identical with our world in containing the fact that S has a Pi it is also true that S has Q.11 We can talk here of a sort of structural realization. Properties are conceived of as structures of things that are realized in other structures. Thus, the property of a set of gas molecules of having a certain temperature is multiply realized in the properties of kinetic energy that each of the molecules displays; or the relation between two persons A and B such that A is greater than B is multiply realized in the various exact mathematical relations between A and B. Likewise, different algebraic groups multiply realize the general pattern of the axioms of group theory. And if so-called functional laws in physics of the general form Q = f (P), where Q and P are measurable parameters, are specified in concrete measurements Q* = f (P*), then the general law Q = f (P) is, as a pattern, multiply realized in its specific forms Q* = f (P*).12 Furthermore, maybe the property of being an animal is realized in the property of being a cat, but it could also be real8 9 10 11 12

Compare Beckermann (2002) for an extensive discussion of this idea. This is argued convincingly in Block and Stalnaker (1999). Papineau (1993), Block and Stalnaker (1999). Jackson (1998). I am talking here, of course, not about types but about tokens of patterns. Compare Armstrong (1978); Johansson (2000).

Mental Causation and the Notion of Action

55

ized in the property of being a dog; and the same goes for the properties of being coloured, red and scarlet.13 It seems that in most cases if Q(S) is necessarily realized in Pi (S), in the sense of (4), then Q is a more general property or pattern than each Pi. The fact that some given specific properties necessarily realize a more general property can of course be a matter of scientific discovery. If a more general property Q of a thing S is necessarily multiply realized in a set of more specific properties Pi of S, then S’s having Q is surely ontologically dependent of S’s having P1 or ...or Pn, and S could not be a Q-thing without having either P1 or P2...or Pn. Jackson thinks that in all these cases of a necessary metaphysical structural realization S’s having Q can be completely reduced to, and is therefore strictly identical with, S’s having either P1 or P2...or Pn. Thus Jackson maintains a reductionist position.14 But this is wrong. Property Q can, for instance, provide forms of information different from the disjunctive „property” P1 or P2...or Pn, and in most cases there might exist, or occur in the future, more properties than just the Pi that necessarily realize property Q. Furthermore, as Block pointed out, different laws may operate in Pi and in B.15 We must therefore amend (4) in order to save the intuition of an asymmetrical and necessary multiple realizations of (tokens) of properties in other (tokens) of properties. We could do this by suggesting: 13

14 15

If I talk about properties being realized in other properties, then this is always to be taken in the sense of (4), i.e. as talk about tokens of properties: Strictly speaking it is things having properties that realize things having other properties. But in fact things may be more complicated. According to Millikan (2000) the animal example is not a particularly good one – I put it in too Aristotelian a fashion. Animal is perhaps more like a template, not a genuine kind. Or it is a historical kind, the members of which do not have any „inner” nature in common. I would still insist that (a) even historical kinds supervene, and are multiply realized by, the different individual evolutionary histories of each individual member of a historical kind, and that (b) if this is correct we should look at historical kinds – and ask the question: how are they related to ahistorical physical kinds? See similarly Millikan (2000: 10, fn. 4). Block (2003).

56

Wolfgang Detel

(5) (a) = (4) above; (b) There are cases in which S’s having Q cannot be reduced to, and is therefore not strictly identical with, S’s having either P1 or P2...or Pn. This is an alternative way of putting claim (3), if this claim is taken to refer to structural (not nomological) realizations. The idea expressed in (3) and (5) can also be articulated in terms of the well-known distinction and relation between determinate and determinable properties (or patterns). In a recent paper Ingvar Johansson distinguishes between conceptual and ontological determinates and determinables. While I am interested here mainly in ontological determinates and determinables, it may be helpful to indicate the way Johansson defines the distinction on both levels: (6) (a) If a determinate concept (e.g. red) can be truly predicated of something, then at least one determinable concept (e.g. coloured) must necessarily also be truly predicable of exactly the same thing; (b) if a determinable concept (e.g. coloured) can be truly predicated of something, then there must necessarily be some (though unspecified) determinate concept (e.g. red) that is also truly predicable of exactly the same thing. (7) (a) If a particular has a determinate property, then this entails that the particular has the determinable property. Necessarily, if a thing is triangular, it has shape; necessarily, if a thing is red, it has colour. (b) If a particular falls under a determinable, then this entails that it has one of the corresponding determinate properties, although it does not entail which. Necessarily, if a thing is coloured, it has a particular colour.16 Of course, if we want to exploit the theory of determinates and determinables for a discussion of essentialist naturalism and the problem of mental causation, it is not sufficient just to introduce the distinction between determinates and determinables; we also need the ontological thesis that not only determinates, but also determinables exist as universals. As Johansson has convincingly shown, this thesis can indeed be defended within the framework of immanent realism. That is to say, we can indeed proceed from the assumption that, for instance, propositions such as This table is coloured or The gas in this container is governed by the Boyle16

See Johansson (2000: 1-2), and Armstrong (1978: 112).

Mental Causation and the Notion of Action

57

Mariotte law, where the notions coloured and the Boyle-Mariotte law refer to determinable universals, can describe real facts in the universe. It follows that we can understand claims (3) and (5) above as saying that, ontologically speaking, mental properties are related to the corresponding physical properties as determinables to determinates. A determinable property is strictly the same in any of its determinate properties: there is this identity in difference. What is more, it is this identity that creates the metaphysical necessity of the relations between determinates and determinables, and, in particular, between physical and mental properties. In fact this is exactly what Yablo suggested in commenting on claim (3). Following Kripke, Yablo makes this point with regard to pain, among other things: suppose that a certain pain is realized by a brain state b that is characterised by a very specific and rich molecular configuration; then since it is conceivable that this pain could have been realized by a slightly different brain state b* that is, however, extremely similar to b, it becomes intuitively irresistible to claim that the pain is possible even in b’s absence.17 Indeed, as has often been emphasized, on the whole mental properties of brain states or other physical things seem to be more general than the corresponding physical properties of these brains or things. For instance, the brain property of a cat that is part of the cat’s seeing my dog Fritz is a quite general one, since it is compatible with a great number of slightly different circumstances in which the cat sees Fritz – for instance slight differences in distance, angle, or brightness, under all of which the cat sees Fritz and which are surely realized in different biological brain states and physiological states of the cat. So the claim that the cat’s seeing Fritz is necessarily multiply realized in more specific physical or biological relations comes down to proposing that these relations are multiple determinate realizers of mental properties as their determinables.18

17 18

See Yablo (1992: 269). This move implies, of course, the claim that zombie worlds are impossible. If the physical realizers of mental patterns are assumed to be rich enough, then they necessitate the mental patterns. For a recent defence of this claim compare, e.g. Block and Stalnaker (1999), Yablo (2000). In her most recent paper (see note 1) Karen Bennett also argues in favour of this sort of “compatibilist” view of mental

58

Wolfgang Detel

2. Causal Powers of Determinables Even if the picture developed so far of the ontological relation between determinables and determinates in general, and between mental and physical properties in particular, is acceptable, it by no means follows that this is of much help for a new approach to the problem of mental causation. The main reason is that usually it is said of causal powers that if some properties realize some other properties in one way or other, it is the realizing properties alone that are the causal agents. However, recent discussions have shown that this consideration is, to say the least, an oversimplification. It may be helpful to offer first some examples from physics in support of this – that is, examples that have nothing to do with mental properties. Indeed, modern physics often theorizes about dominant properties shared by different systems by assuming that other parameters characterizing the difference between the systems are irrelevant for the behaviour of the systems in question.19 If we want, for example, to explain the common behaviour of pendulums, then we actually want to explain why pendulums with bobs of different colours or rods of different lengths, of different masses, or composed of different materials all have periods that are directly proportional to the square root of the length of the rod to which the bob is attached. Basically the explanation consists in solving the equation of motion for pendulums. But in this equation the only relevant factors are the acceleration due to the effect of gravity on the bob, and the length of the rod. Why are other fac-

19

causation which relies on a necessary relation between physical and mental patterns. Maybe it is much too simple to look at the ontological relation between the physical and the mental just as a relation between determinates and determinables. One reason might be that patterns of physical things and pattern of, say, biological or mental things are quite different, because physical patterns form ahistorical kinds, while biological and mental patterns form historical kinds (see Millikan (2000)). But if this is right, this does not seem to free us from thinking about the relation between ahistorical and historical kinds, and it does not seem to be inconsistent with assuming that historical kinds are multiply realized in ahistorical kinds. For a detailed account of this claim see Batterman (2000).

Mental Causation and the Notion of Action

59

tors such as the colour or material, or many other aspects of the micro level structure of the pendulum irrelevant for a causal explanation of the observed universal behaviour of the pendulum? If we write the laws of classical physics in their Langrangian form, then we describe the behaviour of a typical system, i.e. its Cartesian coordinates of position and the coordinates of velocity for the system’s components, as a function of the difference between the kinetic and the potential energy of the system. The question is then why only these common properties are causally relevant for the behaviour of the classical systems with respect to their space-time-position and their velocity. One plausible line of answer is that the different pendulums or other physical systems share some kind of common physical property in which the acceleration and the length of the rod play the crucial role. It is this common physical property that is supposed to be causally relevant for the proportionality which is our explanandum, but without the specific properties that make up the difference between the pendulums; and maybe realize the common property, being causally relevant at all. Other cases point to the same conclusion – for instance certain limit theorems in probability theory. These theorems describe asymptotic regularities in the form of a Gaussian distribution that is they describe the collective or universal behaviour of a sequence of independent trials, in which the individual contributions to the limit become more and more irrelevant as the number of trials increases. A third area that could be looked at is thermodynamics, for instance the behaviour of different fluids at so called critical points where some universal behaviour of these fluids is quite independent of their micro structural constitution. This universality is the result of the system's stability under a state of perturbation of the underlying subvenient details. Interestingly, in some of these cases also it seems that there are physical reasons why this independence holds: it is possible to discover a physical property shared by the realizers that is responsible for the universality at the macro level. There are many other rather trivial examples one could point to in this context. For instance, if buildings collapse as a result of earthquakes that register more than 5 on the Richter scale, then intuitively, and according to basic physics, the value >5 can be taken to be the cause of the collapse; the fact that an earthquake more precisely registers 5.7 or 6.3 is irrelevant. But

60

Wolfgang Detel

according to the exclusion principle, since the values 5.7 or 6.3 realize the value >5, the property of having the value >5 is robbed of any causal power. But if more precise measurements are made, then the value 5.7 or 6.3 is surely realized by, for example, the value 5.734 or 6.398 respectively, and so on. In the end we arrive at the conclusion that only the almost infinite most precisely specified states have causal powers. This is surely a strange view, and what is more, if there is no last most specified state for any more universal state (and many people think there is not), then according to the exclusion argument any causal power whatsoever drains away. The same is true, for instance, if we want to say that a bird pecks at a thing because it is red; but red is also realized by scarlet among other things, and again we can imagine a very long and maybe infinite series of more detailed realizers, so that only at the most specified level of red would a causal explanation be possible. Or again, if one doubts whether there is a last most specified red colour in which all other more general red colours are realized, then no causal explanation whatsoever would be possible. These examples can be backed theoretically by a critical remark about particle metaphysics. The familiar way of looking at causation, multiple realization, physicalism, etc. relies on reductive particles metaphysics. This claims that the basic entities about which physics talks are basic particles having basic internal properties. These entities can adopt certain relations, configurations or patterns that eventually make up new „levels” of organization. But the locus of causal powers is supposed to be in the particles and their internal properties, while their configurations or patterns do not seem to have causal powers of their own. This view supports reductionist moves by encouraging the reduction of what seem to be the causal powers of the configured systems on a „higher” level to the causal powers of the basic particles and their internal properties. However, modern physics tells us that there is a problem with this picture: there are no most basic particles having most basic internal properties. Rather, the most basic entities are quantum fields that have no existence independent of their patterns or configurations. Quantum fields are processes that can only exist in various patterns, and these patterns come in many different sizes and many different physical and temporal scales. To oversimplify somewhat – there are from

Mental Causation and the Notion of Action

61

top to bottom only patterns and configurations of processes. On all levels we must think of the most paradigmatic entities as being essentially configurational, and it is being configurational that is responsible for, and correlated with, causal powers. Change the configurations, and you change the causal powers.20 Thus there does not seem to be a problem with assuming that if new configurations emerge, then new causal powers emerge due specifically to the new configurations (from now on I will therefore talk about patterns or configurations of things instead of their properties). These configurations can also include external (and even historical) relations of the entity in question. In general, therefore, it seems relatively uncontroversial to say that macro level patterns have new causal powers relative to micro-based patterns. It seems to me that we can therefore maintain (8) There are cases in physics in which Q in S is necessarily multiply realized in P in S, in the sense of (5) – (7), such that the fact Q(S) has new causal powers that cannot be reduced to the causal powers of the realizing fact P(S). If we follow the well-known generalization argument that (8) applies to all special sciences, then this should also be true for psychology. In this way we arrive at the claim (8)* Mental patterns too can have causal powers independent of those of neuroscience.21

20 21

Compare the way Heil, following Martin, relates qualities (roughly configurations) to causal dispositions, in Heil (1998: 181-188). According to Kim, however, the generalization argument is mistaken because there is an asymmetry between the way psychology is related to neuroscience and, for instance, the way biology is related to chemistry. Genuine macro properties (such as those found in biology) have causal powers over and above the causal powers of their micro constituents, but, according to Kim, whereas biology and chemistry occupy different levels of organization, psychology and neuroscience are located at the same level of organization. Contrary to the generalization argument, in Kim's view, psychological causation (unlike e.g. biological or chemical causation) cannot be interpreted as a form of macro causation (see Kim 1998). This argument cannot be sustained, however, if it is true that mental properties relate to physical properties as determinables to determinates.

62

Wolfgang Detel

3. Downward Causation and Causal Filters in Physics The approach to the problem of mental causation I wish to recommend here relies on the picture of essentialist naturalism and the idea of causal powers of determinables as outlined above. But there is at least one further important ingredient we need to progress towards a possible solution. This ingredient is an acceptable notion of downward causation.22 Some of the cases in physics looked at in the last section indicate that macro level patterns can be causally effective without more specific patterns of the systems in question being causally effective. Admittedly, not in all these cases do the macro level patterns seem to be determinables of the more specific patterns, although in some cases this seems to be true. In any case, there are interesting specific cases in which macro level patterns of systems are causally effective specifically insofar as they activate some causal powers of their constituents, and sometimes block others. A familiar case in point are boundary conditions, or ceteris-paribus-clauses for natural laws which specify under which conditions the application of a natural law, and thus the activation of the causal power of an entity operating under that law, is possible, and under which conditions this is impossible.23 A simple example is a gas separated from its surroundings by a solid wall. This wall may be heat-conducting or heat-isolating, and it depends on this boundary condition how the temperature, pressure, and volume of the gas are related

22

23

The notion of downward causation has always been under attack by strong physicalists. The main concern has been that this notion implies the violation of laws of physics on the basic physical level (see for instance Kim (1992)). This is unwarranted, however. The examples discussed in recent literature on downward causation (some of which I quote) imply no such violation. This is something Fodor has long been urging. Fred Dretske has turned this thought into a specific form by distinguishing between what he calls structuring causes and triggering causes. Structuring causes are those causes that bring about conditions that must obtain for triggering causes to be triggered. As Dretske points out, the structuring causes rely on extrinsic patterns of things, i.e. on their relations to other states or things. See for instance Dretske (1993).

Mental Causation and the Notion of Action

63

to each other, and thus, how the motion of the microscopic parts of the gas is restricted, i.e. how the laws of physics are applied in this particular case. Another simple example is that the activity of a reagent can be affected by the presence of a catalysing enzyme that forms a complex with the reagent. More complex examples can again be found in the realm of thermodynamics. Thus on the basic physical level all microscopic laws are reversible under time change, but in thermodynamics we have irreversibility as described, for instance, in the law of increasing entropy. If we want to explain why human beings have only a memory of the past which implies a time arrow, then it will not suffice to point to the laws of mechanics and to consider a human body just as a collection of molecules that act together according to the laws of mechanics. Rather, part of the explanation must be to point to the fact that human bodies are inserted into ecological systems which are thermodynamic systems far from equilibrium, and which manifest a time arrow. Such systems have a tendency to relax towards equilibrium, but as long as they are maintained in a state of non-equilibrium by a flow of low-entropic energy from outside, organisms inserted in them are able to survive. If, however, the systems were isolated, they would run down to equilibrium, and the organisms would die. So again it depends on boundary conditions how the parts of the systems behave and which causal powers and laws are put to work.24

24

Compare Christiansen (2000). Many more examples are discussed in the recent paper by Bickhard and Campbell (2000). These authors also distinguish between linear and non-linear cases of downward causality (for example a wheel, as a linear case, and a candle flame, as a non-linear case), and argue that while linear cases do not mean trouble for the physicalist, non-linear cases do. In the latter cases we see far-from-thermodynamic-equilibrium stability persisting for a significant period of time, as in candle flames (as one of the most basic examples) or in (simple) biological organisms. Entities exhibiting this kind of stability are open organized systems that are involved in essential interactions with their environments, and these relations and their causal powers are not reducible to properties of lower-level constituents, but, on the contrary, exercise causal influence on their parts.

64

Wolfgang Detel

Robert van Gulick has recently generalized this point in the following form (where the notion of a causal filter is mine): 25 (9) Let x and y be entities such that x has the patterns P and Q, and y the patterns S and T; furthermore, let R be a relation between x and y that follows from the fact that x has the patterns P and Q, and y the patterns S and T; then if the relation R between x and y brings it about that only the fact that x has Q, but not the fact that x has P, exercises causal effects on y, R is a causal filter of the causal powers of x. We can use the notion of a causal filter to define downward causation: (10) Let x and y be entities such that x has the patterns P and Q, and Q is necessarily realized in P; furthermore, let R be a relation between x and y such that R is a causal filter that brings it about that the fact Q(x), but not the fact P(x), exercises causal effects on y; then R exercises a downward causation on the causal relations between x and y. According to (10), it can be due to downward causation that more general patterns (determinables) of a physical entity can have causal effects on their own, independently of the causal powers of the more specific (determinate) patterns of that entity in which the more general pattern is necessarily multiply realized, and where the more specific (determinate) patterns do not have any causal effects on that fact. The notion of a causal filter is meant to indicate that downward causation restrains, or provides conditions for activating, triggering causes.26 There are different sorts of causal filters, and thus, of downward causation. Campbell and Bickhard distinguish, for instance, four different sorts: (i) consequences outside of a system that are non-linear with respect to the lower levels of the system, but which nevertheless influence lower level external processes – this is among the weakest kinds. An example would be disturbances in the flow or air around a kite; (ii) constraints internal to a system that are non-linear consequences of the organization of the system, such as system-stability in far-from-equilibrium forms; (iii) non-linear con-

25 26

van Gulick (1993), Batterman (2000). This is more or less what Emmeche et al. (2000) call the medium version of downward causation.

Mental Causation and the Notion of Action

65

strains internal to the constituents of a system: for example, the processes internal to cells are strongly constrained by the overall processes of the organism (this kind of meta-internal downward causation can extend to the existence of complex molecules that would not exist otherwise); (iv) constraints on generative processes, i.e. sources of construction variation, such as changes in the organization of an eco-system that can alter the selection pressures on the constituent organisms (downward causation via selection).27 In criticising Kim’s most recent defence of the exclusion principle and of the „draining away” of the causal powers of the mental, Ned Block has articulated similar ideas under the heading of decomposition or multiple composition. Block points out, for instance, that the law of conservation and angular momentum operates on the general level of rigid bodies, no matter how this rigidity depends in detail on the patterns of the molecules making up rigid bodies. This law would remain in place if rigid bodies were made up by alternative decompositions in their mereological configurations. One of the examples Block uses is the relation of the temperature or pressure of a gas to its molecular composition. Summing up, Block puts the problem for Kim-like reductionists as follows: suppose we have good reasons to believe the following three claims: (i) Ua → Ub is a causal law; (ii) Ua has alternative decompositions in mereological configurations L1 and L2; (iii) L1 and L2 participate in different laws; then Ua is multiply constituted and realized, but would still be as such causally effective due to (i), while at the same time not being identical to the disjunction of L1 and L2 because of (iii). In this case, the nomological claim of law (i) is not undermined by the nomological claims at the micro level formulated in (iii). Block applies this idea explicitly to the relation of mental states or actions and their physical or biological realizers. He says, for example, that Cain’s strangling Abel would have caused Abel’s death even if the molecular realizations had been slightly different, so long as the difference was not significant at the macro level.28

27 28

Bickhard and Campbell (2000). See Block (2003).

66

Wolfgang Detel

4. Downward Causation and Formal Causality in Biology The notion of downward causation gets even more interesting if we look at biology. Donald Campbell once remarked that the selective system that works on the level of whole organisms operates by forming, and activating the causal powers of the parts of organisms, for instance by partly determining the distribution of lower level events and substances.29 Some biologists have recently supported this idea by pointing to illuminating cases such as cell organization, which is taken to be a web of causal relations. They show in great detail that the molecules that materially compose cells (DNA, proteins, fatty acids and so on) are fabricated within internal processes of the cell itself, and in this way are generated through the web of interactions of the whole cell. This web can therefore be called the cause of a given component of the cell and its behaviour. This is taken to be a clear case of biological downward causation.30 But in addition, there is a growing sense among scientists that in important biological cases there is a new form of causality at work, which has to do with a specific sort of pattern transmission. My impression is that ultimately this fact can be helpful for our understanding of mental causation. Sometimes this new form of causality is called formal causality, sometimes it is called information processing. In particular, DNA molecules are supposed to be formal causes of proteins in biological cells, since the specific sequence of nucleotides they contain is a form or pattern of these nucleotides that can be mapped into the primary structure of those proteins that are constructed by them. It is also illuminating to look at the way the notion of information is applied to this consideration. To be sure, we are not talking here about conventional information theory, which is a theory about technological communication systems that are designed by inten-

29 30

Campbell (1974). In a recent paper Campbell has re-emphasized this claim, see Campbell (1990). Compare, for instance, Moreno and Umerez (2000).

Mental Causation and the Notion of Action

67

tional human agents.31 Rather, we are talking about the roots of information that are related to natural systems.32 Indeed, in a recent paper on the concept of information in biology, John Maynard Smith proceeds from the assumption that the concepts of traditional information theory do not apply, for instance, to the DNA level, and that therefore the notion of information as used in biology cannot be captured by traditional information theory.33 His own idea is that in nature it is by natural selection that information comes into play: „In biology, the statement that A carries information about B implies that A has the form it does because it carries that information. A DNA molecule has a particular sequence because it specifies a particular protein, but a cloud is not black because it predicts rain”.34 The crucial thought seems to be that if the causal transition of a form or structure F of a thing A to a form or structure G of another thing B takes place due to natural selection, i.e. because A’s having F has been naturally selected since it leads to B’s having G, then B carries information about A, and A confers information on B. One important further point that Maynard Smith makes is that that connection between A’s having F and B’s having G is contingent in the sense that there is usually no feedback from the latter to the former (which indicates the irreversibility of information processing). In his comment on Maynard Smith’s paper, Kim Sterelny points out that the relation between A’s having F and B’s having G is not just a covariation. Furthermore, there must be a reading mechanism in B that reacts reliably to A’s having F, and as a result of this reaction causally produces

31 32

33 34

See the classical study of Shannon and Weaver (1972). One well known way to do this is to suggest that so-called natural signs S carry „information” about the „signed” thing T because T and only T caused S. But this is no more than a metaphorical way of talking about information. The same goes for the suggestion that if p(A/B) > p(A), then B carries the information A. This is in fact the usual probabilistic manner information is defined in standard information theory (compare for instance Dretske 1982). However, there is no theoretical gain in such a definition, see Putnam (1995: ch.19). Maynard Smith (2000). Maynard Smith (2000: 189-190).

68

Wolfgang Detel

B’s having G.35 Eva Jablonka has followed a different approach, and suggests not specifying, but generalizing Maynard Smith’s idea.36 Her proposal is that an entity E can be said to have information if a receiver system R reacts to E in such a way that this reaction changes the state of R in a functional manner and that variations of the form of E are mapped consistently into variations of the form of R. In short, the general picture seems to be this: If a spatial or temporal pattern P in A triggers another spatial or temporal pattern P* in B (as a result of a reading mechanism or a capacity of pattern recognition in B) such that (i) there is a 1:1 correspondence between P and P*, (ii) the change is unidirectional, always going from A (the source) to B (the receiver), and depends almost exclusively on the pattern of A, being to a large extent independent of other circumstances, (iii) the change is discontinuous in the sense that if the pattern in A is changed in a steady continuous way, the response in B may not vary in a steady continuous way at all, or very small changes in the pattern of A may lead to drastic changes in B, and (iv) the energy required for the participating physical process must be provided from an external reservoir (i.e. is external to the interaction process), then we can call P* an image of P, and we can say that B received the information P from A and processed and encoded it into P*. If P* is physically identical with P, then the interaction is a replication, and P*(B) is a copy of P(A).37 This way of looking at information is inconsistent with the widespread idea that a structure of a thing as such, or a purely causal relation as such, or a purely physical interaction as such, can be some sort of information. The process of processing information on this basic level is, however, sometimes more complicated than the description outlined above suggests. For so far we have talked only about a simple transition from an external object A exhibiting a certain pattern P to a pattern or behaviour P* of, for example, a large carbon-based molecule. However, the capacity of these large molecules to react to the patterns of other objects also extends to 35 36 37

Sterelny (2000). Jablonka (2002). For a more detailed account see Roederer (2002).

Mental Causation and the Notion of Action

69

other large carbon-based molecules that themselves already exhibit specific patterns (mostly in the form of specifically structured molecular chains). Thus these molecules are able to code information not only from other external objects, but also from each other, so providing a translation of this information into other and partially different molecules. One of the most basic forms of such a translation is, of course, the replication, reproduction, or copying such as is found in the case of DNA-strings; due to mistakes in the course of the replication process, this can lead to new patterns. It is in this way that we can say that, under these conditions, the interactions between these molecules and their environment (including other molecules) is, at the most basic level, controlled by information and information-processing operations in a way that does not occur in the realm of pure physics. Should we say, as some do, that information, thus conceived, is an agent that controls certain physical operations and is, therefore, an irreducible entity? This way of talking about information seems oversimplified, if not wrong. Information, thus conceived, must instead be taken to be a new sort of physical causality which can indeed be called formal causality. To be sure, if we proceed from the picture of configuration metaphysics, it is always configurations, and thus in a sense forms, that are related to causal powers. What seems special about the level of living things is that configurations are causally transmitted into other configurations by means of pattern recognition such that there is a 1:1 mapping between these forms. For information processing that is faster and more sophisticated, and also for active memory storage of current events, organisms need a central nervous system, which some organisms have developed in the course of evolution. This system facilitates a second type of information besides biomolecular: neural information. This type of information also satisfies conditions (i) – (iv) sketched above, but it is a reaction of organisms that have a central nervous system, and results either in specific spatiotemporal distributions of electrical impulses in the neural network, or in a spatial distribution of the synaptic architecture in the neural tissue. The result of the former is a dynamic transient form that requires a continuous supply of energy to be maintained; the latter is longer-lasting and static, and can, in principal, subsist without any continuous flow of energy.

70

Wolfgang Detel

At this point it may be helpful to draw attention to connectionism as a branch of a theory of the brain. According to the connectionist picture of the brain neural networks are able to recognize – and react to – patterns of input stimuli and map them into patterns of outputs and motor reactions. Indeed, neural networks seem to have the specific capacity of reacting and even of learning to react, to general patterns that certain different patterns of incoming stimuli, such as photons, have in common. In this way these patterns are certainly multiply realized in more determinate patterns of the stimuli in question. According to connectionist theories this also seems to be true if the stimuli flow from the creature's own brain, i.e. if there is internal information processing in the brain. The neural network can also react to internal patterns that can be multiply realized by more specific physical patterns, such as the electric impulses of the brain stimuli in question. In these cases, the pattern recognition of neural networks entails a sort of formal causation, and enjoys the function of a causal filter for selecting the causal powers of internal mental patterns.38

5. A Picture of Mental Causation It seems to me that the material I have presented so far suggests an acceptable notion of mental causation. Specific cases of downward causation involving formal causality flow from certain causal relations between physical entities and representational creatures. For instance, frogs can represent (that is, perceive) little black moving things in their environment. In normal circumstances, these things are flies, but obviously, the pattern of a fly consisting in being small, black and mostly moving is more general than its

38

Compare the following remark in Bechtel (1994: 204): “In feed forward networks, the connections (sc. between neural units) allow the network to respond to patterns activated on the input units by activating patterns on the output units. Networks can be trained so that a variety of different patterns all result in activation of the same output pattern... As a result, such networks can readily be interpreted as performing activities of pattern recognition and categorization.” It should be added that this notion of information does not necessarily imply the richer idea of representation, but the notion of representation can be built on the notion of information as outlined above.

Mental Causation and the Notion of Action

71

pattern as a fly. And this fly pattern belongs to the set of (determinate) patterns in which the more general (determinable) pattern consisting in being small, black and mostly moving is necessarily realized. Nevertheless, due to the relation that the frog sees the fly in a specific way, the frog reacts to a physical entity in its environment by snapping at it purely due to its determinable pattern, i.e. because it is a small black moving object, and not due to its determinate pattern as a fly. To look at a further example, the very fact that a certain gas such as air has a certain temperature at time t may be identical with the mean of the kinetic energies of its molecules at t; but surely having this specific temperature could also be identical with other individual energy states of the gas molecules, provided the mean of these states amounts to the same mean kinetic energy. Thus, for a gas to have a certain temperature is a more general (determinable) pattern of the set of its molecules than a certain specific set of individual energy states of each of its molecules. At the same time, a gas having a certain temperature is always necessarily realized in one of the sets of specific energy states of each of its molecules that are compatible with having this specific temperature. But human lungs (and the associated brains) cannot discriminate between these different sets of individual kinetic energy states; rather, they can only discriminate, and react to, different temperatures of gases. So in this case too the determinable pattern of the surrounding air of having a certain temperature exercises a causal effect on human lungs, but without the much more specific determinate pattern of being in a state that is defined by the sum of specific individual energy states of the air molecules having as such any causal effect on the lungs. I do not wish to deny that in these cases there is always a particular determinate physical specification of determinable patterns that is, in a sense, involved in causing a reaction of general pattern recognition, for instance photons coming from a fly, and not just from a little black moving thing. After all, determinables are ontologically dependent of determinates. The crucial point is, however, that the reaction of the frog is the same for many different determinate physical specifications insofar as these specifications are similar in some respect; for instance, the patterns of photons coming from flies and from other little black moving things are sufficiently similar, i.e. share a determinable pattern, such that the frog reacts in the same way

72

Wolfgang Detel

to all patterns of photons sharing this determinable. It seems to me that we can describe this as a reaction to the determinable which they share. So we can state (11) There are cases of formal downward causation, especially if there are entities x and y such that y is able to mentally represent x so that, due to this relation of representation, y reacts causally to a general pattern of x without reacting, and indeed without being able to react, to more specific patterns of x as such in which the more general pattern is necessarily realized. It should now be obvious how we can use the ideas developed so far to sketch a first general notion of mental causation. Mental patterns are determinable patterns of physical configurations of things like brains and objects in the external world that realize them as their determinates. And representing creatures as we know them can not usually discriminate between, for example, different biological brain states that multiply realize a certain more general pattern, if only because most representational creatures cannot perceive specific brain states. They can only react causally to determinable mental patterns as expressed in behaviour.39 So we can specialize (11) logically to introduce a first notion of mental causation: (12) There are cases of formal downward causation, especially if there are representational entities x and y such that y is able to mentally represent mental patterns M of x so that, due to this relation of representation, y reacts causally to M(x) without reacting, and indeed without being able to react, to more specific physical patterns of x and other physical things as such in which M is necessarily realized. In these cases the causal relation between M(x) and the reaction of y to M(x) is an external mental causation. In (12) we are talking about external mental causation because in the cases mentioned in (12) the fact that a representational entity has a mental pat39

Jaegwon Kim has argued, correctly, that the notion of mental causation implies the idea of downward causation. But he has at the same time denied, falsely, that the notion of downward causation makes sense. Kim fails to see that there are two different notions of causation involved if people talk about downward causation, cf. Kim (1993).

Mental Causation and the Notion of Action

73

tern has causal effects on another representational entity in the world external to the first entity. The problem of mental causation that is usually discussed today, however, is taken to be a problem of internal mental causation. It seems to me – although I can do little more than draw attention to it here – that we can model the idea of internal mental causation on the notion of an external mental causation by saying (13) If, under the conditions specified in (12), x is identical with y, the causation of physical reactions of x by M(x) is an internal mental causation. Are there cases of internal mental causation? This can hardly be doubted. In most representational creatures mental states, i.e. brain states having mental patterns, have causal effects on motor reactions. Certain psychological states (e.g. prolonged anxiety, embarrassment) can cause physiological effects (higher blood pressure, eczema, blushing); love relations can promote health; or viewing a film of Mother Theresa can significantly increase the salivary immunoglobin A level.40 There are devices in these creatures that enable them to react causally to some of their mental patterns, without being able to discriminate between the biological or physical brain states in which the mental patterns are necessarily multiply realized. For instance, a dog will react to seeing a big dangerous cat pretty much in the same way, independently of slight differences in distance or angle under which the dog sees the cat – differences that can of course be mapped 1:1 into specific different physical brain states, all of which will be compatible with the mental pattern of seeing the dangerous cat. It seems to me that this account does not imply any overdetermination claim. The general picture is that in cases of mental causation mental and some physical causes are independent from each other, but work together. Are the physical causes sufficient in these cases to bring about the effects? It simply depends on where we look. In simple cases like the occurrence of structural and triggering causes, for example, we can, of course, say that the triggering cause is sufficient, as long as we neglect how it got activated; but if we look at the conditions of activating it, the triggering cause will seem to be dependent on other causes, including mental ones. The diagnosis of the exclusion problem that follows from this way of looking at 40

Compare McClelland (1987: 366-368).

74

Wolfgang Detel

things is that the completeness assumption is too simple, or if put in a too simple form, is wrong. Seen from a certain perspective, we sometimes have to admit that mental causes are required to bring about certain physical effects. Furthermore, it transpires that the assumption of the identity of physical and mental causality which is presupposed in the formulation of the exclusion problem is also wrong. In contrast to usual physical causation, mental causation is formal, that is it is a processing of information. Therefore the exclusion problem does not arise for the account as outlined so far.

6. Specific Applications My account is obviously still incomplete. It represents at most a first step in approaching the problem of mental causation. It remains to be shown that the analysis can be applied to the specifics of mental patterns. For instance, we must be able to see that a physical state involving a physical brain state, and having a representational content or being conscious, can have causal effects due to the patterns of representing something or being conscious about something. And here lies the rub. Let me demonstrate the problem for the case of representation and semantic content. I assume that we must think about the pattern of having either sub-linguistic or linguistic semantic content in terms of extensionalist theories of content, for example teleosemantics or a broadly Davidsonian theory of meaning. According to these theories, the patterns of having semantic content do, of course, not only supervene on physical patterns of brains, but also on physical patterns of external objects that have interacted with the brains in the history of the organisms in question, i.e. on external relations of the organism. Having semantic content is therefore a pattern extending to all physical things or states that, in the evolutionary, cultural or biographical history of the organisms in question, contributed to the development of proper functions of brains in order to map 1:1 external states to corresponding brain-states. And it is surely difficult to conceive of this extensive physical foundation of semantic content as an event in the normal sense that can have causal effects.

Mental Causation and the Notion of Action

75

Reasons of space prevent me from dealing with this problem adequately here. I merely wish to draw attention to two current strategies on the market, which seem to be promising. One is to suggest that physical causality is of a different kind to mental causality, although at the same time in some cases both sorts of causality work together. Dretske’s distinction between triggering and structuring causes of behaviour was a first attempt to pursue this strategy. More recently Pettit and Jackson have proposed distinguishing between causal efficacy and causal relevance, and suggest that, for certain kinds of behaviour, mental states are causally relevant, while physical states are causally effective.41 The second strategy is to build on configuration metaphysics, i.e. on the claims that at all levels at which their configurations are essential for entities, these configurations are correlated to causal powers; and that they can include external relations. If externalist theories of the mind are then brought into play, for example teleosemantics, we arrive at a notion of wide causation42 by saying that, for instance, for a brain state to have a proper function, or to have semantic content, is essentially to be related to a pattern of evolutionary history.43 In explanations of causes of certain motor reactions, we must point to brain states qua having this sort of external relation; we must not think that, as far as the causal powers are concerned, this relation is accidental to the brain state in this particular case. I am inclined to think that the second approach is more attractive. If we proceed from this intuition, then we must first of all realize that semantic content – for instance on the sublinguistic level a certain function of a brain producing internal brain states that are mapped 1:1 onto external events – is certainly necessarily multiply realized in the physical foundation of semantic contents, since different individual organisms, the brains 41 42 43

See Jackson and Pettit (1990); Jackson and Pettit (1990a); Jackson (1996); compare also the preprint paper by Becker (2003). Compare Yablo (1997). At this point the interesting theory of historical kinds proposed recently by Ruth Millikan could be introduced. As Millikan sees it, historical kinds (or patterns) of things are kinds, the similarity of which emerges not from ahistorical configurations, but from shared patterns in their historical relations to other individuals; compare Millikan (2000).

76

Wolfgang Detel

of which carry the same semantic contents, surely have slightly different evolutionary and biographical histories manifesting slightly different patterns. But these different patterns of histories exhibit sufficiently similar patterns to qualify for being selected for a connection with the same semantic content. And these more general patterns have something to do with the patterns the corresponding representational creatures can represent. In the sublinguistic teleosemantic case, it is therefore evolution itself that operates as a powerful causal filter for selecting semantic contents. So we can at least say (14) Put in terms of an externalist (teleosemantic) theory of semantic content: for a brain state to have sublinguistic semantic content (i.e. to have the proper function f of producing internal states that are mapped 1:1 onto certain external states) is for it to be necessarily multiply realized in the patterns of physical states of the brain, and of the physical states in the various histories of the organisms in question that contribute to the development of the proper function f. Furthermore, the causal filter for selecting f is evolution itself, or unconscious evaluation systems of the organisms in question. Another problem is how sub-linguistic contents are related to propositional contents, and how this relates to accounts of mental causation. It seems to me that mental causation of physical states resulting from their pattern of having propositional content can in principle be modelled on the suggestions presented so far. In a broadly Davidsonian theory of meaning, human brains can adopt proper functions to produce thoughts and sounds that can be mapped 1:1 onto external states. This adoption is based on a cultural history of interpretative interactions between human beings, and on a tradition of language use. It therefore involves more complex processes than in the sublinguistic case. For instance, it requires a mapping between the linguistic expressions of teachers and of learners that shapes the Ttheorems constructed by the learners; and it involves the construction of a holistic interpretation theory to transform empirically confirmed extensional T-theorems into intensionally adequate translations or specifications of meanings. In this case it is cultural tradition and forms of handing over ways of language use to the next generation that take on the role of causal filters selecting those general patterns of producing sounds, and reacting

Mental Causation and the Notion of Action

77

formally to these patterns, and that contribute to building up propositional contents as suggested by a broadly Davidsonian theory of meaning. It seems, then, that the sublinguistic way of describing things also indeed applies to linguistic cases, except that the histories that are causally relevant here are histories of cultural behaviour. Finally, as suggested both in teleosemantics and in a broadly Davidsonian theory of meaning, there is a further step in the development of the theory which consists in an account of the relation between semantic contents of sounds and the contents of inner thoughts. In conclusion I wish to discuss briefly whether there is any payoff of the ideas presented so far for understanding the notion of collective intentionality and collective actions. To do this I proceed from Tuomela’s account44 (not from Searle’s analysis), which can be put as follows: (15) A person P has the collective intention to do A relative to further persons P1,...,Pn iff P intends to do A and believes that all persons Pi intend to do A and that each Pi believes that all other persons P, P1,..., Pi-1, Pi+1...Pn intend to do A. (16) A set S of persons P1,...,Pn have the collective intention to do A iff each of these persons has the collective intention to do A relative to the other persons Pi (in the sense of (15)); each Pi intends to do a specific action AS as her contribution to A; each Pi believes that it is possible to do A; each Pi believes that doing AS is necessary for doing A, and that doing all specific ASs results in doing A; each Pi believes that all other persons in S satisfy conditions (ii) – (iv); each Pi believes that all other persons in S believe that all other persons in S satisfy conditions (ii) – (iv).;

44

In some of his works, e.g. in his contribution to this volume Tuomela (2007), Tuomela seems to call this joint action that is to be taken as a specific (though important) case of collective action.

78

Wolfgang Detel

Conditions (iii) – (vi) belong to the reasons and causes of each Pi for forming the collective intention to do A relative to the other persons (see (i)); and to do AS as her contribution to A (see (ii)). For example, a violinist has the collective intention to play a string quartet relative to the three other players (the second violinist, the viola player, and the cellist) iff she intends to play the string quartet, and believes that the other three intend to do this, and that each of them believes that the others intend to do this. Obviously, if one of these conditions were not satisfied, the first violinist would not intend to play the string quartet with these other players. And for the group of these four players to have the collective intention to play the string quartet not only must each of them have the collective intention to do this. In addition, each of the players must also intend to play her specific part (the part of the two violinists, of the viola player, and of the cellist, respectively) in the string quartet and must believe that this is necessary to play the string quartet together with the other three, and that if they all do so, this will result in playing the string quartet; furthermore, each of them must of course believe that the four players have the opportunity and the capacity to play the string quartet; and finally, each of the four players must believe that the other three intend and believe all this, and that each of them believes that all the others intend and believe all this. In definition (16) it is obviously the notions each Pi and AS that refer to different persons, for instance to the four different string quartet players. Therefore, the contents of the collective intentions of the different persons involved here are not completely identical. But it seems clear that these contents share a common pattern which we get from (16) (i) – (vi) by substituting variables for the expressions each Pi and AS. This common pattern is a determinable relative to, and therefore structurally realized in, the specific determinate collective intentions of the involved persons. So much about the ontological status of collective attitudes. But what about the role of mental causation in this complicated scenario? It seems clear, first of all, that this picture involves external as well as internal mental causation. Each person has some sort of internal mental causation that causes her own specific participation in the collective action. But each person is also epistemically related to all the other persons, for in-

Mental Causation and the Notion of Action

79

stance by mutual beliefs; and since we want to explain how given collective actions are caused by, among other things, collective intentions, we must assume that what each Pi believes according to (16) (iii) – (vi) is correct (otherwise the collective action would not come about), and so external mental causation is involved as well. What is more, the collective (or we-) intention (intention in the we-mode) is more than, and therefore not fully reducible to, the participation intention of each person, for instance because, as Tuomela remarks (correctly, as I believe),45 each we- intender, by intending to participate in a collective action, also endorses doing something together with others, and is wecommitted. Thus the notion of participation intention implies conceptually the notion of collective (or specifically joint) action. This conceptual relation is, it seems to me, in part grounded in the ontological fact that the collective intention pattern that is shared by all persons who have a participation intention is related as a determinable to the participation intentions in which it is necessarily multiply realized as its determinates. Most importantly, though, clause (16) (vii) indicates that the relations between all persons as described in (16) (iii) – (vi) are reasons and causes of forming collective intentions and participation intentions. What we can firmly say, therefore, is that the epistemic relations holding between each person and the others – what each person believes and understands about what the others intend and believe – is part of what causes each person's actions as part of the intended collective action. The set of epistemic relations between each of the persons and the others is therefore a causal filter that activates the causal powers of each of the persons involved. I believe that this is a case of mental downward causation. It is, for instance, the collective intention of four people to play a string quartet together that activates the causal power of the specific intention of each of the four people to play their specific instrument and part in the string quartet, and this, in turn, is likely to lead each of the players to play their specific part in the string quartet. In

45

Tuomela (2007).

80

Wolfgang Detel

order to explain this, and why the collective action came about, we have to point to the collective intentions as the causes of that action.46

References Armstrong (1978): Armstrong, D. Universals and Scientific Realism, Cambridge 1978. Batterman (2000): Batterman, R.W. Multiple Realizability and Universality, The British Journal for the Philosophy of Science, (2000) 51: 115145. Bechtel (1994): Bechtel, W. Connectionism, in: Guttenplan, S. (ed.), A Companion to the Philosophy of Mind, Oxford 1994: 200-210. Becker (2000): Becker, A. Verstehen und Bewußtsein, Paderborn 2000. Becker (2003): Becker, A. Externalism and the Causal Role of the Mind, Frankfurt 2003. Beckermann (2002): Beckermann, A. Analytische Einführung in die Philosophie des Geistes, Berlin 2002. Beckermann et al. (1992): Beckermann,A., Flohr, H. and Kim, J. Emergence or Reduction? Berlin 1992. Bennett (2004): Bennett, K. Why the Exclusion Problem Seems Intractable, and How, Just Maybe, to Tract It, www2.unil.ch/philo/Pages/epistemologie/bio_cv_esfeld/pdf/2004_pdf/ Pragmatismus04.pdf (currently available only in the Internet). Bickard and Campbell (2000): Bickard, M.H., and Campbell, D.T. Emergence, in: Andersen, P.B., Emmeche, C., Finnemann, N.O. and Christiansen, P.V. (eds.), Downward Causation. Minds, Bodies, and Matter, Aarhus 2000: S. 322-348. Block (2003): Block, N. Do Causal Powers Drain Away? Philosophy and Phenomenological Research, (2003) 67: 133-150.

46

I thank Markus Willaschek, Matthias Vogel, Ingvar Johansson, and Raimo Tuomela for making helpful remarks on earlier drafts of this paper.

Mental Causation and the Notion of Action

81

Block and Stalnaker (1999): Block, N. and Stalnaker, R. Conceptual Analysis, Dualism, and the Explanatory Gap, The Philosophical Review, (1999) 108: 1-46. Campbell (1974): Campbell, D.T. Downward Causation in Hierarchically Organised Biological Systems, in: Ayala, F.J. and Dobzhansky, T. (eds.), Studies in the Philosophy of Biology: Reduction and Related Problems, Berkeley/Los Angeles 1974: 179-186. Campbell (1990): Campbell D.T. Levels of Organization, Downward Causation, and the Selection-Theory Approach to Evolutionary Epistemology, in: E. Tobach and G. Greenberg (eds.), Scientific Methodology in the Study of Mind: Evolutionary Epistemology, Hillsdale, NJ 1990: 1 - 17. Dretske (1982): Dretske, F. Knowledge and the Flow of Information, Cambridge 1982. Dretske (1993): Dretske, F. Mental Events as Structuring Causes of Behavior, in: Heil, J. and Mele, A. (eds.), Mental Causation, Oxford 1993: 121-136. Emmeche et al (2000): Emmeche, C., Koppe, S. and Stjernfeld, F. Levels, Emergence, and three Versions of Downward Causation, in: Andersen, P.B., Emmeche, C., Finnemann, N.O. and Christiansen, P.V. (eds.),, Downward Causation. Minds, Bodies, and Matter, Aarhus 2000: 13-34. Heil (1998): Heil, J. Philosophy of Mind, London 1998: 181-188. Jablonka (2002): Jablonka, E. Information: Its Interpretation, Its Inheritance, and Its Sharing, Philosophy of Science, (2002) 69: 578-605. Jackson (1996): Jackson, F. Mental Causation, Mind, (1996) 105: 377-413. Jackson and Pettit (1990): Jackson, F. and Pettit, Ph. Causation in the Philosophy of Mind, Philosophy and Phenomenological Research, (1990) 50, 1990, 195-214. Jackson and Pettit (1990a): Jackson, F. and Pettit, Ph. Program Explanation, Analysis, (1990) 50: 107-117.

82

Wolfgang Detel

Johansson (2000): Johansson, I. Determinables as Universals, Monist, (2000) 1: 101-121. Kim (1984): Kim, J. Concepts of Supervenience, Philosophy and Phenomenological Research, (1984) 45: 153-17. Kim (1992): Kim, J. “Downward Causation”, in: Emergentism and NonReductive Physicalism, in: Beckermann, A., Flohr, H. and Kim, J. (eds.), Emergence or Reduction? Berlin 1992: 119-138. Kim (1993): Kim, J. The Non-Reductionist’s Troubles with Mental Causation, in: Heil, J. and Mele, A. (eds.), Mental Causation, Oxford 1993: 189-210. Kim (1998): Kim, J. Mind in a Physical World: An Essay on the MindBody-Problem and Mental Causation, Cambridge/Mass 1998. Maynard Smith (2000): Maynard Smith, J. The Concept of Information in Biology, Philosophy of Science, (2000) 67: 177-194. McClelland (1987): McClelland, D. Human Motivation, Cambridge 1987. Millikan (2000): Millikan, R.G. On Clear and Confused Ideas, Cambridge 2000. Moreno and Umerez (2000): Moreno, A. and Umerez, J. Downward Causation at the Core of Living Things, in: Andersen, P.B., Emmeche, C., Finnemann, N.O. and Christiansen, P.V. (eds.), Downward Causation. Minds, Bodies, and Matter, Aarhus 2000: 99-117. Papineau (1993): Papineau, D. Philosophical Naturalism, Oxford 1993. Perner (1991): Perner, J. Understanding the Representational Mind, Cambridge/Mass. 1991. Putnam (1995): Putnam, H. Worlds and Life, Cambridge/Mass. 1995. Roederer (2002): Roederer, J.G. On the Conception of Information and its Role in Nature, http://www.mdpi.net/ec/papers/fis2002/145/ Roederer1.htm. Shannon and Weaver (1972): Shannon, C., and Weaver, W. The Mathematical Theory of Information, Urbane/Chicago/London 1972.

Mental Causation and the Notion of Action

83

Sterelny (2000): Sterelny, K. The „Genetic Program” Program, Philosophy of Science, (2000) 67: 195-201. Tuomela (2007): Tuomela, R. Joint Action, in: Psarros, N. and SchulteOstermann, K. (eds.), Facets of Sociality, Frankfurt 2007: 169-207. van Gulick (1993): van Gulick, R. Who’s in Charge Here? And Who’s Doing All the Work? In: Heil, J. and Mele, A. (eds.), Mental Causation, Oxford 1993: 233-256. Christiansen (2000): Christiansen, V. Macro- and micro-levels in Physics, in: Andersen, P.B., Emmeche, C., Finnemann, N.O. and Christiansen, P.V. (eds.), Downward Causation. Minds, Bodies, and Matter, Aarhus 2000: 51-62 Yablo (1997): Yablo, St. Wide Causation, Philosophical Perspectives, (1997) 11: 251-281. Yablo (2000): Yablo, St. Textbook Kripkeanism and the Open Texture of Concepts, Pacific Philosophical Quarterly (2000) 81: 98-122. Yablo, St. (1992): Yablo, St. Mental Causation, Philosophical Review, (1992) 101: 245-280.

From Individual Mind to Forms of Human Practice Pirmin Stekeler-Weithofer, Leipzig

1. Mind and Spirit as Titles for Particular and General Competence It is not my brain that thinks. I think with my brain, just as I act with my body. The body as such does not perform an action. These are, in a sense, truisms. They are part of the grammar of our language. And yet, they have acquired the status of a philosophical theory, too: mind-body dualism. In the end, this is the view that the I (or the Self, the Soul or the Mind) is an entity whose existence is known with higher certainty – and which is, perhaps, even of some higher form of being – than my body, including my brain. The soul ‘governs’ and ‘uses’ the body. This view goes back at least as far back as to the Socrates of the probably pseudo-platonic, yet very influential dialogue Alcibiades (I). It was taken over by Christian theology and then by Cartesian philosophy, with only minor changes in content. St. Augustine and Descartes argue, accordingly, that the thinking I exists beyond any reasonable doubt. This existence is even more certain than the existence of any body. Even if I were a brain in a vat, the fact that I exist as the subject of my thoughts (and my experiences, for that matter) would still be more certain than any truth about the objective world. There is some truth in this last sentence, if we read it properly: In any act of thinking, as in any speech act, even in the act of doubting, an actor or thinker is presupposed. I can doubt that I see or touch a keyboard or that I am sitting at a desk. But I cannot doubt that it is I who doubts. Who is this actor or thinker, however? Or rather, who am I? Am I not identical with my brain or body, despite the fact that we talk as sketched above? And if not, how could we understand the difference?

N. Psarros and K. Schulte-Ostermann, Facets of Sociality, Frankfurt 2007: 85-115; ontos verlag

86

Pirmin Stekeler-Weithofer

We should not underestimate these questions, on the one hand. On the other hand, we should not presuppose that there is some kind of entity that is being referred to by expressions like “the (thinking) I” or “my Self” or “my soul” – and that all we require to know are its characteristic properties. It is precisely this presupposition that leads to the (problematic) alternative between mind-body-dualism and the different doctrines of ‘materialist’ (or ‘psycho-physical’) monism. Dualism presupposes an ontological – or rather ‘ontic’ – difference between mind and body. Monism can be characterized, roughly, by the idea that ‘in reality’ mental states or events, consciousness or personality are a ‘product’ of our brain, or rather, that talking about mental events and mental processes is only a helpful, perhaps even unavoidable, way of speaking about bodily processes. But in reality, says monism, there is no self. This (reductive) claim, however, is no less dogmatic, and no less confused about the logical status of our use of words like “I” and “mind”, “intelligence” and “consciousness” than dualism. For the program of replacing the soul or mind by the brain – or rather, speaking about the soul or mind by speaking about the brain – resembles the attempt to replace talking about a poem by talking about sequences of letters, or talking about numbers by talking about marks on papers. In any case, the mind-body-problem rests on the questionable assumption that mind is either identical with or distinct from the body or the processes it is engaged in. The standard ‘solutions’ to the mind-body in the contemporary ‘philosophy of mind’, mentalism (dualism) and naturalism (monism) share this assumption. All versions of mentalism, interactionism (i.e. proper Cartesianism) or parallelism (as proposed by Leibniz), assume that there are special entities, minds, having peculiar properties as being conscious or being intentionally related to something. Naturalists, on the other hand, like Hobbes, take pride in claiming that there is only one nature and only one realm of existing things, namely physical ones – plus, perhaps, abstract objects like pure numbers. In reality, i.e. in nature, naturalism claims, there are no such entities as spirits or souls, minds or ghosts. Furthermore, reductionists assume that it could be possible to replace ‘mental’

From Individual Mind to Forms of Human Practice

87

explanations by ‘physical’ (in some vague sense of the difficult word “could”), non-reductive monists do not.1 But both, mentalists and naturalists alike forget to ask what we do when we talk about the mind. As a result, naturalism and mentalism are equally confused. In what follows, I will try to show that the source of the confusion is this: In talking about human consciousness, intelligence, mind, or spirit we do not talk about any ‘real entity’ or ‘object’ with certain properties which is doing something at all. Rather, we use a nominal form of language in order to express features of a person as taking part in a common practice and being competent to do so. More precisely, we use a figurative form of speech when we speak, with Plato, about our soul or with Kant about mankind in us. We use it in order to reflect on forms that are related to different areas of competence. In order to see this, we should, in a first step, distinguish two kinds of presuppositions: ontic presuppositions and pragmatic presuppositions. To assume that there is an entity that can be called “the I” or “consciousness” or “mind” or “soul” is to make an ontic presupposition. To do so is to leave totally unclear or unanalyzed what we do when we use nominalizations of this form. The analysis of pragmatic presuppositions, on the other hand, shows that among the complex faculties involved in leading a human life there is some methodological order of the form: one could not do X if one could not do Y. For example, we could not do science if we did not already know practically what it is to think or to do things consciously or intentionally. I.e. we must already know how to take part in a (well-known) practice of speaking, acting and reflecting about actions. In order to avoid misleading connotations of the expression “implicit knowledge”, I will follow Karl Bühler and Paul Lorenzen and use the very helpful neologism “empractical”2 for pointing to the mode of being of forms and norms that show up only in the context of our knowing how to do things.

1

2

It is not at all clear, however, what the word “non-reductive” means for such monists. Usually, it is supposed to stand for a use of mental terms, which does not contradict the newest biological theories of cognition. Cf. Bühler (1982) and Kamlah and Lorenzen (1967).

88

Pirmin Stekeler-Weithofer

‘Ontic’ presuppositions result from a formal feature of language: Whenever we use a name-like expression, we tend to assume that it is clear what it refers to, i.e. that the object or entity named is in some sense known. Or rather, we presuppose (as known) that it ‘exists’ and that its ‘identity’ is defined. Parmenides had already grasped this most general logical structure – and some of its problems. As a result, critical thinking always requires an analysis that turns implicit presuppositions of this sort into explicit and self-conscious understanding of how to use the corresponding name-like expressions. In order to avoid ontic presuppositions, we should therefore understand a question like “Who am I?” as a meta-level question. As such it asks: What is the general grammar or use of the word “I”? The same holds for the question “What is consciousness?” And what does it mean, then, when I sometimes distinguish myself from my (present) body or from bodily parts like my hands or eyes or brain? This turn from the object- to the meta-level is the most important insight of analytic philosophy, common to Frege, Carnap and Wittgenstein, at least insofar as Frege’s ‘linguistic’ turn can be interpreted as a first step in a transformation of the objectlevel question: “What is a number?” or “What is a set?” into the meta-level question: “What is the logical grammar of number-terms or set terms and arithmetical or set theoretical predicates?”. With respect to pragmatic presuppositions, my basic claim is this: Scientific investigation of the function of the brain and of the empirically accessible behavior of humans empractically presupposes some understanding of what rational or intelligent action is. This holds especially with respect to the standards of rationality in the sciences. The importance of this truism is regularly overlooked. Brain and behavioral scientists tend to claim, for example, that their investigations reveal not only necessary, but also sufficient conditions for consciousness or intelligence. But precisely this is not the case. It is not the case because it is always already presupposed that we know how to think and how to do things consciously or in an intelligent way. It is even absolutely unclear what it would mean to specify these sufficient conditions in the terms of the behavioral and brain sciences. As a result, behavioral and brain scientists usually claim to know much more than they actually do know, or promise more than we could sensibly expect from their investigations. They do so because they are using lax, unscien-

From Individual Mind to Forms of Human Practice

89

tific language. For example, they usually do not differentiate conceptually between human consciousness, mere awareness, and vigilance as a state of being awake, or between intentional actions as consequences of experience and rational thinking on one side, more or less appropriate reactions of a vigilant animal, on the other. I.e. it is left unclear how to distinguish consciousness, as it is specific to humans, from the sort of awareness humans share with animals. Moreover, it is absolutely crucial to distinguish different forms of necessary conditions for a ‘proper’ working of the brain from sufficient conditions for having or developing a rational competence of some sort. If, for example, certain neuro-physiological conditions are not fulfilled, the competence of consciousness or reasonable comprehension may disappear – together with awareness and proper animal reactions. But there are other types of necessary conditions. Semantic understanding, for example, requires a proper social and educational development. This already shows why the specification of brain states alone could never be the specification of sufficient conditions for any ‘higher’ mental competence. To say, as Joseph LeDoux does, that synapses underlie personality since synapses underlie everything the brain does, is, therefore, as true as to say that oxygen or water underlies personality, since oxygen or water also underlies everything the brain does.3 To say that charitable cooperation and mutual respect underlie personality would, in a sense, be much closer to the facts. Because of his ontic presuppositions, Descartes was wrong to reconstruct the human mind or soul, the subject of thought, as a ‘thinking thing’, a res cogitans. But he was right in emphasizing the following pragmatic observation: Any act of thinking – and therefore any real human action – is in some way accompanied by (what he calls) conscientia, i.e. by consciousness and self-consciousness. This (self-)consciousness is, in a sense, nothing but a certain kind of (self-)control.4 And this kind of (self-)control is

3

4

Joseph LeDoux, the author of The synaptic self: How our brains become who we are (LeDoux 2002), exemplifies like, Gerhard Roth and Wolf Singer in Germany, a type of media scientist promoting cognition theory and the brain sciences. Cf. Hennig (2006).

90

Pirmin Stekeler-Weithofer

different from mere awareness. What kind of control is it, then? Any appropriate answer to this question will have to refer to social and cultural traditions, which lie far beyond the scope of any sort of proper physiological and biological investigation. In fact, mere biological approaches to mental competence and its actualizations grossly underestimate the importance of education, communication, and social cooperation, factors, which are necessary preconditions for developing and actualizing the competence of comprehension and understanding. It is crucial to see, on the other hand, that words like “(self-)consciousness”, “reason”, “will” and “intention” refer neither to some transcendent powers of a soul nor to mere bodily powers of my brain or my neurophysiological system altogether. These words are means for talking about forms of human action, practice and life. From a logical point of view, all of these name-like expressions should be interpreted as nominalizations of evaluative words like “meaningful”, “spirited”, “intelligent”, “(self)conscious”, “rational”, “reasonable”, “willful” or “intentional”. Such ‘metonymic’ nominalizations are used as linguistic means in order to refer, by way of a kind of rough gesture, either to the whole practice the corresponding evaluative adjectives are for or to the particular competence of the individuals taking part in this practice. In order to make explicit the peculiar status of these name-like expressions, we must distinguish them from ordinary names. Ordinary names refer to objects in a limited realm of concrete things or abstract entities. In such a normal or limited domain the object-level properties are clearly defined. In order to understand the nature of such limited domains of (abstract) objects, it is best to go back to Frege’s logical analysis of the realm of numbers and arithmetical properties.5 Here, we have a whole set of names like 1,2,3,… and a set of elementary sentence forms or predicates like n