Dispositions and Causal Powers

Bruno Gnassounou, Max Kistler

• Publisher: Routledge
• Year: 2016
• Print ISBN:
  9781138275744
• eBook ISBN:
  9781317149491
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

Dispositions are everywhere. We say that a wall is hard, that water quenches thirst and is transparent, that dogs can swim and oak trees can let their leaves fall, and that acid has the power to corrode metals. All these statements express attributions of dispositions, be they physical, physiological or psychological, yet there is much philosophical debate about how far, if at all, dispositional predicates can have complete meaning or figure in causal explanations. This collection of essays, by leading international researchers, examine the case for realism with respect to dispositions and causal powers in both metaphysics and science. Among the issues debated in this book is whether dispositions can be analyzed in terms of conditionals, whether all dispositions have a so-called categorical basis and, if they do, what is the relation between the disposition and its basis.

PART 1
The Metaphysics of Dispositions and Causal Powers

Chapter 1

Dispositions and Counterfactuals. From Carnap to Goodman’s Children and Grandchildren

François Schmitz

In good Fregean orthodoxy, a concept or a relation is scientifically acceptable only if it has sharp boundaries, in other words, if it can be decided, for any object name filling the empty place of the conceptual word, whether the resulting proposition is true or false. It is well known that this requirement of sharp boundaries is the consequence of a twofold thesis: first, a concept does not subsist outside a propositional context, and second, a ‘scientific’ proposition is true or false, any third possibility being precluded.

In good positivist-empiricist orthodoxy, a descriptive concept is scientifically acceptable only if it can be reduced to a ‘phenomenal basis’ or to ‘observational predicates’, for it must ultimately be possible to establish the truth or falsity of every proposition on the basis of the data of ‘experience’ (sense-data or observable characteristics).

In good Russellian orthodoxy, ‘inferred’ entities are definitely suspect and it is desirable to be able to replace them by ‘logical fictions’, on the model of the construction of mathematical entities on the basis of a small number of ‘logical’ entities provided by Principia Mathematica, which means to construct them as classes.

Carnap was one of the few who admired Frege; when he read Our Knowledge of the External World in 1921 he was enthusiastic, and he did not feel any antipathy for Mach’s ‘phenomenalist’ theses. The result was The Logical Construction of the World,¹ written in the years around 1925 and published in 1928.

For what will concern us here, Frege’s requirement of ‘sharp boundaries’ entails that the construction/definition of objects or quasi-objects makes use only of explicit definitions: definitions in the strict sense of the term (‘4=_df 3+1’, so that 4 belongs to the same ‘object sphere’ as 3 and 1), or definitions in use (‘n is a whole number that is the square of a whole number =_df n is a whole number and there exists m such that m is a whole number and n = m × m’, ‘whole number that is the square of a whole number’ designates a quasi-object relatively to integers). Such definitions being considered as bi-conditionals between definiens and definiendum, it is always possible to eliminate those ‘constructed’ objects and to put in place of them the objects on the ground of which they have been constructed. So, every statement including expressions for ‘constructed’ concepts can be translated into a statement not including them.

On the other hand, so called ‘implicit’ definitions (introducing concepts through axiomatic systems) are definitively ruled out because it is not possible to decide, for each object, whether, or not, it falls under so ‘defined’ concepts, as Frege had so many times emphasized in his articles on Hilbert’s Foundations of Geometry and as Carnap himself had settled anew in his 1927 paper, ‘Eigentliche und Uneigentliche Begriffe’.²

Applied to the descriptive concepts which appear in empirical sciences, Frege’s requirement has effects which seem puzzling: a scientific sentence in which appear concepts constructed in the constitutional system of Carnap sketched in the Aufbau, is, after substitution of the definientia to the definienda, reduced to a statement (very complicated) which expresses only a state of affairs concerning the basic relation; and this state of affairs is made of a finite number of basic elements (so-called ‘elementary experiences’) connected by the relation of ‘recollection of similarity’.³ In this way, we are sure that it is always possible to decide whether or not the state of affairs in question is the case.

More accurately, this means that a scientific statement speaks only of a finite number of past experiences even if it seems to state the future occurrence of an event or a general law. As E. Kaila said in his critical study of the Aufbau: ‘… as regards its logical meaning – but not also as regards its “content” – no concept which does not directly refer to what is given can denote anything else than classes of basic elements, classes of classes or classes of relations of them, etc. […] Just as little as pure mathematics, if natural numbers constitute its basis, can contain anything other than statements about classes, classes of classes, classes of relations,… of natural numbers, just so little can empirical science contain anything other than statements about classes, classes of classes,… of “my” past elementary experiences’.⁴

Carnap was not saying anything else in ‘Testability and Meaning’⁵ when he wrote: ‘The laws of physics as well as all predictions were interpreted [in the Aufbau] as record of present and (remembered) past experiences, namely those experiences from which the law and the prediction is usually said to be inferred by induction.’

Yet, one year before the publication of the Aufbau, the physicist P.W. Bridgman, following the well known formula of Einstein’s (‘this notion [sc. the simultaneity of two events] exists, for the physicist, only if he has found the possibility to verify in the concrete case whether or not it is accurate’), exposed his operationist conception of physical concepts: ‘In general, we mean by any concept nothing more than a set of operations; the concept is synonymous with the corresponding set of operations’.⁶ The immediate consequence of this thesis is that to each set of operations corresponds a different concept, even if, ‘intuitively’, it would seem that we deal with the same concept (for example, the concept of ‘length’ explodes into so many different concepts as there are different procedures of measuring length): ‘If we have more than one set of operations, we have more than one concept, and strictly there should be a separate name to correspond to each different set of operations’.⁷

Another consequence of this point of view was that a physical concept has a determinate meaning only in the domain or region in which the operations can be performed; where it is not possible to perform them, the corresponding concept lacks any meaning: ‘… the concepts can be defined only in the range of actual experiment, and are undefined and meaningless in regions as yet untouched by experiment’.⁸ In Frege’s words, all this amounts to say that a physical concept has no sharp boundaries: it is possible to decide for an object whether, or not, it falls under the concept only if the object in question belongs to the domain where the corresponding operations can be performed; for objects outside this domain, the question whether or not it falls under the concept, is merely meaningless. In other words, Bridgman’s theory results in giving to the definition of a concept the form of a conditional definition, whose antecedent includes experimental data.⁹

The Vienna Circle took notice of Bridgman’s book in 1929 and quickly recognized in its ‘operationism’ a point of view in keeping with its own empiricism. After having admitted, at the beginning of the 1930s, that the translation requirement championed in the Aufbau was too strong, Carnap brought up the weaker requirement of reducibility (and confirmability) which was a sort of formalization of a weak form of Bridgman’s operationism.

This ‘liberalisation of empiricism’, which will take new forms after the Second World War, recognized that it could be interesting to introduce new concepts on the basis of ascertained regularities, that is to say, if one is ‘regularist’, on the basis of physical laws: for example (a simple one!) if I remark that some bodies, every time they are subject to a pressure and change their shape, recover their initial shape when the pressure is released (while others do not), I could consider advisable to subsume them under the concept of ‘elastic body’. Afterwards, on the demand of the Royal Society, I investigate the behaviour of two elastic bodies when they collide, and, after some experiments, I establish the ‘law’ according to which, when ‘two bodies running with equal and opposite velocities collide, they part company with the same velocities as before’. From this, it follows that if an elastic body collides with another elastic one at rest, it transmits to it all its velocity and stays at rest after the collision, etc. Then, I come perhaps to the idea that an elastic body is such that every change of its shape corresponds to a change in the system of pressure to which it is subjected, and reciprocally, etc.

A property like that of ‘elasticity’ appears only when I remark that in some circumstances, or when subjected to some tests, some objects (here, bodies) behave regularly in a determinate way. This experimentally ascertained regularity suggests introducing a new concept as far as it seems reasonable to expect the objects falling under this concept to present other common characteristics in other circumstances (here, to obey particular laws of shock); it is clearly a bet on the future which could prove wrong, only future experiments being able to decide on this. As one knows, the so-called disposition-concepts, those which in everyday language end in ‘able’ (‘adaptable’), ‘ible’ (‘visible’) or ‘uble’ (‘soluble’) furnishes the prototype of such concepts. The aim of Carnap in ‘Testability and Meaning’ was to legitimate the introduction of such concepts, once it is admitted that all ‘theoretical’ concepts can be handled as disposition concepts.

What does ‘to legitimate’ mean here? This means:

1. that it must be possible to continue distinguishing concepts of this kind – which, while they are not eliminable to the benefit of observational concepts or terms alone, are not without support in ‘experience’ – from ‘metaphysical’ concepts (such as the ‘entelechy’ of Driesch, for example);

2. that the distinction between analytic statements and synthetic ones can be made sharply.

The historical situation of this question is rather curious. Carnap had accepted getting out of the frame of strict Fregean orthodoxy and had endeavoured to preserve, as much as possible, the empiricist orthodoxy. Later on, he eventually recognized that the solution, in terms of ‘disposition’, of the epistemological problem posed by the ‘theoretical terms’ was not fully satisfactory; this led him, by steps, to re-elaborate the solution sketched by Ramsey in 1929 and to appeal to the so-called ‘Ramsey sentences’.

On the other hand, the introduction of disposition predicates by means of the ‘reduction sentences’ put forward by Carnap in ‘Testability and Meaning’, gave rise, in the 1940s/50s to an abundant literature resting on the idea that one had to get off the extensional logic for the treatment of the disposition predicates because definitions of such predicates ought to be in terms of ‘counterfactual (or subjunctive) conditionals’, something like: ‘b is soluble =_df if b were put into water, b would dissolve’.

As it is well known, later on, the developments of ‘possible worlds’ semantics for modal logic led some authors in the 1960s (Stalnaker and D. Lewis are the best known) to treat again the question of a ‘logic’ of counterfactuals in this new frame. The idea was that the seemingly insuperable difficulties, which have been encountered in this question in the 1940s, could perhaps be solved by using this semantic approach. So, a new topic got in and grew in an autonomous fashion, but only loosely related with Carnap’s initial concerns.

Yet, as the logics of counterfactuals have their remote origin in the way Carnap was treating disposition predicates in ‘Testability and Meaning’, it may be amusing, or at least instructive, to come back and to consider the following question: is the handling of the counterfactuals from the point of view of possible worlds semantics, more precisely, from that of the ‘ptolemaic’ semantics of D. Lewis, able to cope with the objections raised by Chisholm and Goodman against the introduction of disposition predicates by means of reduction sentences? Carnap, in 1963, expressed his open-mindedness on this topic and declared in his answer to the criticisms of A. Pap: ‘Only in the future, when the logic of modalities has been investigated much more thoroughly, will it be possible to judge whether an extensional or a modal language has the greater overall advantage’.¹⁰ Perhaps it is the moment to pass judgement on this matter.

Vices and Virtues of Reduction Sentences

A disposition property is ascribed to an object only under the condition that the object in question responds in a determinate way to a change in the surrounding circumstances: a piece of wax, when heated, may be said to be flexible if it changes its shape when subjected to a pressure. Strictly speaking, the piece of wax cannot be said to be flexible before it has been verified that it changes its shape when subjected to a pressure. So long as it has not undergone such a test, the assertion that it has this property is not supported. With the aid of standard logic, one is led to introduce a predicate ...

Table of contents

• Cover Page
• Half Title page
• Title Page
• Copyright Page
• Contents
• Notes on Contributors
• Acknowledgements
• Introduction
• Part 1 The Metaphysics of Dispositions and Causal Powers
• Part 2 Dispositions and Causal Powers in Science
• Bibliography
• Index