Abstract
In the previous chapter we examined the nature of evidence for truth that is independent of the epistemic. But one variety of anti-Realism is not concerned with the relation between Realism and the epistemic at all. In the first part of this book we called this form of opposition to Realism ‘conceptual-relativistic’ anti-Realism. Whereas the epistemic-truth anti-Realists were concerned to deny that truth is independent of our means of knowing, or verifying or confirming claims about the world, the conceptual-relativistic anti-Realists were concerned to deny that truth is independent of discourse. It can be found in the writings of Nelson Goodman and in some of the writings of the later Hilary Putnam. The aim of this chapter is to describe what would constitute evidence for truth that is independent of discourse.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
Notes
I will continue to use Putnam’s expression ‘cognitively equivalent’, although it is, I think, rather doubtful whether this is the best expression. To say that two theories are cognitively equivalent suggests that the thought or idea of one theory is the same as that of the other. But it is very doubtful whether Putnam’s examples of cognitively equivalent theories do express the same thought or idea. He says, for example, that particle and field formulations of a theory can be cognitively equivalent. But the thought or idea of a particle is obviously not identical to the thought or idea of a field. From a Realist perspective, so-called cognitively equivalent theories are linguistically or conceptually different ways of expressing the same ‘underlying facts’. So, from a Realist perspective, a better name might be ’factually equivalent’. But it is obvious that this name is preferable only from a Realist perspective. (An opponent of Realism would not agree that cognitively equivalent theories were different ways of expressing the same underlying facts. I do not have a ’non-partisan’ and clearly better way of describing the theories with which we are here concerned, and so I will stick with Putnam’s ‘cognitively equivalent’.
As noted in chapter V, this should not be taken to mean that such claims cannot be true independently of the Intentional, nor even that we could not have evidence that they are. It is merely an acknowledgment that here our concerns are with empirical truth, particularly with the claims of empirical science. But the idea that there can be a kernel of a mathematical theory that is true independently of any particular formulation of the theory leads naturally to ‘Structuralism’ in the philosophy of mathematics.
I have fairly frequently encountered objections either identical or very similar to this in conversation with anti-Realists concerning science. The argument is generally advanced as showing that Realism is obviously ridiculous. I suspect that something like this argument is behind a great deal of opposition to Realism.
This is done in my ‘Realism and Equivalence’ in Erkenntnis 31 (1989), pp.109–128.
The terms ‘designative’ and ‘attributive’ are borrowed from Michael Devitt’s Designation (Columbia University Press, 1978). The way I use these terms is not quite the same as Devitt: According to Devitt, a referring refers designatively if it refers to that to which it is linked by what Devitt calls a d-chain. (A d-chain is Devitt’s own explicated version of a ’causal chain of the appropriate kind’.) A term refers attributively if it refers to that which fits some description. On my use of these terms, a theoretical term designatively refers to that which is causally responsible for the observed effects by which we recognise the referent of the term, whereas a theoretical term refers attributively if it refers to that which fits a theoretical description not containing reference to observable effects.
Unless, of course, the properties in terms of which ‘t’ is defined are themselves to be defined in terms of those observational properties that we normally take to be indicative of t. But I will assume here that this situation does not occur.
It is, for example, compatible with the ‘tA’ component being much more specific, or more detailed, or having greater empirical content than ‘te’. Even such a well-known opponent of the causal theory of reference as Michael Dummett would, it seems, accept (1). See Dummett’s Truth and Other Enigmas,especially the paper ‘The Social Character of Meaning’.
See H. Mellor ‘Natural Kinds’ in the British Journal for the Philosophy of Science (1977), pp.299–312.
One question that naturally occurs here is: Do physicists refer to hypothetical fundamental entities or properties, not actually yet found? The answer to this seems to be clearly ‘Yes’. One such hypothetical ‘entity is the quark. If quarks exist scientists have, presumably, been referring to them. yet no quarks have ever actually been observed, and there is even speculation that it may be in principle impossible to observe them. And it is perhaps doubtful that anything that could reasonably be called ’quark effects’ have ever been observed. Therefore, it would appear that the meaning of ’quark’ does not have what we are here calling a ’designative component’.
It will be argued that there are reasons for doubting that the term ‘quark’ creates a difficulty for the view of cognitive equivalence to be defended here. First, we should note that it is only essential to this view that those terms for the objects of theories admitting cognitively equivalent versions need to have designative components. But I am not aware of cognitively equivalent versions of quark theory. Second, I think a case can be made for saying that quarks do have observable effects,even though they themselves may not be observable. There appears to be no evident reason why we cannot regard the quarks as being causally responsible for the (indirectly) observable properties of leptons and baryons, much as we regard electrons as causally responsible for the directly observable chemical behaviour of atoms, and molecules as causally responsible for the directly observable behaviour of material objects. If it is granted that quarks do have detectable effects, then maybe ‘quark’ does have a designative component after all. I conclude that further work needs to be done before it can be concluded that the case of ‘quark’ creates a difficulty for the view of equivalence defended here.
Unfortunately, the statement of the Law of Universal Gravitation requires a little more ingenuity. However, we can do so if we note that each point-mass occupies the same position as the centre of the corresponding field. Obviously, the centre of a point-mass will therefore occupy the same position as the centre of the corresponding field. Let us represent ‘the centre of X’ by c(X) and ‘the centre of Y’ by c(Y). Clearly, c(X) will refer to the same position in space, whether X is taken to be a point-mass, or the corresponding field. Using m(X) and m(Y) to mean ‘the mass of X’ and ‘the mass of Y’ respectively, and D(a,b) to mean ‘the distance between a and b’,we can state the inverse square law of gravitation as follows: U.G. will represent the kernel of the law of gravity upon which all formulations of Newtonian theory will agree.
See my ‘Realism and Equivalence’ in Erkenntnis, loc. cit.
This is not to assert that the opponent of Realism cannot explain this. Goodman’s Many Worlds Thesis, Putnam’s Many Interpretations Thesis and the Kuhnian Constructivists Semantic Incommensurability Thesis are three examples of how an opponent of Realism might consistently assert that non-equivalent theories in the same domain can all be true.
This is, of course, not the only defect of the attempts to explain novel success in epistemic terms. In chapter eight it was argued that such attempts either disallow from saying that some sentences, whose truth is needed to explain novel success, are true, or else make the connection between truth and novel success conceptual and therefore not explanatory.
It is apparently this view that is the object of Putnam’s attack in his article, ‘Is the Causal Structure of the Physical Itself Something Physical?’. Putnam advances a very wide range of arguments and considerations in this paper, but, as far as I can see, there is only one that might be thought to tell against the view presented here. (Even this argument is sketchy and allusive.) The Realist says that the causal structure of the world is metaphysically real. So, presumably there will be, according to the Realist, some family of statements of the form ‘A caused B’ that are true, objectively and independently of anybody’s point of view. But now, Putnam-asks, precisely which class of statements will this be? Putnam attributes to the Realist the causal theory of reference, that is, roughly, the theory that a referring term refers to that to which it is linked by a causal chain of the appropriate kind. So, according to the Realist as interpreted by Putnam, the terms ‘A’ and ’B’, as they appear in ‘A caused B’, will refer to those things to which they are linked by a causal chain of the appropriate kind. But which entities, precisely, will they be? The crucial step in Putnam’s argument is to assert that when we say, for example, that the term M refers to mass because it is the property mass that was causally responsible for the observable effects whereby we recognised the presence of M,we are using cause in the sense of ’the cause’, rather than ’a cause’. We say that mass is the cause of the effects, not just a cause, as we might say that the cause of the bushfire was the sparks, whereas the oxygen was merely a cause. But now, Putnam plausibly says that what is to count as the cause is interest-relative. Different groups of speakers with different interests may count one thing as the cause of an event, while other speakers may class another thing as the cause. Putnam apparently draws the conclusion from this that if we say for example that the referent of ‘A’ is whatever was responsible for the effects by which we recognise A, the thing that we thereby fix as the referent of A will depend on our interests. Other speakers, with different interests, will attach different referents to A. We can now glimpse how Putnam’s argument is meant to create a difficulty for Realism. The Realist, according to Putnam, says that there is some family of claims of the form ‘A caused B’ that are true objectively, independently of anybody’s point of view. But Putnam, if I understand him correctly, has argued that this is not so. Precisely which family of statements it is that is claimed to be true will depend on the speaker’s interests. On speaker, S„ with one set of interests, may assert that ‘A caused B’, and thereby mean that, for example, event E, caused event E2, while another speaker, S2, with another set of interests, may also assert that ‘A caused B’ but thereby mean that E3 caused E4. Now, let us assume the Realist asserts that ‘A caused B’ is true objectively, and independently of anybody’s point of view. But according to Putnam, if I understand him correctly, precisely which things the Realist will thereby be asserting to be causally related will depend on the Realist’s own interests. Putnam sees this as a kind of reductio ad absurdum of the Realist’s view that the claim is true independently of anybody’s point of view.
I am by no means confident that I have correctly stated Putnam’s argument. But there is one step in the argument that I am confident Putnam does need, and which is simply false. Putnam assumes that the Realist will say that a term M refers to that which is causally responsible for certain effects, where the thing that is ‘causally responsible’ for those effects is what the speaker would take to be the cause of those effects. But a Realist need not say this. A Realist - even one who subscribes to the causal theory of reference - can allow that if a speaker can use a term M to refer to that which is responsible for certain effects, then the speaker must have approximately true beliefs about the mechanism whereby the theoretical property caused those observable effects, and that it is the approximately correct beliefs about the mechanism that pick out one property as the referent of M. (See R. Nola ‘Fixing the Reference of Theoretical Terms’.) Such an account of reference-fixing need make no mention of any notion of ’the cause’ of observable effects, and so Putnam’s argument fails.
For an account of the shape of space and time as an explainer, see G. Nerlich The Shape of Space (Cambridge University Press, 1975).
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 1997 Springer Science+Business Media Dordrecht
About this chapter
Cite this chapter
Wright, J. (1997). The Evidence for Realism — II: The Evidence for Independence from Discourse. In: Realism and Explanatory Priority. Philosophical Studies Series, vol 71. Springer, Dordrecht. https://doi.org/10.1007/978-94-017-2844-7_11
Download citation
DOI: https://doi.org/10.1007/978-94-017-2844-7_11
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-4823-3
Online ISBN: 978-94-017-2844-7
eBook Packages: Springer Book Archive