Charles Travis: Thought's Footing: Themes in Wittgenstein's Philosophical Investigations.

Whiting, Daniel

Charles Travis

Thought's Footing: Themes in Wittgenstein's Philosophical Investigations.

Toronto and New York: Oxford University

Press 2006.

Pp. 240.

Cdn$72.00/US$55.00

(cloth ISBN-13: 978-0-19-929146-5).

The central question of this book is, 'How does thought ... get footing?' (2) That is: in virtue of what is what one thinks answerable to how things are? In answering this, through a reading of Wittgenstein's Investigations, Travis discusses an impressive variety of issues--proper names, privacy, perception, logic, nonsense--and the views of Austin, Dummett, McDowell, Russell and Strawson. Threaded throughout are beautiful (and compelling) examples concerning, among other things, pigs in parlours and the etiquette of fire-doors. Travis' writing is often very difficult but, given the subtlety of the points being made, not gratuitously so.

Wittgenstein's principal target, Travis says, is a Fregean idea according to which thoughts are 'beholden ... to the way things are' just if they are 'about an object and a concept in the right structured way.' Such a view is endemic in contemporary philosophy. Truth-conditional semantics, following Davidson, is committed to each sentence, as such, possessing a certain truth-evaluable content in virtue of the referents of its constituent parts. Likewise, representationalists in philosophy of mind, such as Fodor, hold that thoughts are sentences (or sentence-like entities) that, qua representations, bear truth-conditions. Opposing this is Wittgenstein's view: 'What words name (by way of concepts and objects), and the structured way they do that, does not determine, uniquely, when they would be true' (2-3). Consider, for example, the sentence, 'The car is blue'. That this names a car and applies the concept blueness to it does not yet determine whether one would count it as expressing a thought whose truth requires or precludes its having black leather upholstery, or the engine's being blue, or red paint covering blue fibreglass, and so on (30).

For Travis' Wittgenstein, what determines which thought is expressed are the 'consequences of thinking, or saying, something', what one might reasonably expect in 'the particular circumstances'. Suppose, for example, that I am choosing which car to buy. The colour of its bodywork would reasonably be expected to matter, the colour of its carburettor not. In turn, this means that what thought is expressed 'is fixed, in part, at least, by our parochial sense for things' (3-4). That is, the specific truth-evaluable content expressed on this occasion is partly determined by what we (humans) idiosyncratically take, as a matter of fact, to be of import.

In Chapter 2 Travis proceeds to consider, in the light of Wittgenstein's remarks on family-resemblance, the application of these views to singular thought, an especially important case since it there that thought has its feet most firmly in the world. Then in Chapter 3 he suggests that, for Wittgenstein, the idea that what determines what is required for the truth of a thought is our parochial judgment extends equally to thoughts concerning thought, truth, and the like. Here a threat of idealism emerges. Our judgments as to what reasonably to expect are equally determinative of the contents of thoughts about whether a stance is answerable. That is, 'a particular parochial sensibility' provides 'the measure' of whether an attitude or utterance--concerning, say, the size of a table or how nice kumquats taste--is truth-apt (146). We now seem to lack 'assurance that our would-be answerable stances are really that' (158). Travis' response, based on a novel reading of Wittgenstein's private language argument, is that someone pressing this issue is really failing to say anything at all, since (very roughly) there are no consequences that one might reasonably expect her words to have.

There are two ways in which one might think of the role that parochial judgments play in individuating thought. First, what a representation names delimits a range of possible thoughts it might express, and what is reasonably expected settles which of those thoughts is actually expressed. Second, what a representation names provisionally delimits a range of possible thoughts it might express, and what is reasonably expected might determine that a thought is expressed lying outside that range. Travis suggests the second when he writes: 'for any condition, C, on being some given thing to speak of ... we are prepared to recognize, and can conceive of, circumstances in which C would not hold of that thing there is to speak of' (60; cf. 106-7).

Adapting an example from Putnam which Travis endorses, consider the sentence 'There is a vixen'. One might take the established usage of 'vixen' to determine that what it expresses is true if and only if there is a female fox. This, perhaps, leaves it open whether it expresses a thought that would be true of flattened road-kill or a stuffed, mounted hunting-trophy. But Travis also appears to hold that, should a creature turn out not to be a female fox (but, perhaps, a robot), it might be reasonable nonetheless to take the thought expressed to be true, that is, as lying outside the range laid down in advance.

It is hard to make sense of this picture. If a word is applied on two occasions on the basis of entirely distinct set of features, why say that the same concept is expressed on those occasions? Indeed, since the conditions of application are wholly different, one's expectations on both occasions should be wholly different. Hence, by Travis' own lights, there seems no reason to hold that on each occasion the same concept is named.

Another concern is that, for Travis' Wittgenstein, thought is indeterminate. Since it is conceivable for there to be two competing judgments as to what expectations are reasonable regarding a given utterance, it appears possible for there to be no fact of the matter as to which of two thoughts a person expresses. Travis seems to acknowledge this, stating that 'there is no longer any supposition of a unique right answer to the question in which way' a person 'meant her words' (128).

It is unclear that one can continue viewing thought as real if it is indeterminate whether any given person has a certain thought. More specifically, indeterminacy appears hard to reconcile with thought's playing a role in explaining behavior, or with the (Wittgensteinian) idea that thinkers are authoritative as to what they think.

These qualms aside, Travis' rich and subtle book is a stimulating investigation of what it takes for thought to be answerable to reality. He makes a very strong case for the idea that the truth of a thought is bound up with the consequences of thinking it, that those consequences are fixed by our judgments regarding what is reasonable, and that there is no external standpoint from which to assess those judgments. These are all recognizably Wittgensteinian themes, but Travis makes them independently attractive and eloquently demonstrates their ongoing importance.

Daniel Whiting

University of Southampton