As Edward Becker acknowledges in the first sentence of The Themes of Quine's Philosophy: Meaning, Reference, and Knowledge, Willard Van Orman Quine ‘was the most important and influential hilosopher of the second half of the twentieth century.’ From ‘Truth by Convention’ (1936) to From a Logical Point of View (1953), Word and Object (1960), (1969) to The Pursuit of Truth (1990 at the end of his life, he has continued to make important contributions to philosophy.

The most distinctive features of Quine's philosophy include a developed and distinctive naturalistic outlook and an emphasis on the philosophical importance for philosophy of contributions from formal logic. This is most clearly manifested in his rejection of some familiar philosophical dichotomies: Quine makes no philosophic al use of the distinctions between the necessary and the possible, between the a priori and the a posteriori, and he appears to make little philosophical use of the concepts of meaning and synonymy. He was initially sympathetic to the logical positivism of Rudolf Carnap, and he must have been influenced by his teacher C.I. Lewis. But much of his work was marked by the rejection of what he would have learned from his teachers.

His engagement with the analytic synthetic distinction was fundamental to the development of Quine's work. The concept of analyticity had been intended to provide a contribution to explaining a range of concepts and phenomena by appeal to which logical positivists saw as fundamental to their work. We can ask about the apparent differences between for example, logic and mathematics, on the one hand, and empirical science or other descriptive phenomena on the others: mathematical propositions appear to be necessary truths, things that could not have been otherwise in other circumstances; we acquire mathematical knowledge by proof or calculation, while our empirical knowledge appears to be obtained through observation and inductive reasoning. Some positivists explained these features of our knowledge by arguing that mathematical propositions, for example, are analytic: they are true by virtue of their meanings; or their truth consists in the fact that we have adopted a convention to treat such propositions as necessarily true. How to understand analyticity may be unclear, but we can recognize that if such propositions are analytic, this promises an explanation of their distinctive epistemological and metaphysical characters. In his important paper ‘Two dogmas of empiricism’, Quine explored how we might understand analyticity, and just how being held to be true by a conventional decision. In a series of important papers, Quine made evident the problematic features of analyticity, and his paper finished with suggestion that we should explore the consequences of abandoning these concepts for serious philosophical purposes.

Central to Quine's presentation of his views is the defence of two important semantic theses. The first is the Indeterminacy of Meaning which holds that there is no fact of the matter about whether two sentences or propositions have the same meaning. Quine's favoured formulation of this is the Indeterminacy of translation. When we attempt to translate a hitherto unkown language on the basis of evidence of when people endorse sentences or propositions, then there will always be different, inconsistent manuals of translation available such that there is no evidence which determines which of the translations is correct. If this is correct, Quine holds, then the semantic facts that are used to ground claims about what is necessarily true are simply not available. If there are no facts about synonymy, then there are no analytic truths, so the linguistic account of necessity cannot be defended.This indeterminacy thesis undermines philosophical uses of analyticity.

The second of these theses is the ‘inscrutability of reference’. The ontology of a theory is fixed by the objects that must exist if the theory is true; In ‘On what there is’ Quine argued that we can identify the ontology of a theory be identifying what objects must exist for the theory of be true, by observing the contents of the universe of discourse over which the values of variables must range. Quine argued that we could describe a body of theory in different ways, which would involve different ontological commitments. For example, compare ‘Some rabbits are brown’ involves quantification over rabbits; and ‘Some undetached rabbit parts are brown has different quanitification, over rabbit parts. The question whether our discourse is really about rabbits or about undetached rabbit parts makes little sense. A trivial matter, but one that helps us to see way we should ot take questions of ontology seriously. By 1970, and ‘Ontological Relativity and other essays’ saw a further development in Quine's views: what ontological commitments a theory carries depends upon context. It is a great strength of Becker's book that he provides us with a subtle and an impressive exploration of just what Quine's position takes ontology to be relative to. Becker has evidently been studying Quine's texts for a long time, and he examined each development with great care.

Chapters one and two attempt to understand Quine's views and his arguments for them. The second and third Chapters of Becker's book provide an invaluable account of these views: the first of them is a detailed exploration of the arguments of ‘Truth by Convention’ and the ‘linguistic conception of necessity. He provides a impressively well informed and detailed accounts of Quine's theses and of the arguments he employed to defend them. For many of his readers, the most interesting and challenging part of the book will be chapter five in which Becker ‘is somewhat critical of some of Quine's doctrines’, although the criticisms are ‘largely constructive’. The character of Becker's attitude towards Quine's views is measured complexity: his admiration from Quine is combined with a readiness to conclude that many of Quine's views are mistaken. In section 5.7, for example, Becker argues that Quine's arguments for the indeterminacy of translation are not conclusive; ‘the analytical apparatus developed by Quine on the basis of stimulus meaning can, with judicious adjustments, be used to reconstruct synonymy in terms that meet Quine's standards of clarity.’ (280). Given the fact that Quine emphasizes the connections between synonymy and a host of other philosophically important concepts such as analyticity, Becker thus argues that a ‘Quinean’ should have the right to use a lot of concepts which Quine himself has repudiated. Most important, section 5.8 is called ‘Rehabilitating analyticity and the linguistic doctrine.’ (281). As we have already seen, most of Quine's readers interpret him as rejecting both the linguistic conception of necessity and a priori knowledge; but now we find him offering at least a qualified defence of this view. It is a natural reading of Becker that he holds that the framework for understanding language and knowledge that Quine constructed as an alternative to the analytic/synthetic distinction can actually be used to defend a version of analyticity and the linguistic theory of necessity.

Some readers have not recognized the distinction between the indeterminacy of translation and the inscrutability of reference. The latter seems relatively trivial, as Quine acknowledges when he defended it using the argument from proxy functions. The indeterminacy of translation appears to be much more threatening. If there are no determinate meanings, this appears to present a challenge to community and conversation. And, of course, it also resents a challenge to some attractive philosophical theses, for example those that involve the linguistic theory of necessity. Towards the end of his life, Quine himself drew distinctions between his different ‘indeterminacy theses’: the indeterminacy of translation and the inscrutability of reference. For example, he distinguished the epistemic statuses of the two claims: the indeterminacy of translation is an empirical hypothesis which may well turn out to be false; while the inscrutability of reference, by contrast can be proved. This is because translation is usually a shared activity, involving people with a strong commitment to finding a way to speak together and understanding each other. If different participants come up with different manuals of translation or interpretation, then this just gives us a reason to try to do better, to find a better translation manual which we can all accept. Becker quotes Quine's claim that ‘the thesis of the inscrutability of reference “admitted of conclusive and trivial proof by proxy functions, hence the model theory”.’ (256fn.) In an important passage, Becker argues that ‘Quine's appeal to proxy functions, like is other arguments for the inscrutability of reference, is unsound, and rests on the assumption that sameness of truth conditions is sufficient for synonymy.’ (260)

By the end of the book, Becker is arguing that we can adopt a broadly Quinean approach to the philosophy of language without having to reject the analytic/synthetic distinction. The indeterminacy of translation is rejected, largely on the grounds that speakers eager to communicate successfully will be able to find translations that they and agree to. And then, once the determinate meanings have been secured, Becker hopes to undermine his inscrutability of reference and a full-blooded analytic/synthetic distinction and the linguistic theory of necessity.

The value of the Becker's book is clear. It provides a detailed and analytical account of Quine's philosophy, and a detailed exploration of the arguments he employed for defending them. But the book also provides a challenge. Becker's criticisms of some of Quine's most important views are powerful and interesting, but he also provides a valuable case for Quine's general approach to philosophy.