Dummett's model for a theory of language makes an essential use of the notion of sense, and thereby implies some notion of an analytic sentence. Hence, it is exposed to Quine's general criticism of those cognate notions and must therefore be confronted to it; Dummett fulfils that task by showing the need for some notion of sense and, at the same time, by defining a suitable notion of analytic sentence - which quite naturally goes with. Here a diagnosis is proposed of what underlies Quine's criticism of sense and prompts his holism. The point is to make clear what, in his approach to the theory of language, may possibly determinate his final outcomes: namely, the adoption of the stance of an observer in describing the activity of speaking a language. Moreover, two different notions of 'determinate content' are at work in each of the two philosophers, which are equally tied to their approaches to the theory of language. Quine could possibly only admit a sense which would be a piece of forniture of the world; since it is not, he straightforwardly rejects it. However, if some notion of sense were to be admitted, it should be one of something objectively determined - both as a part of the world and as in relation to some observer. Dummett, on the other hand, can and does allow sense to play a role just because he does not need, to that effect, to reify it. The linguistic content is determinate fully independently from any objective - wordly and observational - determinacy. Quine could simply not conceive of anything like that: for him, only would be respectable a (notion of) sense which would warrant its own reification. Which actually amounts to say that a good meaning is a dead meaning. Finally a criticism of Quine's holistic outcomes is sketched which is due partly to Simon Blackburn and partly to Dummett himself, who in fact elaborates his anti-holistic argument together with his own notion of analiticity. That criticism, however, now turns out to be possibly an argument against the very approach of Quine's. The issue whether or not arguments against Quine's model of a theory of language are also, straightforwardly, arguments against his approach depends on whether or not one can safely shapen a different model, immune to that criticism, within the same approach to the theory of language. This, however, remains an open issue.
