IV. Reasons for Holism Thesis and the Web-of-Belief Model: Language Acquisition
Why reasons offered by TDE are not enough. While Quine’s positive account is as compelling as it is gripping, his negative account falls short of out and out refutation of old dogmas. For one, neither the Duhem Thesis nor the failure of Carnap’s reductionist program is sufficient to firmly establish the fallacy of the Dogma of Reductionism. Duhem originally stated his thesis as an empirical observation strictly about the challenges that physicists faced when confirming theories of physics (Ariew, 2011). To establish that the same challenges arise in confirming any statement whatsoever, Quine would need to cite corresponding empirical observations, but more importantly, explain why they arise. TDE attempts neither. A failure of a single reductionist program, or even several, is not sufficient evidence for the impossibility of such a program. Quine’s argument against The Dogma of Analytical/Factual duality also falls short of being conclusive. Some authors have noted three aspects of the “closed curve” argument that lack foundation. First, TDE is unclear as to the necessity of the main assumption of the “closed curve” argument, that an inexplicable concept must be illusory. Second, Grice and Strawson find Quine’s standards of clarity arbitrary (Grice, 1957). In particular, Quine’s standards for a clear explanation prohibit the explanandum and the explanan from belonging to the same circle of interrelated concepts. Third, they also point out that Quine arbitrarily assumes the impossibility of explanation of analyticity based on the failure of only the few cases that he considers.
Additional Reasons for Holism Thesis. However, the arguments of TDE lose their arbitrariness when placed in context of Quine’s beliefs about language and the steps to acquire it in our early years, which he summarizes in “The Nature of Natural Knowledge” (henceforth, NNK). According to NNK, we learn to establish links between sense data and statements when we learn our first language. Therefore, the process of establishing the truth of a statement from the sense data is determined by the process by which we acquire language, specifically, when we progress from stating observations to more general statements: “The paths of language learning, which lead from observation sentences to theoretical sentences, are the only connection there is between observation and theory” (NNK, p. 79).
Language acquisition consists of two stages: straightforward induction from observations to observation sentences and vague, inscrutable “analogical synthesis.” Observational sentences are a natural springboard into language for multiple reasons: (1) they are more connected to sense data than all other statements, (2) their truth value depends on the circumstances of the communication, observable by everybody at the time, (3) as a consequence, they can be learned by induction. Induction is made...