Consider the two Duhemian theses S, according to which no single or individual theoretical hypothesis by itself has any observational consequences, and F, which states that no single theoretical hypothesis can be conclusively falsified by any observations.

Does the truth of F follow from the truth of S? I had conjectured that Duhem may have thought so, but I argued that it does not [1]. Nancy Tuana contends that my argument misses the mark, and she goes on to argue that it does [2]. It seems to me that her arguments are decisive on this point, and so I welcome the correction.

In response to the growing number of paradoxes relating to the hypothesis of universal predictability, Sellars has written:

In a recent article [1], Kenneth Sayre presents what he takes to be an alternative to Salmon and Reichenbach's screening-off model of causal relations. His statistical model, based on communication theory, supposedly “has the added advantage of distinguishing cause and effect without reference to temporal distinction” (p. 203 abstract, cf. 212–213). Unfortunately his model falls far short of this grand claim, faltering before the same sort of examples I had previously used to assail the screening-off model [2].

Alberto Cortes, in [1], attempts to show (a) that Leibniz's Principle of The Identity of Indiscernibles (L) is a principle restricted to individuals (as distinct from a class of entities Cortes takes to be non-individuals), and (b) that photons (light quanta) appear to violate L (since they don't obey Pauli's Exclusion Principle). L is stated by Leibniz as “no two substances are completely similar, or differ solo numero.” In second-order quantification theory with identity L becomes:

Having noted my intention to improve upon other statistical analyses of causal relatedness by providing theoretical means to distinguish cause and effect without reference to temporality [1], Professor Shrader argues [3] that the alternative model I offer falters “before the same sort of examples he had previously used to assail” these other analyses. Although the assailant seems both to have misunderstood the model in question and to have overestimated the effect of his counterexamples, his objections provide a welcome opportunity for clarification.

In his article “On Mechanical Recognition” ([6]) R. J. Nelson brings to bear a branch of mathematical logic called automata theory on problems of artificial intelligence. Specifically he attacks the anti-mechanist claim that “[i]nasmuch as human recognition to a very great extent relies on context and on the ability to grasp wholes with some independence of the quality of the parts, even to fill in the missing parts on the basis of expectations, it follows that computers cannot in principle be programmed to recognize or learn to recognize all patterns” ([6]), p. 24). Nelson proposes, contrary to this claim, that “gestalt recognition is not beyond digital automata” ([6], p. 24). in particular, he claims, he will establish by what he calls “Turing machine arguments” ([6], p. 25) that the following four theses are true:

1. (1) automata can recognize different pattern types in one and the same set of instances;

2. (2) automata can recognize the “same” pattern in different (even completely disjoint) sets of instances;

3. (3) automata can recognize incomplete, degraded patterns having missing or indeterminate parts;

4. (4) automata can recognize family resemblances. ([6], pp. 24–25)

Aaron Snyder and Fred Dretske present an argument for the proposition that singular causal sequences, such as S caused b, need not be related to a general regularity, that an event of type S is always followed by an event of type b. Therefore, they assert that a claim of causal relation does not require a regular observation of effect following cause or cause preceeding effect. To reinforce their assertion, they present the following case:

In Fact, Fiction, and Forecast ([1]) and numerous publications in its wake Nelson Goodman has tried to define projectibility on the basis of a notion of entrenchment. While he and his associates have been able to defend the theory ingeniously against repeated and varied attacks, there seems to be one obstacle to the applicability of that theory which has not been widely discussed and to which we can only hope that Goodman will turn his attention.

One recent problem in philosophy of science is, “Ought a methodology be construed, not merely as an instrument of appraisal, but also as a source of advice to the practising scientist(s)?” Imre Lakatos ([7], p. 174 and 178) and John Worrall ([11], p. 163), among others, have answered the question in the negative. Alan Musgrave disagrees. In a section entitled, “Appraisals and Advice,” in [9], Musgrave attempts to give us a deductive argument in support of his claim that methodologies should be construed as giving advice. After briefly explaining the plausibility of the claim, I shall present Musgrave's argument in full, and then examine each of his premises. I will try to show that his attempt to answer the above question in the affirmative raises some interesting questions, and which must be answered if his venture is to succeed. I should add, that if my objections are successful against Musgrave's attempt, it would be a Pyrrhic victory: I am neither an inductivist nor do I think that the answer to the question should be in the negative.