Enunciados nomológicos y condicionales contrafácticos

Abstract

One of the distinctive like features of law-like statements has been taken to be that warrant the inference of counterfactual conditionals. For example “Metals expand when heated warrants the inference of the statements “If x were a metal, x would expand” as well as “If this metal had been heated, it would have expanded” were the conclusions are two forms of counterfactual conditionals. Accidental generalizations would only warrant the inference of a special type of counterfactual conditional, called by Goodman “counteridenticals”, an example of which is “If this coin were one of those which actually are in my pocket, it would be silver”. The assumption that there is a distinctive type of inference associated whit law-like statements has led to the view that such an inference provides information about Natural Laws.1 However, philosophers concerned with the problems presented by counterfactual conditionals have not considered in detail what the conditions in which a conditional is counterfactual are. These conditions have been regarded as suficiently clear and uncontroversial. Here I shall argue that this is not case. When these conditions are analyzed, the thesis that whereas counterfactuals are implied only by law-like statements counteridenticals are implied by accidental generalizations turns out to be tautological. It neither provides information about the nature of law-like statements nor warrants the conclusion that the derivation of counterfactual conditionals involves a new type of inference in need of clarification.

A first restriction in the application of the terms “counterfactual” and “counteridentical” is called for. We shall apply them to those conditionals where there is a presumption that the antecedent is false.2 Such a presumption is normally conveyed by using the subjunctive mood. Of course there are uses of the subjunctive in which this presumption is not made, for example, on occasions the subjunctive mood is used to indicate what the consequences or certain premises are without making any suggestion of their truth values. Conditional of this latter type many be called “merely subjunctive” to distinguish them from those of the former type to which the term “counterfactual” may be applied. To simplify our argument, in what follows we shall call “counterfactuals” only those conditionals which are not counteridenticals, i.e., those conditionals which can only be derived from law-like statements. Now is should be clear that the distinction between counterfactuals and counteridenticals. But how is it possible to characterize the distinction between these conditionals?

A characterization of the distinction between these conditionals might be attempted assuming that there are typical logical forms of statements regarded as counterfactuals or counteridenticals. However, this assumption is attended by the difficultly that there are counterfactuals with similar logical form to counteridenticals with similar logical form to counteridenticals ; in what follows we shall consider examples of both.

If I have three ravens in a box which cannot be seen by another person and I want to test his zoological knowledge, I might ask what would be the colour of a certain animal if it were one of the ravens in this box . A specific answer to this question, particularly when it is derived from a law-like statement and not from an acquaintance whit the animals in the box, may be regarded as stating a counterfactual conditional in spite of the fact that it has similar logical form to some counteridenticals and only a restricted classis mentioned. On the other hand Goodman has noted that “descriptions of particular facts can be cast in a form having any desired degree of syntactical universality. “Book B is small” becomes “Everything that is Q is small” if Q stands for some predicate that applies uniquely to B”.3 Thus is possible to formulate a counteridentical in purely general terms similar to those of counterfactuals. At this stage it seems reasonable to conclude that the logical form of a statement, by itself, is neither a necessary nor a sufficient condition for its being a counterfactual or a counteridentical. An alternative proposal might be that only the logical form of standard . Examples of counterfactuals will be shown to be different from that of counterfactuals will be shown to be different from that of counteridenticals. The counterexamples because sp far advanced would not be normally regarded as standard examples because for special reasons either there are pieces of knowledge withheld from the hearer and which are required for the confirmation of these statements, or they suggest information which in fact does not hold. For example, in the case of our example of a counterfactual with similar logical form a counteridentical, the linguistic form used (“What would be the colour of this animal if it were one of the ravens in this box?”) suggests that there is a law-like connexion between the colour of a raven, and its being in a box. The information that all ravens are black is naturally withheld from the hearer.

But if we want to distinguish between standard and non-standard examples of these conditionals all information provided and no information relevant to that purpose should be withheld from the hearer when it is available to the speaker. Thus, even if we were to succeed in proving logical forms characteristic of would be necessary, to distinguish between different types of evidence for counteridenticals the evidence for counterfactuals is not complete in the above sense. Counterfactuals formulated in restricted terms or counteridenticals expressed in purely general terms or counteridenticals expressed in purely general terms are odd, for they do not indicate that type of evidence can be provided to support the corresponding statements.

However, our observation concerning different types of evidence reintroduces an analogous distinction to that existing between law-like statements and accidental generalizations because, as pointed out by Goodman “A general statement in law-like if and only if it is acceptable prior to the determination 1f all its instances”.4 The law-like character of a statement and the fact that predictions can be derived from it is in many cases marked off by using the subjunctive mood: “If this were a piece of metal, it would expand”.
Thus, we may conclude that when we regard a statement as counterfactual we are indicating, among other things, that the statement is a law-like one . We can arrive at a similar conclusion if we reflect upon the fact that the since the distinction between counterfactuals and counteridenticals does not merely depend on the logical form of the corresponding statements, it will have to depend on what is stated by them. The sort of connexion holding between these antecedent and consequent is the most plausible candidate for distinguish between these conditionals.

Thus, as a result of our observations we may conclude that when we regard a statement as a counterfactual we are indicating, among other things, that it can be inferred from a law-like statement. Thus, to say that only law-like statement. Thus, to say that only law-like statements which suggest the falsehood of their antecedents and either are derived from law-like statements or can be characterized in terms of them. Counteridenticals would be those statements which suggest the falsehood of the antecedent without being deductible from law-like statements. Even if the existence of a distinctive pattern of inference associated whit law-like statements has not been refuted, at least it may be granted that the introduction of the concepts of counterfactual conditional and counterfactual inference is not likely to shed too much light on the nature of law-like statements.

Notas a pie de página

1Goodman, N., Fact, Fiction & Forecast, London 1954; pp. 23-25.
2 I have examined other uses of the subjunctive mood not considered here in “On subjunctive Conditionals”, Journal of Philosophy, vol. LXIV, No.10, May 25, 1967.
3 Goodman, op. cit., p. 26.
4 Goodman, op. cit., pp. 27-28.