In this article I discuss the justification of scientific change and argue that it rests on different sorts of invariance. Against this background I consider notions of observation, meaning, and regulative standards. I sketch an account of the rationale of scientific change which preserves the merits and avoids the shortcomings of the approach of Feyerabend, Hanson, Kuhn, Toulmin, and others. Each of these writers would hold that transitions from one scientific tradition to another force radical changes in what is observed, in the meanings of the terms employed, and in the metastandards involved. They would claim that “incommensurable” replacement is what does, and should, occur during scientific “revolutions.” Such writers, however, have trouble in comparing different scientific theories. My account permits different theories to be compared in various ways. These comparisons, I argue, are possible through appeal to “first-level” and “second-level” invariance. In this connection I argue that observation, meaning, and regulative standards should be, and in fact usually are, nontrivially invariant with respect to scientific change.

The conditions are examined under which one theory is said to be replaceable by another, while preserving those features of the original theory which made it serviceable for a given purpose. Among such replacements, special attention is given to ones which qualify as so-called reductions of a theory, and some theorems are proved concerning the notion of a reduction.

If, as it is usually understood, incommensurable theories must be compatible then one need never choose between two such theories. But if theories were incompatible and incommensurable one would have to choose between them. What if they are incompatible only outside the domain of observation? The fact that Darwin's biology can clash with Kelvin's physics (each with their respective auxiliary assumptions) regarding the age of the earth shows how commensurable theories may yet be incompatible. But it also shows that they need not be alternatives—i.e. that one may not be able to simply and satisfactorily replace the other in our world view. But standard examples of scientific revolutions consist of the replacement of one theory by another in one's world view. These alternative theories must therefore be more than merely incompatible—what do they share if not content ? (I.e. they must be commensurable.)

Putnam and Finkelstein have proposed the abandonment of distributivity in the logic of quantum theory. This change results from defining the connectives, not truth-functionally, but in terms of a certain empirical ordering of propositions. Putnam has argued that the use of this ordering (“implication”) to govern proofs resolves certain paradoxes. But his resolutions are faulty; and in any case, the paradoxes may be resolved with no changes in logic. There is therefore no reason to regard the partially ordered set of propositions as a logic—i.e. as embodying a criterion for soundness of proofs. Its role in quantum theory ought to be understood in an entirely different way.

It is suggested that there is a strong connection between intentions and plans, and these plans are then taken to be programs of the sort suggested by Miller, Galanter, and Pribram in Plans and the Structure of Behavior. There is then a hierarchy of programs connected with intentions stretching from the macroscopic level of ordinary discourse to the neurological level. It is argued that as we proceed downwards we arrive at a threshold below which we can still describe the phenomena but below which we can no longer speak of intentions. The paper concludes with a discussion of the criteria for the identity of intentions at various levels.

A serious problem for covering law explanation is raised and its consequences for the Hempelian theory of explanation are discussed. The problem concerns an intensional feature of explanations, involving the manner in which theoretical law statements are related to the events explained. The basic problem arises because explanations are not of events but of events under descriptions ; moreover, in a sense, our linguistic descriptions outrun laws. One form of the problem, termed the problem of weak intensionality, is apparently solved by a simple logical move, but in fact the problem arises in a new, strong form. It is found that Hempel's model for deductive explanation (to which this discussion is confined) requires modification to handle the weak intensionality problem but then is faced with the problem of strong intensionality. In consequence, it is suggested that Hempel's important concept of explanation sketch is not as widely applicable as usually claimed, especially for explanations in the behavioral and social sciences and history. Reason is found to reject the covering law thesis that every scientific explanation must contain at least one law statement. An important feature of the discussion is that some of the main reasons given for altering the deductive model and for considering other forms of explanation are internal to the covering law theory.

In order to account for the actual function of analogue models in extending theories to new domains, we argue that it is necessary to analyze the inference involved into a complex two dimensional form. This form must go horizontally from descriptions of entities used as a model to redescriptions of entities in the new domain, and it must go vertically from an observation language to a theoretical language having a different and exclusive logical syntax. This complex inference can only be intelligible if we interpret theoretical terms in a platonic manner, a la Körner.

Concepts originating in the philosophy of science generally are used only ritualistically and in careful isolation from research practice in political science. But philosophical considerations are fundamental to political research, and critically influence its decisions. The question is whether ideas offered by philosophers of science have practical (that is to say, theoretical) significance for political researchers. This essay argues that philosophy of science has extremely relevant ideas to offer. The argument proceeds through an initial presentation of some elementary notions drawn from reconstructions of the nature of concept formation-theory construction. These are then utilized in a critique of the research of quantitative political scientists. Three rather central concerns of this, still very young, discipline are discussed: measurement problems, the use of recursive and structural systems in causal modeling, and the primary logical function of multivariate analysis in political studies. The discussion is viewed as supporting the general point that applied philosophy of science ought to be, not an adjunct, but one of the key critical contributors to political research.

In a recent paper in this journal which was part of a panel discussion of Grünbaum's philosophy of science, M. G. Evans discusses the Lorentz-Fitzgerald contraction hypothesis and related matters [2]. The purpose of this note is to clarify and correct some of the points in Evans' paper.