In formal theories of measurement meaningfulness is usually formulated in terms of numerical statements that are invariant under admissible transformations of the numerical representation. This is equivalent to qualitative relations that are invariant under automorphisms of the measurement structure. This concept of meaningfulness, appropriately generalized, is studied in spaces constructed from a number of conjoint and extensive structures some of which are suitably interrelated by distribution laws. Such spaces model the dimensional structures of classical physics. It is shown that this qualitative concept corresponds exactly with the numerical concept of dimensionally invariant laws of physics.

The Conventionality of Simultaneity espoused by Reichenbach, Grünbaum, Edwards, and Winnie is herein extended to mechanics and electrodynamics. The extension is seen to be a special case of a generally covariant formulation of physics, and therefore consistent with Special Relativity as the geometry of flat space-time. Many of the quantities of classical physics, such as mass, charge density, and force, are found to be synchronization dependent in this formulation and, therefore, in Reichenbach's terminology, “metrogenic.” The relationship of these quantities to 4-vectors and their physical significance is discussed.

The purported failures of ergodic theory (seen in its often proved ineptitude to ground a mechanical explanation of thermodynamics) are shown to arise from misconception of the functions served by scientific explanation. In fact, the predictive failures of ergodic theory are precisely its points of greatest physical utility, where genuinely new knowledge about actual physical systems can be obtained, once the links between explanation and reconstructive estimation are recognized.

This paper consists of two parts. In Part I, an attempt to get around certain well-known criticisms of the trace theory of memory is discussed. Part II consists of an account of the so-called “logical” notion of a memory trace.

Trace theories are sometimes thought to be empirical hypotheses about the functioning of memory. That this is not the case, that trace theories are in fact philosophical theories, is shown, I believe, in the arguments which follow. If this is so, one may well wonder about psychologists' insistence that any empirical theory of memory must involve the postulation of traces (or trace-like entities: engrams, schemata, etc.).

Taking reduction in the traditional deductive sense, the programmatic claim that most of genetics can be reduced by molecular genetics is defended as feasible and significant. Arguments by Ruse and Hull that either the relationship is replacement or at best a weaker form of reduction are shown to rest on a mixture of historical and logical confusions about the nature of the theories involved.

These days ‘evolution’ is usually defined as any change in the relative frequencies of genes in a population over time. This definition and some obvious alternatives are examined and rejected. The criticism of these definitions points out the need for a more holistic analysis of genotypes. I attempt such analysis by introducing measures of similarity of whole genotypes and then by grouping genotypes into similarity classes. Three sorts of measures of similarity are examined: a measure of structural similarity, a measure of functional similarity and one of relational or historical similarity. The functional approach is shown to be superior and a definition of ‘evolution’ is suggested.

The distinction between discovery and justification is ambiguous. This obscures the debate over a logic of discovery. For the debate presupposes the distinction. Real discoveries are well established. What is well established is justified. The proper distinctions are three: initial thinking, plausibility, and acceptability. Logic is not essential to initial thinking. We do not need good supporting reasons to initially think of an hypothesis. Initial thoughts need be neither plausible nor acceptable. Logic is essential, as Hanson noted, to both plausibility and acceptability. An hypothesis needs good supporting reasons to be either plausible or acceptable. Such reasons need not be relative to the particular scientific theory undergoing test at the time. There is no fundamental difference between reasons relevant to plausibility and acceptability. The difference is one of degree. Acceptability requires more than plausibility.

The relation that holds between two singular sentences just if one of them can figure in an acceptable deductive nomological (DN) explanation of the other is not symmetric. As Sylvain Bromberger ([1]) pointed out some years ago, a sentence giving the height of a flagpole can figure in a DN explanation of a sentence specifying the length of the flagpole's shadow, but interchanging the explanans and the explanandum does not result in an explanation even though it may result in a valid argument. Recently Evan Jobe ([2]) has offered an ingenious account of such asymmetries, one which if successful would permit their characterization entirely within the framework of the DN model. The purpose of this note is to show that despite its attractiveness Jobe's idea will not work.

In a recent note ([3]) Roger Carlson presented a rather negative appraisal of my treatment of the logic of Fisherian significance testing in [10]. The main issue between us involves Carlson's thesis that, within the limits set by Fisher, standard significance tests are valuable tools of data analysis as they stand, i.e., without modification of the structure of the reasoning they employ. Call this the adequacy thesis. In my paper I argued that (i) the pattern of reasoning employed by tests of significance needs to be justified in spite of its unquestioned acceptance by most researchers, (ii) The best justification offered to date—by R. A. Fisher—is seriously defective. Therefore (iii) the adequacy thesis is not justified. I proposed some alterations of the pattern of reasoning employed by tests of significance and argued that they guarantee that tests are adequate epistemological tools.

In pursuing philosophical enquiries in the foundations of physics, I have found myself operating with a concept of explanation which has a quite different logical syntax from the concept which Hempel has explicated. I suspect that the concept of explanation I operate with is not especially idiosyncratic, is not fully captured in the existing philosophical literature (though I am open to correction here), and that it may be the more appropriate concept if we are to consider explanation as the goal of physics. I therefore propose to give here a short account of the structure of this alternative concept of explanation.

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