The Modern Synthesis, according to Gould (1980, p. 119), “has broken down on both its fundamental claims: extrapolationism (gradual allelic substitution for all evolutionary change) and nearly exclusive reliance on selection leading to adaptation.” In another place (1982b, p. 380), he writes that “the essence of Darwinism lies in the claim that natural selection is a creative force, and in the reductionist assertion that selection upon individual organisms is the locus of evolutionary change.” There is a trivial question of terminology here, and perhaps a less trivial question of historiography. Does Darwinism have an essence, and was the Modern Synthesis as monolithic as all that? Sewall Wright used to be routinely cited as one of the main synthesizers. Does it now turn out that his models of interdemic selection retroactively cast him into the outer darkness (on which see Orzack 1981 and Gould 1981)?

As a biologist who has been looking to philosophers of biology for fifteen years for help with epistemological problems, it was a surprise to me to be asked by Marjorie Grene and Richard Burian to prescribe some philosophical problems which have arisen in molecular biology developments such as those described by Alan Maxam. I thoroughly enjoyed the NEH Summer Institute on the “Philosophy of Biology” which they organized at Cornell and feel that they have raised many challenges to my own thinking in my field of molecular evolution and the origins of life. I believe that an analysis of Professor Ayala's, Kauffman's, and Maxam's papers is relevant to developments in molecular evolutionary theory. First, just as the punctuated equilibria versus phylogenetic gradualism debate in macroevolutionary theory has been seen by some as a challenge to Neo-Darwinism, notably not Professor Ayala, debates about neutralism versus selectionism has been viewed by others as a similar challenge.

There are many ways to characterize what we are talking about when we speak of modernity and post-modernity. But one description—as it pertains to philosophy—might go something like this. The “core problem” for philosophy in the modern world has been to resolve what Michael Dummett has called the “scandal” of philosophy:—“the scandal caused by philosophy's lack of a systematic methodology.”

Professor Bernstein has shown how, coming from three distinct orientations in twentieth century philosophy, Gadamer, Habermas and Rorty have arrived at strikingly similar conclusions. The upshot of their thought, as Professor Bernstein points out, is a powerful critique of traditional foundationalism. There is no uninterpreted “given”, no theory-neutral set of “facts”, no timeless structure of reason, no disinterested vantage point for research, outside of the interpretations and interests of the community of interpreters that are embodied in our current social practices. The collapse of the bad old dream of firm foundations and of ultimate convergence on the truth is taken as a fait accompli by Bernstein. The question that confronts us now is what to make of this situation. On the face of it there appear to be two types of stance open to us.

With his usual deftness and sensitivity to the spirit of our times, Richard Bernstein has brought us to the center of current philosophical debates concerning modernity and postmodernity. He has sought to identify the common ground shared by three of the principals in these debates, in contradistinction to the usual stress placed on the differences separating them. He finds this common ground in the vision of a “non-foundational” or “anti-foundational pragmatic humanism” which is said to inspire all three, and next to which the usually perceived differences — which he does not’ deny — fade to differences of emphasis.

One job of a commentator is to set a discussion in motion, and the easiest way to do that here is obviously to recall the familiar differences, to argue that they are basic, and to suggest that next to them the alleged common ground shades to philosophical insignificance.

Albert Einstein distinguished sharply between the context of discovery and the context of invention. For example, in a letter of 6 January 1948 to Michele Besso, Einstein wrote:

By “invention” Einstein meant the mind's ability to leap across what he took to be the essential abyss between perceptions and data on the one side in order to create concepts and axioms on the other. Although sometimes Einstein interchanged the terms discovery and invention, he deemed invention to be the route of creative scientific thinking.

Einstein himself had a few words to say about the writing of the history of science. From the first of his famous conversations with Einstein, Robert Shankland relates: “Then he gave me his ideas on historical writing of science. ‘Nearly all historians of science are philologists and do not comprehend what physicists were aiming at, how they thought and wrestled with their problems… .’ The struggle with their:problems, their trying everything to find a solution which came at last often by very indirect means, is the correct picture.” (Shankland 1963, p. 50).

Why should we take these words seriously? After all, Einstein was not an historian, and, unless we are to pose him as an authority on every topic, these comments represent but a personal opinion. This, however, is precisely why they should be taken seriously, if we are interested in the history of Einstein's own accomplishments.

In the design and analysis of an experiment there are several places where an element of randomization can be used: the design can be selected at random, the result can have a random element adjoined to it, or the random element already present can be used in the analysis. The first technique is much used by statisticians; for example, in making a survey of a population, Basu (1980) calls it prerandomization.

R.A. Fisher began defending the use of randomized assignments of treatments to plots in agricultural field experiments in 1926 (Fisher 1926). His claims on behalf of randomization were challenged by many other statisticians including his friend and colleague W.S. Gosset, the celebrated “Student” (Gosset 1937). After Gosset's death, Fisher's insistence upon the critical importance of randomization to good experimental design became a cornerstone of methodological orthodoxy, a status it has continued to enjoy until recent years. Even so, an undercurrent of resistance to some of the more excessive claims on behalf of randomization has continued to influence thoughtful statisticians and methodologists.

Fisher's rationale for. randomization was grounded on the need to control or eliminate systematic error or bias. For example, the average yield of wheat on plots treated with one kind of fertilizer may be compared with the average yield of wheat of the same variety on plots treated with another kind of fertilizer.

I have organized my remarks about randomizing under four headings: computation, communication, causal Inference, and complexity. It Is hard to think of a wore controversial subject than that of randomization. My remarks are simpler and more extreme than they ought to be. I have put them in a rather bald and definite way in order to draw the lines more sharply and to make my message as clear as possible. I do not doubt that under extended debate it would be necessary to qualify some of the things I have to say, but I would insist on the point that I would be offering qualifications, not retractions.

It is often said by pure Bayesians that once the likelihood function is available knowledge of any randomization scheme used is superfluous Information. It seems to me that this argument misses an Important point which I want to illustrate by a simple artificial example.

A recurrent theme in the “new philosophy of science” is the importance of temporally extended conceptual entities termed variously disciplinary matrixes (Kuhn 1970), research programs (Lakatos 1970), scientific disciplines (Toulmin 1972), theories (McMullin 1976), and research traditions (Laudan 1977). Each of these macro-conceptual entities contains a rich heterogeneity of constituent elements. For example, Kuhn's (1970, pp. 183-187) disciplinary matrixes (“paradigm” in his global sense) include symbolic generalizations, metaphysical views, models, values, and exemplars as concrete problem solutions (“paradigm” in the narrow sense). All of the conceptual entities listed above are “historical entities” or “continuants”, the sorts of things that can change through time. Toulmin (1972) and Laudan (1977) permit total changeover in elements just so long as the transformation is gradual and the system remains cohesive in the process. Others, such as Lakatos (1970), insist on the retention of a “hard core” of some kind.

Taxonomy is not destiny. In the context of scientific theories, it is explanation rather than taxonomy which should play the leading role. (The domain of any theory consists not of mere items—as in philately or genealogy—but of phenomena.) Unless biological taxonomy is guided by underlying theory—presumably, evolutionary theory, genetics, and molecular biology—systematics risks degenerating into mere genealogy.

David Hull believes that certain taxa employed in history of science should be conceived on the model of biological species. He first discusses the nature of species, indicating (in 1.) what they are not— they are not “natural kinds“—and then (in 2.) what they are—i.e., they are “historical entities”, “chunks of the genealogical nexus” which are individuated in relation to some arbitrarily selected and quite often atypical “type specimen”. The connection between the two points, the negative and the positive, is that conspecificity is determined by actual spatiotemporal contacts of the sort required by descent, rather than by any putative similarities among conspecific organisms.

The thesis of this paper is that perception is intelligent in that it is based on the same kinds of operations that characterize thought. However, the dependence of perception on sensory information makes for certain differences between it and “higher” cognitive functions such as imagination and thinking. Clearly the phenomenal experience of a percept is different from that of thought. Still the processes underlying these different mental end products may be quite similar. Although we know little about the nature of thought, a demonstration that perception results from cognitve operations does constitute an explanation. Others seek to explain perception in very different ways.

To turn the problem around so to speak, it is entirely possible that thinking evolved from perception. Perception might be the evolutionary link between low level sensory processes that mediated simple detection of environmental changes in phylogenetically primitive organisms and high level cognitve processes.

Here is one way to portray the history of research on the visual system. It consists of a rivalry between those who discern the benchmarks of intelligence in visual perception, and those who seek to show how the appearance of intelligence can be stripped away to reveal the reality of essential stupidity. In the main the rivalry has been exciting and productive, as the two egg each other on to ever more extreme exertions; that is to exhibit demonstrations of ever more cunning ways in which stupid elements can be wired and assembled to yield smart looking results, and by reply, to exhibit demonstrations of ever more subtle and striking performances of the visual system which betoken intelligence and defy reduction by existing reductive hypotheses. The demonstrations of intelligence have typically come from behavioral experiments concerning, for example, illusory contours, and constancy in such things as size.perception and orientation perception.

The controversy over inference in perception turns on the nature of the processes that intervene between the stimulus.and the perceptual experience or percept. For those who find it useful to conceptualize such matters in boxological shorthand, the dispute can be viewed as focusing on the contents of the box in Figure 1. (Figures will be found at the end of the paper.)

Gibsonians have stressed the informational richness of the stimulus and have argued that in virtue of that informational richness no complex inferential processes must be postulated in the box. Rather, the perceptual mechanism can be thought of as “picking up” or “resonating to” the information in the stimulus. But, of course, the suggestion that the perceptual system can pick up information “directly” from the stimulus does not really speak to the issue of what is in the box.

It was with some trepidation that I agreed to speak today, because of a strong doubt that I could say anything substantial not already to be found in the literature of the subject. I cannot say that this trepidation has been subsequently relieved: all I can claim to offer in this paper is a review of certain basic characteristics or themes in the quantum-mechanical situation (which by now should, I think, be thoroughly understood by everyone engaged with the matter), supplemented by some rather general reflections on our philosophical predicament. In aid of these more general reflections, I shall indulge a proclivity for calling on historical matters—some fairly recent, some older, some ancient—which I hope may serve to place current issues in a useful perspective; and I ask your forgiveness for allowing myself to quote certain previous, but hitherto unpublished, remarks of my own.

Allow me to begin by thanking Professor Nickles for his organizing the session on “Technology and Scientific change” in order, as he wrote me in his kind letter of invitation to participate in this session, to explore ways in which technological developments have stimulated or obstructed scientific change. I take it there is no established list of ways in which technological progress stimulates scientific progress, yet some ways are all too obvious. The simplest, perhaps, are the cases of the technological progress which facilitate the life of the scientist, the technological progress which makes more free time available, some of which is used for scientific education and research, the innumerable technological advances which raise the standard of living in myriads of ways, and which help science grow in all sorts of manners. I suppose these are the indirect ways.

As was remarked two years ago at a symposium of the PSA on locality and hidden variables (Shimony 1981), the Bell theorems together with experiments supporting the quantum correlations (violating the inequalities of the theorems) provide a rather remarkable opportunity for philosophy. Shimony summed this up in the nice phrase, “experimental metaphysics”. For, as he pointed out, arguments based on Bell-type theorems, concluding that certain experiments should exhibit correlations obeying a certain inequality, generally employ premises of a metaphysical character, e.g., that the systems in question are deterministic (in the experimental observables). If, in fact, the experimental predictions are violated, and if one has reason to believe any other premises involved, then one must give up the metaphysical hypothesis.