Notes

1. . Henceforward frequently shortened to “physicochemical.”
2. . Francis H. C. Crick, Of Molecules and Man (Seattle: University of Washington Press, 1966), p. 10.
3. . W. D. M. Paton, Oxford discussions, 1971–72.
4. . R. S. Cohen, “Causation in History,” in Physics, Logic and History, ed. W. Yourgrau and A. D. Breck (New York: Plenum Press, 1970), pp. 231–45.
5. . J. D. Lambert, Oxford discussions, 1971–72.
6. . J. A. Russell, Oxford discussions, 1971–72.
7. . D. W. Millard, Oxford discussions, 1971–72.
8. . H. M. Robinson, “The Mind‐Body Problem in Contemporary Philosophy,” in this issue.
9. . Morton Beckner, “Reduction, Hierarchies and Organicism,” in Studies in the Philosophy of Biology: Reduction and Related Problems, ed. Francisco J. Ayah and Theodosius Dobzhansky (London: Macmillan Press, 1974), pp. 163–76.
10. . See Francisco J. Ayald, introduction to Ayala and Dobzhansky (n. 9 above), pp. vii–xvi.
11. . Theodosius Dobzhansky, “On Cartesian and Darwinian Aspects of Biology” in Philosophy, Science and Method, ed. S. Morgenbesser, P. Suppes, and M. White (New York: Macmillan Co., 1969), pp. 165–78.
12. . Ayala (n. 10 above), p. viii.
13. . Theodosius Dobzhansky, “Introductory Remarks,” in Ayah and Dobzhansky (n. 9 above), p. 1.
14. . C. N. Hinshelwood, The Structure of Physical Chemistry (Oxford: Clarendon Press, 1951), p. 449.
15. . Note that the ambiguity of “nothing” has to be guarded against at this point. Sometimes “thing” is used generally (as in “something,”“anything,”“everything,”“nothing”) of any possible entity that could be referred to by a referring expression within some context of discourse. But sometimes “thing” means a visible and tangible object.
16. . D. M. MacKay, The Clockwork Image (London: Inter‐Varsity Press, 1974), pp. 42–44.
17. . Theodosius Dobzhansky, The Biology of Ultimate Concern (New York: New American Library, 1967); Ian Barbour, Issues in Science and Religion (London: SCM Press, 1966), esp. pp. 324–37; J. C. Eccles, Facing Reality (London: Longman, 1970); Arthur Koestler and J. R. Smythies, eds., Beyond Reductionism (London: Hutchinson, 1969); A. R. Peacocke, Science and the Christian Experiment (London: Oxford University Press, 1971); Marjorie Grene, ed., Interpretations of Life and Mind (London: Routledge & Kegan Paul, 1971); William H. Thorpe, Animal Nature and Human Nature (London: Methuen, 1974).
18. . Ayala (n. 10 above), p. ix.
19. . Ayala and Dobzhansky, (n. 9 above).
20. . As quoted by W. H. Thorpe, “Reductionism in Biology,” in ibid. p. 111.
21. . Ernest Nagel, “Wholes, Sums and Organic Unities,” Philosophical Studies 3 (1952): 17–32. His treatment may be summarized as follows: The word “whole” is used to refer to something with a spatial extension, to some temporal period, to any class of elements, to a property of an object or process, to a pattern of relations, to a process, to any concrete object, or to any system those spatial parts stand to one another in various relations of dynamical dependence. Corresponding to each of these meanings, the reference of the word “parts” can be explicated, and this suffices to indicate at once not only the ambiguity of these words themselves but also of the word “sum,” which is being attributed, or not attributed, to the relation of the “wholes” to the “parts,” and so also the ambiguity of the word “addition,” which is the process whereby wholes are putatively derived from parts. “Organic wholes” or “organic unities” are those systems which exhibit a mode of organization that is often claimed to be incapable of analysis in terms of an “additive point of view” (p. 26). Two kinds of this supposed “addition” have furthermore to be distinguished: “The question whether a given system can be overtly constructed in a piecemeal fashion by a seriatim juxtaposition of parts, and the question whether the system can be analyzed in terms of a theory concerning its assumed constituents and their interrelations…. However, this difference between systems does not correspond to the intended distinction between functional and summative wholes; and our inability to construct effectively a system out of its parts, which in some cases may only be a consequence of temporary technological limitations, cannot be taken as evidence for deciding the second of the above two questions” (p. 28).
22. . Ibid., p. 29.
23. . Ibid., p. 30.
24. . Thorpe (n. 20 above), pp. 109–38.
25. . J. R. Lucas, Oxford discussions, 1971–72.
26. . Nagel (n. 21 above), pp. 24–25.
27. . Ibid., pp. 29–30.
28. . Peacocke (n. 17 above), chaps. 1–3; Thorpe (n. 20 above), pp. 111 ff.; H. H. Pattee, “The Problem of Biological Hierarchy,” in Towards a Theoretical Biology, ed. C. H. Waddington (Edinburgh: Edinburgh University Press), 3:117–36.
29. . Beckner (n. 9 above).
30. . Ibid., pp. 164–45.
31. . Ibid., p. 166. Here i enumerates, from 0 to n, a series of hierarchical levels Li, to each of which is assigned a part Pi; every Pi (except Pn) is part of exactly one part at each level above i and is (except for Lo) exhaustively composed of parts at each level elow i. This is Beckner's definition of a perfect hierarchy i, The larger i, the “higher” the level; h (and 1) will denote higher (and lower) in this sense.
32. . Ernest Nagel, The Structure of Science (New York: Harcourt Brace & Co., 1961), chap. 11.
33. . Ibid., p. 364.
34. . C. G.Hempel,“Reduction: Ontological and Linguistic Facets,”in Morgenbesser  et al. (n. 11 above), pp. 179–99.
35. . Lucas (n. 25 above). The relevant passage in his contribution is as follows: “Only rather restricted theories are complete. As soon as a theory is at all rich, propositions can be expressed in it which are neither necessitated nor disallowed by the axioms (i.e., which neither can be proved from the axioms nor will lead to an inconsistency if added to the axioms). This means that although the theory will be able to make many detailed predictions about the course of events, granted some set of initial conditions, it will not be able to divide all conceivable descriptions of states of affairs into just two classes, those which must occur, granted those initial conditions, and those which cannot, in view of them, possibly occur. There is some range of uncertainty so far as any particular theory is concerned, and therefore some room for further theories and further modes of explanation. If it had been otherwise–if the relevant theories were complete–then the mechanist could have maintained that even if the biologist found biological explanations more explanatory, it could only be a matter of personal taste, for either the biological explanations accounted for exactly the same phenomena as the mechanist's explanations or else they were wrong. But incomplete theories can be compatible without being the same in all their consequences. And so we can accept the possibility of physical and chemical explanations without thereby excluding that of biological explanations that are essentially different. The incompleteness theorems of mathematical logic give a further bonus. Gödel not only gave a formal proof that any consistent theory rich enough to include elementary number theory must contain a well‐formed formula which was consistent with the axioms but not provable from them; he also argued, but this time of necessity informally, that this well‐formed formula was in fact true. He could explain why it was true, although not within the confines of the formal theory he was considering–if he had been able to explain within elementary number theory why it was true, he would have proved it, and the formula would have been a provable one just like all the theorems. But the Gödelian formula, although it cannot be formally proved within the theory, can nevertheless be informally shown to be true. So that beyond the formal proofs of the theory, there are others which we find cogent, although they cannot be expressed within the formalism of that particular theory. In a similar fashion we may expect that, besides the explanations offered by any one scientific theory, there will be insights from outside that theory, which we find quite convincing and entirely explanatory, and which can explain some things which the original theory could not. Thus not only are explanations of one, in some sense more basic, type compatible with explanations of a more sophisticated type, but they positively require them.”
36. . Conveniently grouped and summarized into four paradigms by K. F. Schaffner in “Approaches to Reduction,” Philosophy of Science 34 (1967): 137–47, where he elaborates a general reduction paradigm yielding the earlier ones as special cases.
37. . If, like one group of authors (Karl R. Popper, Paul K. Feyerabend, and Thomas R. Kuhn), one considers that a complete reduction of a Th to a Tl is not likely, then in the development of sciences new theories will tend to be seen as making cataclysmic breaks with the old; whereas, if one thinks reduction is possible, then old theories may be regarded as reduced by the new ones which replace them. (E.g., Schaffner [ibid.]; cf. also the stretching of reduction to deal with cases where there is an abrupt contradiction [and not a deductive relation] between the old and the superseding new theory by relating them both to the observations they both explain [J. G. Kemeny and P. Oppenheim, “On Reduction,” Philosophical Studies 8 (1956): 6–19].) The latter position represents some widening of the scope of what is meant by reduction, although the central idea is always that of explaining a theory Th in terms of a theory Tl from a different branch of science, corresponding (usually) to a different level in the hierarchy of systems which the sciences study. E.g., from this more historical perspective, Popper (“Scientific Reduction and the Essential Incompleteness of All Science,” in Ayah and Dobzhansky [n. 9 above], pp. 259–84) asserts that, while in all the sciences the attempt to make reductions is justified on methodological grounds and must continue because we learn so much that is fruitful even from unsuccessful attempts and while nothing is as great a success in science as a successful reduction, yet hardly any major reduction in science has ever been completely successful, for there is, he argues, almost always an unresolved residue left by even the most successful attempts at reduction. There is an unresolved residue even in the reduction of chemistry to physics since the heavier elements have an evolutionary history and their coming into existence is rare in the cosmos, so that cosmological considerations enter in addition to those of quantum physics–and a theory of evolution is even more indispensable in biology. The “deductions” involved in reducing one branch of science to another are not, in practice, strict, for they involve all sorts of approximations, simplifications, and idealizations.
38. . C. F. A. Pantin, The Relations between the Sciences (Cambridge: Cambridge University Press, 1968); P. W. Anderson, “More Is Different: Broken Symmetry and the Nature of the Hierarchical Structure of Science,” Science 77 (1972): 393–96; Peacocke (n. 17 above), chap. 2; Peter Medawar, “A Geometric Model of Reduction and Emergence,” in Ayah and Dobzhansky (n. 9 above), pp. 57–63; and L. L. Whyte, A. G. Wilson, and D. Wilson, eds., Hierarchical Structures (New York: Elsevier Publishing Co., 1969).
39. . Medawar, p. 61.
40. . Peacocke (n. 17 above), pp. 89–90.
41. . Using the metaphors of the subsection entitled “Hierarchies of Systems, Theories, and Sciences” rather than those of Medawar himself.
42. . Beckner (n. 9 above), p. 170, and Nagel (n. 21 above).
43. . Beckner, p. 170.
44. . Ibid.
45. . The confusions to which Beckner (pp. 168–70) draws attention, partly as a result of current analyses, inter alia, of the description of events, are (1) saying that, because a theory T0 appropriate to the zeroth level provides some basis (“cash value”) for regarding every part P0 at this level L0 as of a certain kind K0, then the set of wholes w contains only parts of kind K0 (this ignores the distinction between being exhaustively composed of parts Ki and being composed only of parts Ki); (2) saying that, because every part Pi at level i in a hierarchy is composed of parts Pi‐l, of the level below (i‐1), any theory Ti must be reducible to a theory Ti‐1; but alternative descriptions of the same thing, whether or not they belong to theories at different levels, need have no logical connection, and neither description entails the other; (3) the fallacy of concluding from the premise that since a j‐level phenomenon is of kind Kj it is not of kind Ki either, where i < j; the fact of hierarchical organization is sufficient ground for holding that events at the j‐level, in addition to their status as being of kind Kj, are also events of each kind Ki, where i (4) if a j‐level theory is autonomous with respect to i‐level theory (i<j), then there are j‐level processes (events, effects, phenomena) that are autonomous with respect to i‐level processes. (See the subsection entitled “Hierarchies of Systems, Theories, and Sciences” above, where this is discussed more fully.)
46. . The label “reduction” is applied only to theories, in the sense used by Nagel (n. 32 above), so the opposite of theory autonomy is theory reduction. The word “reduction” is not applied to processes, so the opposite of process autonomy is not usually called “process reduction,” but the description is simply negatived.
47. . Beckner (n. 9 above), p. 174.
48. . Ibid.
49. . Ibid.
50. . Medawar (n. 38 above) cities “heredity,”“infection,”“immunity,”“sexuality,” and “fear” (p. 57), but many others were also pointed out in the conference on problems of reduction in biology by other authors who develop this emphasis in their own particular ways. Ernest Boesiger (“Evolutionary Theories after Lamarck and Darwin,” in Ayala and Dobzhansky [n. 9 above]) stresses that the patterns of explanation of the adaptive or “purposeful” character of organisms which invoke use and disuse or natural selection are distinctive of biology and are “organismic” (p. 42). Gerald M. Edelman (“The Problem of Molecular Recognition by a Selective System,” in ibid., pp. 45–48) has described how, methodologically, the current selective theory of antibody formation emerged only when attention was focused on two levels in the system, that of the cell and that of antibodies at the molecular level; the latter alone failed as an explanation. G. Ledyard Stebbins (“Adaptive Shifts and Evolutionary Novelty: A Compositionist Approach,” in ibid., pp. 285–306) sees the study of organic evolution as polarized around two widely distant forms: evolution in the broad sense (“a succession of events that took place over billions of years of time, and gave rise successively to living matter… and finally to man…” [p. 285]), which can be studied only by a compositionistic activity drawing on information from a wide variety of sources (systematics, paleontology, population genetics), and evolution at the level of populations, which looks for changes that can be observed by a scientist in a much shorter time through quantitative, experimental methods. Theodosius Dobzhansky (“Chance and Creativity in Evolution,” in ibid., pp. 307–38) shows that evolutionary theory requires concepts that play no part outside biology (“Mutation, sexual recombination and natural selection are linked together in a system which makes biological evolution a creative process” [p. 336]). Henryk Skolimowski (“Problems of Rationality in Biology,” in ibid., pp. 221–22) emphasizes that evolved man is cognitive, comprehending evolution; and this itself is life enhancing and involved in evolution, so that ideas such as “feedback,”“information,”“environment,” and “past experience” become normative, to be made sense of only in a system that admits values and norms. This necessitates an “evolutionary rationality” introducing “open‐ended concepts,”“growth concepts,” and “normative concepts.”
51. . J. J. C. Smart, Philosophy and Scientific Realism (London: Routledge & Kegan Paul, 1963), chap. 3.
52. . Michael Polanyi, Personal Knowledge (London: Routledge & Kegan Paul, 1958); The Study of Man (London: Routledge & Kegan Paul, 1959); The Tacit Dimension (London: Routledge & Kegan Paul, 1967); “Life Transcending Physics and Chemistry,” Chemical and Engineering News (August 21, 1967), pp. 54–66; “Life's Irreducible Structure,” Science 160(1968): 1308–12.
53. . Much is sometimes made of the distinction that machines are artifacts and biological organisms are not. Clearly, machines are designed by man to have certain relationships and interactions among their components, so that we can predict how they will behave, within limits determined partly by our lack of knowledge of what we need to know about the components and their properties over a period of time and under stress. The labyrinthine organization of biological organisms is only gradually becoming apparent to us, and we are profoundly ignorant of it at most levels–only a small, though central, part has been unveiled by the molecular biology of the last few decades. We are largely ignorant of the organization of biological systems; we are unable to predict any but the smallest fraction of their behavior; and we have to admit that we are ignorant even of how we ought to think about them, what conceptual tools are needed, at different levels (see below). But this relative difference in our knowledge need not of itself invalidate Polanyi's arguments, even if they are vulnerable on other counts, as discussed in this paper. It is interesting to note that even systems composed of molecules, etc., obeying the essentially deterministic laws of classical physics can be shown to have behavior which is predictable only within limits: E.g., in meteorology, it is becoming clear that “no conceivable improvement of the observing network can increase this period [of weather prediction] to longer than a value lying somewhere between ten days and three weeks … The atmosphere is a physical system which is demonstrably unpredictable in practice beyond a certain time, even though it is com posed of a finite collection of objects whose interactions are governed by the laws of physics” (R. S. Harwood, Oxford discussions, 1971–72).
54. . K. F.Schaffner,“Anti‐Reductionism and Molecular Biology,” Science  157 (1967): 644–47, and “The Watson‐Crick Model and Reductionism,”British Journal for the Philosophy of Science  20 (1969):325–48.
55. . Schaffner, “Watson‐Crick Model,” pp. 345–46 (my italics).
56. . See n. 21 above.
57. . Polanyi, “Life Transcending Physics and Chemistry” and “Life's Irreducible Structure” (n. 52 above).
58. . See n. 50 above.
59. . Popper (n. 37 above).
60. . Ludwig von Bertalanffy, “Chance or Law,” in Koestler and Smythies (n. 17 above), pp. 56–76. Other relevant references are to be found in this same article.
61. . Ted Bastin, private communication, 1972; see also his “Timeless Order,” in Biogenesis, Evolution and Homeostasis, ed. A. Locker (New York: Springer‐Verlag, 1973), pp. 137–45.
62. . D. Berlinski, “Philosophical Aspects of Molecular Biology,” Journal of Philosophy 69 (1972): 319–35.
63. . Bastin, private communication (n. 61 above).
64. . See Beckner (n. 9 above), p. 166.
65. . See n. 45 above.
66. . It seems to be the case that the cosmos has in fact evidenced a succession in time wherein more and more complex arrangements of matter have appeared with these distinctive qualities, properties, etc., at least in the terrestrial corner of the universe that we can observe in detail. In this sense, the cosmic process displays emergent qualities the significance of which, I think, is pivotal for its interpretation but which it is not appropriate for me to develop here (see my Science and the Christian Experiment [n. 17 above], chap. 3, esp. pp. 102–8). “Emergence” has been used in this paper with reference only to the relation of higher‐level to lower‐level phenomena.
67. . Tim Appleton, “Consciousness in Animals,” in this issue.
68. . Robinson (n. 8 above).