The theory-observation distinction

Philosophical Review

Volumn 105, Issue 2, 1996, Pages 173-230

  Kukla, Andre ^a  

^a UNIVERSITY OF TORONTO   (Canada)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 53149128812     PISSN: 00318108     EISSN: 15581470     Source Type: Journal    
DOI: 10.2307/2185717     Document Type: Article

References (75)

1. Bas C. van Fraassen, The Scientific Image (Oxford: Oxford University Press, Clarendon Press, 1980)
2. Jerry Fodor, "Observation Reconsidered," Philosophy of Science 51 (1984): 23-43.
3. Unless otherwise indicated, all occurences of 'realism' and 'antirealism in this article should be understood as referring to the epistemic variety of realism and antirealism. I'll try to avoid van Fraassen's term, 'constructive empiricism', for its use in the literature is ambiguous between (1) the general thesis of epistemic antirealism, and (2) the complete, multifaceted philosophy of science expounded in The Scientific Image. What is it about the observational that's supposed to warrant its special epistemic status? There's a tradition of attributing to van Fraassen the view that we should never believe theoretical hypotheses because they can never have more than extremely low probabilities. I'll take exception to this interpretation in a later section. Most of what I have to say, however, can be said without specifying the source of the special privilege that is supposed to be enjoyed by the observational. In particular, when I talk, as I sometimes do, about some hypotheses' incurring greater "epistemic risks" than others, I don't mean to invoke the probabilistic interpretation. I use the phrase to refer genetically to whatever deficiencies our theoretical claims are supposed to be afflicted with (according to antirealists) that renders them unsuitable as objects of belief. The same applies to such stylistic variants as 'epistemic privilege', 'epistemic trouble', 'epistemic problems', etc.
4. Grover Maxwell, "The Ontological Status of Theoretical Entities," in Scientific Explanation, Space and Time, ed. Herbert Feigl and Grover Maxwell (Minneapolis: University of Minnesota Press, 1962), 3-27;
5. Michael Friedman, "Review of van Fraassen (1980)," Journal ofPhilosophy 79 (1982): 274-83
6. Jeff Foss, "On Accepting van Fraassen's Image of Science," Philosophy of Science 51 (1984): 79-92;
7. Richard Creath, "Taking Theories Seriously," Synthese 62 (1985): 317-45;
8. Alan Musgrave, "Realism versus Constructive Empiricism," in Images of Science, ed. Paul M. Churchland and Clifford A. Hooker (Chicago: University of Chicago Press, 1985), 197-221.
9. Fodor, "Observation Reconsidered," 26
10. Thomas S. Kuhn, The Structure of Scientific Revolutions (Chicago: University of Chicago Press), 113.
11. Jerome Bruner, "On Perceptual Readiness," Psychological Review 64 (1957): 123-152
12. Kuhn, Scientific Revolutions, 150.
13. Daniel Gilman, "What's a Theory to do .. with Seeing? or Some Empirical Considerations for Observation and Theory," British Journal for the Philosophy of Science 43 (1992): 287-309.
14. It could be argued that 'all' is stronger than necessary here-that the impossibility of theory-neutral observation already follows from the narrower thesis that the cognitive differences that are implicated in the choice among competing theories have an effect on perception. This hypothesis is still far broader than any that the New Look experiments can reasonably be said to have addressed. Moreover, the supposedly narrower hypothesis may well be equivalent to the stronger one in the end, for it is plausible that for any cognitive difference, there is some theoretical choice in which this difference is implicated. See also note 26.
15. Fodor, "Observation Reconsidered" 34.
16. Jerry. Fodor The Modularity of Mind (Cambridge: MIT Press1983),64
17. Fodor" Observation Reconsidered"38
18. Observation Reconsidered Ibid., 38.
19. Observation Reconsidered Ibid., 30.
20. Observation Reconsidered Ibid., 31.
21. Observation Reconsidered Ibid.
22. Paul M. Churchland, "Perceptual Plasticity and Theoretical Neutrality: A Reply to Jerry Fodor," Philosophy of Science 55 (1988): 169-170
23. Jerry Fodor, "A Reply to Churchland's 'Perceptual Plasticity and Theoretical Neutrality'," Philosophy of Science 55 (1988): 189.
24. Churchland, "Perceptual Plasiticity," 20.
25. Kuhn, Scientific Revolutions, 97, emphasis added.
26. Fodor, Reply to Churchland 195.
27. Reply to Churchland Ibid.
28. Is the output of P prepositional? Does P provide C with the proposition that it looks as though there's an apple on the desk, or does it pass along a picture of an apple on a desk, which C then converts into a proposition on its own? This is the kind of detailed scientific question that can only be settled by a program of empirical research. One thing is certain, however: for perception to influence belief fixation, its content has to be converted to a proposition somewhere along the line. Moreover, if it's to provide us with a pool of commensurable beliefs, at least part of the conversion to propositions has to be performed by mechanisms that function in the same manner in all of us. For present purposes, it will do no harm to allocate this function to P itself.
29. Gilman What's a Theory to Do... 307
30. Jerry Fodor, "The Present Status of the Innateness Controversy," in his Representations: Philosophical Essays on the Foundations of Cognitive Science (Cambridge: MIT Press, 1981), 257-316.
31. It might be maintained that this 'all' is too strong-that the necessary condition for incommensurability is only that our theories influence those of our perceptions that confirm or disconfirm them. I suspect that this weakened claim is equivalent to the original, for it seems overwhelmingly likely that every perception is epistemically relevant to some possible theoretical choice. Whether or not this is so, the thesis that Churchland argues for does not, by itself, produce an incommensurability problem. To produce an incommensurability problem, you have to establish at least that every theory influences all the perceptions that are relevant to its own confir-mational status. But the hypothesis that Churchland tries to establish is only that every theory influences some perceptions. Compare note 9, which makes essentially the same point about the previous universal quantifier.
32. Kuhn, Scientific Revolutions, 113.
33. 28Andre Kulda, "Some Limits to Empirical Inquiry," Analysis 54 (1994)
34. Fodor, "Observation Reconsidered," 23. Granny distinguishes here between two processes of belief fixation rather than two categories of scientific hypotheses. These distinctions don't line up in a one-to-one fashion, for we may learn the truth of an observation statement by a process of inference (for example, by inducing that the emeralds we've never seen are green). But it's clear what Granny's distinction between hypotheses would be: a hypothesis is observational just in case it can by established noninferentially.
35. Kukla,Some Limits to Empirical Inquiry."
36. Bas C. van Fraassen, Laws and Symmetry (Oxford: Oxford University Press, Clarendon Press, 1989).
37. Van Fraassen, Scientific Image, 14.
38. "Theoretical Entities," 13.
39. Theoretical Entities Ibid., 9.
40. Foss, On Accepting van Fraassen's Image of Science 85-86
41. Lazw and Symmetry , 193.
42. Maxwell Theoretical Entities 11
43. Paul M. Churchland, "The Ontological Status of Observables: In Praise of the Superempirical Virtues," in Images of Science, ed. Paul M. Churchland and Clifford A. Hooker (Chicago: University of Chicago Press, 1985), 35-47.
44. Churchland, "Perceptual Plasticity."
45. Churchland, "The Ontological Status of Observables"
46. Fodor "Reply to Churchland." 190
47. Van Fraassen, Scientific Image, 18.
48. Bas C. van Fraassen, "Empiricism in the Philosophy of Science," in Churchland and Hooker, Images of Science, 256.
49. Images of Science Ibid., 256-257
50. Paul K. Feyerabend, "Science Without Experience," Journal of Philosophy 66 (1969).
51. Churchland, "The Ontological Status of Observables" 42-43
52. Van Fraassen, "Empiricism in the Philosophy of Science" 256
53. Van Fraassen, "Empiricism in the Philosophy of Science" 256 Ibid.
54. Van Fraassen, "Empiricism in the Philosophy of Science" 256 Ibid.
55. Maxwell "Theoretical Entities" 11
56. "Theoretical Entities" Ibid., 9-10.
57. Foss, "Van Fraassen's Image of Science"
58. Ronald N. Giere, "Constructive Realism," in Churchland and Hooker, Images of Science, 75-98;
59. Musgrave, "Realism versus Constructive Empiricism"
60. A. Rosenberg, "Protagoras Among the Physicists," Dialogue 22 (1983): 311-317
61. Van Fraassen,"Empiricism in the Philosophy of Science" 256
62. Friedman, "Review of van Fraassen."
63. "Review of van Fraassen"Ibid., 278.
64. Van Fraassen, Scientific Image, 46.
65. Scientific Image Ibid., 46-47.
66. Scientific Image Ibid., 12.
67. Scientific Image Ibid., 64.
68. Giere, "Constructive Realism."
69. Van Fraassen, Scientific Image, 58.
70. I am grateful to an anonymous referee for pointing this out to me.
71. Stephen Leeds, "Constructive Empiricism," Synthese 101 (1994): 187-221.
72. Synthese Ibid., 195-196
73. Synthese Ibid., 196.
74. What one does in a situation like this is pardy a matter of philosophical taste. There are many who would feel that there's no need to pay any further attention to antirealist arguments unless and until the undreamed of vindicatory concept materializes. I don't think that this attitude can be faulted. But neither would it be a mistake to investigate the second-round case for antirealism, wherein arguments about the epistemic differences between the theoretical and the observational are exchanged, under the supposition that we have a coherent theory-observation distinction. Of course, it may turn out that the ultimate disposition of some of the second-round arguments will depend on how, precisely, the distinction is drawn. There are bound to be some limits on the work that can be done with a placeholder in lieu of a real concept. In practice, however, much of the second-round argumentation seems to be unaffected by this limitation.
75. Of course, as noted above, there's always the possibility that a fourth distinction might be devised that vindicates antirealism. Thus it might be considered an overstatement to claim that the rejection of hypothesis (2) would eliminate antirealism from consideration. But the rejection of hypothesis (2) would constitute as strong a case against antirealism as there can presently be.