Is It Justifiable to Abandon All Search for a Logic of Discovery?

Abstract

In his influential paper, ‘Why Was the Logic of Discovery Abandoned?’, Laudan contends that there has been no philosophical rationale for a logic of discovery since the emergence of consequentialism in the 19th century. It is the purpose of this paper to show that consequentialism does not involve the rejection of all types of logic of discovery. Laudan goes too far in his interpretation of the historical shift from generativism to consequentialism, and his claim that the context of pursuit belongs to neither discovery nor justification is based on narrow interpretations of the contexts of discovery and justification. As a result, Laudan draws unwarranted conclusions concerning both the early and contemporary defenders of a logic of discovery. A methodological logic of discovery—which involves self‐corrective methods of hypothesis generation that promote the long‐term goals of science and which require consequential support for justification—is a type of logic of discovery that survives the shift to consequentialism.

Acknowledgments

I would like to acknowledge the assistance of John Greenwood and two anonymous referees of International Studies in the Philosophy of Science on earlier versions of this paper.

Notes

[1] Laudan’s article was originally published in Nickles (1980b, 173–183) but he revised it for Laudan (1981, 181–191). Since the revisions are relevant to the argument of this paper, page numbers correspond to the 1981 version.

[2] This term is from Gutting (1980).

[3] One exception is Hoyningen‐Huene (1987).

[4] Laudan’s argument against the philosophical relevance of heuristics in the context of discovery is an historical one. It is not to be confused with the conceptual argument against a logic of discovery offered by both Popper (1959) and Reichenbach (1938). The flaws in the conceptual argument can be found in McLaughlin (1982b, 74–75).

[5] In a response to McLaughlin, Laudan (1983, 321) concedes that fallibilism does not preclude the search for a logic of discovery, so his rejection of a logic of discovery is based entirely on the shift to consequentialism.

[6] Nickles (1987a, 515) also argues that the shift was mainly a philosophical fiction and that working scientists continued to develop heuristics for problem‐solving.

[7] A full explanation of the various uses of empirical consequences in Bacon and Descartes is beyond the scope of this paper. See Blake, Ducasse, and Madden (1960) and McMullin (2001) for extended treatments of this issue.

[8] Popper holds that novel consequences are the only epistemically relevant evidence. See Nickles (1987b) for an extensive treatment of the various types of consequentialism.

[9] This rationale for a logic of discovery differs from those of both McLaughlin and Nickles. McLaughlin argues that methods of generation have epistemic relevance because he denies the divorcing of discovery from justification such that the same methods of justification are also used in the context of discovery. Nickles argues for discoverability, in which a logical method of discovery is not temporally necessary, but in which a logical discovery procedure can be rationally reconstructed. This reconstruction provides the epistemic relevance.

[10] Certain ‘friends of discovery’, such as Achinstein (1970) and Thagard (1978), have argued that principles of justification are the same principles used to generate hypotheses. This type of non‐autonomous logic of discovery differs from a methodological logic of discovery.

[11] For a contrary view, see Buchanan (1982).

[12] McLaughlin (1982a) also challenges Laudan’s context of pursuit, but only for a non‐autonomous logic of discovery.

[13] Kordig (1978) draws a similar distinction.

[14] This term was coined by Hanson (1958).

[15] See Giere (1973) for a dissenting view.

[16] As was shown in section 2, above, both McLaughlin and Nickles have challenged Laudan’s claim that there is no epistemic warrant for a logic of discovery. McLaughlin has done so by re‐interpreting the discovery/justification distinction, and Nickles (1987a, 510) has argued for discoverability. Nickles (2000, 91) also argues that the context of discovery is necessary to epistemology because it provides the material for the context of justification. Further, Hintikka (1998) has argued that abduction is the fundamental problem of epistemology. So Laudan’s claim that there is no epistemic warrant for a logic of discovery has been challenged.

Additional information

Notes on contributors

Mehul Shah

Mehul Shah is at the Department of Philosophy, Baruch College.