Terence Hutchison and Frank Knight: a reappraisal of their 1940–1941 exchange

Abstract

The person arguably most responsible for the view of Hutchison as the positivist who introduced positivism into economics was Frank Knight. I argue that Knight in 1940 failed to demonstrate that Hutchison was a positivist, at least in the narrow logical positivist sense of the term. By questioning Knight's charge, I aim to challenge the conventional wisdom that identifies ‘Hutchison’ with ‘positivism’. The paper is then a first step in the argument that positivism, even in 1938, played only an inessential role in a consistent methodological position that Hutchison developed alongside his work in the history of economic thought.

Keywords:

• Hutchison
• Knight
• positivism
• uncertainty
• history
• inductivism

JEL Codes::

• B20
• B40
• B41

Acknowledgements

An earlier version of this paper was presented at the 36th Annual Meeting of the History of Economics Society, University of Colorado, Denver, 26–29 June 2009. I would like to thank Christopher Torr, Victoria Chick, Derek Wang and Ross Emmett as well as two anonymous referees for helpful comments and suggestions.

Notes

 1. For Hutchison's acknowledgement of Kaufmann's influence, see Hutchison (2009, pp. 304–306).

 2. Contrary to Emmett (2009b, p. 349, n. 1) Hutchison therefore was far from maintaining a ‘strict [Popperian] falsificationism’.

 3. Hutchison (1998, p. 73) himself cites Hausman's call with approval.

 4. While Knight (1940) appears to emphasize the conception of truth as a value (p. 26) or as a ‘social consensus’ (p. 7), he also answers his title question ‘What Is Truth?’ by saying that it is equivalent to the questions, What is observation? and What is inference? (p. 10).

 5. While it is closer to Popper's falsifiability criterion, it is not the same – see section 1.1 above.

 6. Hutchison's response concerning mistaken observations presumably refers to Knight's (1940, p. 7) examples of how everyday observations may be mistaken. According to Hammond (1991, p. 371), Knight's argument involves the notion that the testing of such primary observations ‘is inherently a theoretical exercise’, a point which Hutchison appears to miss. Whatever the case, Knight here places more importance on the notion that testing is a social activity, a point with which Hutchison ‘entirely agrees’.

 7. Hutchison (1938) draws liberally on Poincaré (1905), one of the few scientist-philosophers respected by Knight (1944, p. 262).

 8. Contrary to Caldwell (2004, p. 202), Hutchison accepts, rather than challenges, the use of introspection in economics: see Hutchison (1938, pp. 137–143).

 9. According to Knight (1941, p. 752), although there is no ‘ultimate’ difference between theoretical economics and theoretical physics, ‘yet the difference in degree and in the actual meaning of the fundamental propositions in the two fields is so great that it must be recognized as a difference in kind’. This seems to imply, contra Hutchison, a ‘fundamental discontinuity’ between the two methods.

10. One of the points Hutchison may have had in mind is Friedman's (1977, p. 452) view that ‘differences are as great among, say, physics, biology, medicine and meteorology as between any of them and economics’.

11. Contrary to Emmett (2009b, p. 347), I argue that Hutchison's methodological argument does not depend on logical positivism.

12. Although Hutchison (1938, pp. 65–72) reveals the extent to which he is influenced by positivist notions of cause, for the most part he interprets behaviour in terms of motives.

13. The Weber reference was translated by Dr Sabine Marschall.

14. Given Machlup (1955, p. 8, n. 24), it may be well to note here that Hutchison is by no means urging any ‘political agenda’ in economics. And, contrary to Emmett (2009b, pp. 347–348), the purpose of Hutchison's appendix is not to argue ‘that classical economics was utopian while social planning on the basis of empirically verified economic knowledge was not’; nor to defend ‘Keynesianism and socialism’ (see Hutchison 1977b, 1981, pp. 108–154). Hutchison is not in favour of planning and against laissez-faire. Instead his concern here is entirely methodological. It was sparked off by the ‘quite preposterous’ claims made by von Mises regarding economic liberalism (Hutchison 1960, p. xxii). For example, von Mises (1933) claims that ‘science has succeeded in showing that every system of social organisation that could be conceived as a substitute for the capitalist system is self-contradictory and unavailing’ (Mises 1985, pp. 88–89; the 1933 reference is cited in Hutchison 1938, pp. 184–185). In particular, Hutchison sought to confront von Mises's extreme apriorism and to emphasize that, because of the deductive a priori approach adopted in much of the 1930s debate on planning versus laissez-faire, ‘there was no legitimate a priori conclusion either way’ (1938, p. 181; see also 1960, p. xxii).

It is not very convincing for anti-planners to prove that the individual capitalist is the better entrepreneur by, apparently, tacitly assuming a world of perfect foresight where no entrepreneuring … is necessary … Equally unconvincing … are the airily confident generalisations of the planner unsupported by any agreed or concrete proposals for economic organisation. (Hutchison 1938, p. 182)

Hutchison's point is that, given the complex issue of planning versus laissez-faire, detailed empirical investigation in addition to purely deductive theorizing is required before any sweeping claim can be made that ‘science has succeeded in showing [that every non-capitalist social order] is self-contradictory and unavailing’.

15. In particular Caldwell has explained that ‘when I said Hutchison introduced positivism into economics, what I chiefly meant was that he was the first economist in a major book to quote logical positivist philosophers in arguing about methodology’; see Hart (2002, p. 376, n. 4). In the footnote that immediately follows, he accepts that one might infer from Caldwell (1982, p. 115) that he associated positivism with mathematization but explains that, ‘the major point of my book was that positivism means many things, that the rhetoric of positivism had entered economics, and that its entrance was idiosyncratic and haphazard, and that our actual practice is often different from our rhetoric. I agree with Hutchison that the association of positivism with math is a doubtful association’.