ABSTRACT
John Rae, in 1834, objected to The Wealth of Nations on the grounds that it constituted a totally unscientific work. Smithian procedure, as he saw it, was non-experimental for Smith failed to justify empirically his axioms (and this despite Bacon’s earlier warning against the propensity to generalize from a few familiar notions, and indictment of such procedures as evidence of a disinterest in attaining ‘a knowledge of nature’) (Rae 1834:334-5). Marshall took a more positive attitude towards Smithian procedure but it appears from the celebrated Appendix to the Principles on the history of economics that he greatly played down the ‘inductive’ dimension (1961 [1920]: 759n). According to Marshall’s evaluation, history provided Smith with ‘illustrations’ and thus made his exposition more palatable, but close induction from the evidence played little part in his procedure. In particular, the basic behavioural assumptions are said to be purely axiomatic or, at the most, only supported by very casual observation.1 Much the same view seems to have been taken by J.E.Cairnes who observed that ‘when [Smith] has recourse to history, it is always in illustration or confirmation; he never makes it the basis of his doctrines. He first lays the foundation deep in the principles of human nature and the physical facts of the external world; the subsequent reference to historical events is merely in illustration of the mode in which the laws thus established operate’ (1888:114-15).2 And Bagehot’s assertion that Smith maintained as a fact that ‘there was a Scotchman inside every man’ on a feeble basis of evidence is well known (1911:125-6).
