I recently read this excellent essay by Smith scholar Erik W. Matson, who identifies and explores many connections between the ideas of Adam Smith and his intellectual mentor David Hume. (The final version of Dr Matson’s essay was published in Volume 11 of the Adam Smith Review; here is an ungated pre-print via SSRN.) To the point, among the myriad connections between Smith and Hume’s thought that Matson identifies, the one that resonated with me the most is Hume’s probabilistic picture of human knowledge, i.e. the idea that experience and observation can at best give us probable knowledge, not demonstrative or complete knowledge.

In summary, this idea that scientific theories or explanations are probabilistic (never certain) can be traced back to David Hume’s 1739 Treatise of Human Nature, where Hume himself explains that “all knowledge resolves itself into probability, and becomes at last of the same nature with that evidence, which we employ in common life.” (See Matson 2019, p. 274, citing Book 1, Part 4, Section 1, Paragraph 4 of Hume’s Treatise. As an aside, this idea can perhaps be traced even further back to John Locke’s 1689 “Essay Concerning Human Understanding”! See also here for more details about Humean probability.) For his part, Dr Matson refers to a lesser-known essay by Adam Smith titled “Of the external senses” or “OES”, which appears in Smith’s posthumous collection of philosophical essays (Smith 1982, pp. 135-168). According to Matson: “Smith uses the language of probability in reference to the status of systems of explanation on multiple occasions [in his essay on the external senses]. He speaks of the ‘probable foundations’ of philosophical doctrines; he speaks of ‘this great probability … still further confirmed by the computations of Sir Isaac Newton’ (OES 147.41).”

Why did this aspect of Hume’s (and Smith’s) thought resonate with me? Because I would argue that the concept of “truth” is itself probabilistic in nature; see here or here, for example. Alas, although Matson may very well be correct that Smith may have adopted a “Humean attitude” toward theory choice in his (Smith’s) early philosophical work, I don’t see much evidence of Adam Smith thinking in explicitly probabilistic terms in his more mature works, like Theory of Moral Sentiments or Wealth of Nations. (In fact, Matson himself describes Smith’s own back-sliding toward “truth-talk” in the last part of his (Smith’s) essay on the history of astronomy, where Smith praises Newton’s discovery of gravity as the culmination of thousands of years of sundry scientific endeavors.) Either way, as I mentioned in a different context in my previous post, the battle between absolute truth and probabilistic truth in Smith’s thought could be the subject of another paper.

P.S.: Below the fold is a bibliography of some of the works I have cited in this blog post:

• Hume, David. 2000 [1739]. A Treatise of Human Nature. Edited by David F. Norton and Mary J. Norton. Clarendon Press.
• Locke, John. 1825 [1689]. An Essay Concerning Human Understanding, available here.
• Matson, Erik W. 2019. Adam Smith’s Humean attitude towards science; illustrated by “The History of Astronomy”. Adam Smith Review, Vol. 11 (2019), pp. 265-280.
• Smith, Adam. 1982 [1795]. Essays on Philosophical Subjects. Edited by W. P. D. Wightman and John Cameron Bryce. Liberty Fund.
• Bonus cite: Guerra-Pujol, F. E. 2020. Frank Ramsey’s contributions to probability (and legal theory). SSRN, available here.