Caplan and Smith on Surprise

Bryan Caplan and Adam Smith on Surprise

by Daniel B. Klein, dklein@gmu.edu
21 Feb 2012

Back in 2002 Bryan Caplan and Pete Boettke held a debate, focused on Bryan’s opening statement, “Why I Am Not an Austrian Economist … And Why You Shouldn’t Be Either” (transcript). The debate is available on Youtube (part 1 of 13).

In the transcript, Bryan writes:

Neoclassical surprise. Pete is wrong to say that probability theory cannot explain surprise. It has a very simple explanation: the occurrence of extreme, low-probability events is inherently surprising. It is surprising to win the lottery. Why? It is an extreme, low-probability event. It is surprising to have a big surprise 40th birthday party a couple days after you have a small party with your family. I know because I was there at Pete’s surprise 40th birthday party. I took a picture of him. He was surprised. But presumably he knew that such things as surprise birthday parties exist, that they are more likely to happen near your actual birthday, and that they are more frequent for birthdays divisible by 10.

Suppose we asked Bryan what probability Pete had put – or would have put – on having a surprise party, and Bryan says 1 in 10,000.

Now suppose Bryan and I sit down at a table with a deck of cards. I shuffle them thoroughly and deal five cards to Bryan. The probably of the hand received, whatever it is, is, I believe, 5!/(52x51x50x49x48). I might have gotten the formula wrong, but the point is, Is Bryan surprised when he looks down and sees the hand that he has received? No, he’s not, even though its probability is a lot smaller than that assigned to the surprise party.

Adam Smith said that surprise comes from “what is unexpected” – and I think we can safely rephrase that as what is contrary to the expected. He distinguishes surprise from wonder, which comes from encountering “What is new and singular” (HA in Essays on Philosophical Subjects, pp. 33). Surprise and wonder arouse our reflection and often admiration. All this is how Smith starts the History of Astronomy. He says:

Those panic terrors which sometimes seize armies in the field, or great cities, when an enemy is in the neighbourhood, and which deprive for a time the most determined of all deliberate judgments, are never excited but by the sudden apprehension of unexpected danger. Such violent consternations, which at once confound whole multitudes, benumb their understandings, and agitate their hearts, with all the agony of extravagant fear, can never be produced by any foreseen danger, how great soever. Fear, though naturally a very strong passion, never rises to such excesses, unless exasperated both by Wonder, from the uncertain nature of the danger, and by Surprise, from the suddenness of the apprehension.

Surprise, therefore, is not to be regarded as an original emotion of a species distinct from all others. The violent and sudden change produced upon the mind, when an emotion of any kind is brought suddenly upon it, constitutes the whole nature of Surprise. (HA in EPS, pp. 35)

So Smith would disagree with Bryan where Bryan writes that “the occurrence of extreme, low-probability events is inherently surprising”.

You experience surprise when something happens that is quite contrary to what you expected and, to quote Smith, “an emotion of any kind is brought suddenly upon” the mind. Your expectations are the projections from your working interpretations of what is going on, of what you are up to. You are surprised when experience supervenes your working interpretation. The occurrence of a low-probability event might be necessary, but it is not sufficient.

Surprise prompts us to wonder whether we should rethink our working interpretation and the judgment we exercised when we settled on that now-embarrassed interpretation. No doubt surprise has been vital evolutionarily, for it gets us to fashion new and better interpretations.

In The Theory of Moral Sentiments, we feel surprise and wonder when we encounter an example of conduct that is exceptional, particularly in any of the virtues, and the admiration may inspire emulation and the internalization of the conduct and judgment exemplified. Surprise, therefore, is an important part of the invisible hand in morals. We also find many mentions of surprise in the Lectures on Rhetoric Belles Lettres.

Anthony J. Evans and Jeffrey Friedman talk a lot about interpretation and surprise in their article, “’Search’ Vs. ‘Browse’: A Theory of Error Grounded in Radical (Not Rational) Ignorance, 2011, Critical Review 23(1-2): 73-104. To search for earlier discussion, some names I would pursue are Jack C. High, Martin Ricketts, Brian Loasby, G.L.S. Shackle, and Mario Rizzo and Gerald O’Driscoll (pp. 107-109 in The Economics of Time and Ignorance).

Why do I set out these and other captious remarks about Bryan? It is because I consider him one of the great liberal thinkers of our times, and yet I am uncomfortable with a certain quality in his thinking, which I guess is best described as undue rationalism. One manifestation of that quality is his flattening of knowledge down to information. The fascinating thing about Bryan’s statements about surprise is that he himself is a wonderful and creative artist.