On Ostracism and proto-mechanism design. With Hume, Spinoza, and Toland showing up.

A few days ago, I noted that in the context of defending the European balance of power, Hume illustrated and compared its underlying logic to ancient practices of ostracism: “The same principle, call it envy or prudence, which produced the Ostracism of Athens, and Petalism of Syracuse, and expelled every citizen whose fame or power overtopped the rest; the same principle, I say, naturally discovered itself in foreign politics, and soon raised enemies to the leading state, however moderate in the exercise of its authority.” (“Of the Balance of Power” (1752)). This was first published in a collection, Political Discourses.

Now, by 1752 Hume had published at least three major collection of essays that included topics related to political economy and political philosophy. One notable feature of these essays (that has been noted before!) is that many of the essays involve an interest in what one may call social mechanisms. The interest is empirical and prescriptive or normative. I dub this ‘proto-mechanism design.’

In particular, in many cases Hume is willing to isolate a few major foundational social causes (which he often calls ‘principles’) and rely on a fairly sober account of human motivation, and how we respond to incentives, in order to explain recurring social mechanisms, including equilibrium and disequilibrium processes, unintended consequence processes, and many kinds of social regularities (which he often calls ‘maxims’).

His treatment of causation, especially the rules, is a further background commitment that enters into his analysis of social mechanisms. In particular, the conjunction of Hume’s fourth and seventh rules produces a new rule: “An effect always holds proportion with its cause” (“Of Interest” (1752)). This I have called, ‘Hume’s ninth rule.’ In a very fine (2006) essay, David Wootton, “Liberty, metaphor, and mechanism: the Origins of Modern Constitutionalism." Liberty and American experience in the eighteenth century. (Liberty Fund) also singles it out (p. 210).

Proto-mechanism design shows up throughout Hume’s essays. For example, Hume writes in his (1741) “That Politics may be reduced to a Science,” that self-governing states (we might call them republics or democracies) are “the most happy for those who partake of their freedom; yet are they the most ruinous and oppressive to” those they have conquered or lack full citizenship. (I often think of it when reflecting on the current state of the Israeli-Palestine conflict.) Once the operation of a social mechanism has been analyzed and is abstracted from local details, it becomes prescriptive in the design of constitutional form or more lower level policy.

For obvious reasons students of the American founding and modern economists have traced their own version of interest in social mechanisms back to Hume’s maxim, that, “in contriving any system of government, and fixing the several checks and controuls of the constitution, every man ought to be supposed a knave, and to have no other end, in all his actions, than private interest.” (From (1741) “Of the Independency of Parliament.”) (Spinoza and de Lolme have similar pithy statements as does Kant later.) Modern scholars often don’t note that Hume quickly adds that “it appears somewhat strange, that a maxim should be true in politics, which is false in fact.”

That is, Hume explicitly restricts the domain of operation of his own knave maxim. He does not presuppose it in all his accounts of social mechanism. For example, his brief analysis of ostracism actually requires as a presupposition that citizens not act as knaves (which may motivate them to join a would be oligarch), while recognizing that the motives they are expected to act on (prudence and jealousy) are not elevated. Of course, prudence and jealousy (which in some senses is closer to our fear and animosity) may be fully self-interested, but they need not be. (The anatomist of the social mechanism can be agnostic about that!)

Hume really makes three claims about ostracism: (i) it rests on one — or if two then they are mutually supporting — major sources of motivation; (ii) it operates by way of a fairly simple principle/rule: viz, ‘identify leading citizens and then expel them.’ Hume hints at a further simple rule: a leading citizen can be identified in virtue of what one may call formal characteristics about their source of power or fame (usually extensive land-owning, being a creditor, an admired public speaker), and the procedure to expel them is, in turn, characteristic of the form of government (a vote). (iii) it is analogous to the operation of the balance of power in virtue of (i) and (ii). This analogy only works if human motivation and the grounds of state action are structurally alike (which means states are treated, in Westphalian fashion, as if a unified person in a certain respect).

Now, the Humean practice of proto-mechanism design is not original with Hume. For this reason alone it is worth reading the Wootton essay I mentioned. Wootton is right to say that analysis of social mechanism is in European intellectual history a relatively modern project (although I am myself more inclined than he is to see more antecedents among the ancients [consider the iconic and competing ancient analyses of Thales’ cornering of the market in olive presses] and locate an early flowering in Medieval Islamic writings). Hume makes his debt to Machiavelli explicit. Bacon (who also had read his Machiavelli as the Vatican censors noted—the latter I learned from scholarship by Marta Fattori) is also an important antecedent. As regular readers know, I think of Thomas More and Spinoza as important practitioners in the genre (see here for scholarship on Spinoza). Hume’s relationship to Spinoza is speculative, but we know that Hume is a careful reader of More, who (despite More ending his life as a religious fanatic which Hume does not find endearing) he, in fact, praises rather highly in the History: “besides the ornaments of an elegant literature, possessed the highest virtue, integrity, and capacity.”

Now, Wootton’s essay alerted me to the fact that Walter Moyle (1672–1721) is a major balance of power theorist, and also an important practitioner of proto-mechanism design. In fact, for Wootton, “Moyle’s essay of 1698…represents the birth of a new language and a new paradigm: he writes of checks, of controls, of the balance of power (although perhaps not in its modern meaning), of [social] machinery.” (p. 226) Moyle is a fascinating character, and since reading Wootton I have happily been exploring Moyle’s writings and scholarship on them.

Now in a lovely (2011) essay, “Walter Moyle's Machiavellianism, declared and otherwise, in An Essay upon the Constitution of the Roman Government,” Vicki Sullivan argues,

Moyle becomes increasingly reliant on Machiavelli's analysis that shows how the Roman republic became vulnerable to its ambitious citizens. Moyle echoes Machiavelli's conclusion that the Roman republic declined because it failed to seek timely reductions to its beginning. At this point, Moyle prescribes the harsh remedies that Machiavelli claims can counter such ambition. (p. 121)

So, I kind of wondered whether Moyle was Hume’s source, and looked whether he had written on ostracism or petalism (or banishment) in republican government. But I was unable to locate anything. In fact, much to my surprise there is very little joint discussion of ostracism and petalism in English. Harrington (who draws on Aristotle) has an account of Athenian ostracism throughout The Oceana (a book Hume praises).

The only source I found that conjoins ostracism and petalism is a translation from the Latin of Evenkellius, entitled ‘Gymnasiarchon, or the School of Potentates,’ (the translator is Thomas Nash (1588–1648)). A rather obscure work. Rather than being a dead end, here’s where things become neat.

The Gymnasiarchon actually treats ostracism and petalism together (go to p. 100) and cites Diodorus Siculus, presumably the Bibliotheca historica. This work (much of it lost) discusses petalism at length in Volume 11, chapter 87, alongside Athenian ostracism. I had looked at it when I wrote my earlier digression on the analogy between ostracism and the balance of power.

Now we know Hume was an avid reader of Diodorus Siculus and often cites and quotes him. (I don’t think this has attracted much attention.) So it is highly likely that he is Hume’s source here.

Another avid reader of Diodorus Siculus is, Harrington’s editor, Johan Toland (who is also rather important to Wooton’s argument on the modern invention of proto-mechanism design). This I already knew because Toland draws on Diodorus in key moments in The Letters to Serena. Toland is also a deep reader of Spinoza (and disguised it badly). Much to my sadness, I found no treatment of Volume 11, Chapter 87 in Toland. So, I couldn’t close this digression showing how Toland had anticipated Hume.

However, in his explosive and hilarious Nazarenus, Toland does note that ancient Hebrews had a version of the practice of ostracism (citing Genesis 17:14), or banishment. Toland is here actually quietly correcting (or at odds with) Spinoza, who at TTP 17:78 somewhat surprisingly claims that the Ancient Hebrews did not condemn citizens to exile.

In fact, Spinoza (who shows familiarity with Diodorus) is a critic of the practice of punishment by exile by Republics. The criticism is easily missed because it is part of his argument for free speech: “What greater evil can be imagined for the Republic than that honest men should be exiled as wicked because they hold different opinions and don’t know how to pretend to be what they’re not?” (TTP 20:34 in Curley’s translation.) The self-undermining nature of banishment is, in fact, the lesson Diodorus’ Syracusians draw (and they eventually repeal the law). What’s surprising is that Hume does not note that ostracism is self-undermining in the very same way as the balance of power is, as liberals since Kant have claimed.