Ulysses Unbound
A Review of Three Essays on Self-Constraint
In any book review, one must pick and choose the things which are discussed. To discuss everything would be to, as in Borges, make a map as big as the territory. It would be useless. In this essay, I will take a scattershot view of the material – you will have to settle for a series of watercolors, in preference to a grand landscape.
i. Self
Some of our most fruitful ideas are simply recombinations of other ideas. Different fields have different tools, and someone knowledgeable in two worlds can solve problems in one with the methods of another. Would economics be where it is today without the application of math? Biology would not be where it is today without the development of game theory,1 and history is enlivened by the injection of causal reasoning of economics.
Jon Elster, in his book “Ulysses Unbound”, is a superb classicist, and by dint of striving (he seems to have talked to absolutely everyone) has made himself into a fine economist too. It is striking how much his classics training constitutes a lense through which to see the world, as economics is a lense to see the world. Where I might bring up a casual analogy to the increasing returns and gains from trade, he will cite an essay of Seneca, the memoirs of Montaigne, a play by Racine, and an incident from the life of Caesar. It is awe-inspiring to read, simply as a feat of erudition. Professor Elster is the model of the educated man from another century.
But he is not a hidebound, dusty old scholar. The economics in this is modern, and excellent. Elster is concerned with what our true desires are, and how we can make ourself happiest. A naive view of the world might say that our preferences can be revealed only through action, and that whatever we choose will make us happiest. To expand the field of choice is necessarily an improvement – we can simply not take up any choice which leaves us worse off.
Yet, this does not take full account of people’s actions. We have conflicting desires – we might desire to devour every Christmas cookie in the house, but also not want to grow fat. We might desire not to grow fat far more than the temporary pleasures of the cookies, and we know this keenly, but when the cookies are baked we will undoubtedly find ourselves gorging on them in the middle of the night. What is it that we truly wanted?
There are a couple reasons why we might behave like this. Firstly, we might have a discount rate (an expression of the tradeoff between present and future consumption) sufficient to make temporary pleasures worth it. This discount rate could be exponential, and the utility of waiting declines at a constant percentage rate; or it could be hyperbolic, and there is an enormous gulf between consumption today and consumption now.2 Alternatively, this behavior could be due to a strategic inconsistency. Imagine we ask for money now, and we will pay back later. If there is nothing outside of this set-up, we could just keep the money! We lack an ability to meaningfully bind ourselves across time periods.3
Things grow philosophically knotty when the only party is ourselves – or perhaps, us today and us tomorrow. Suppose our preferences change? We might want to give up drinking, and instruct someone to take every action, up to physical force, to keep up us from it – and then the next day, give it up and instruct them to let me drink. I may give my money to a safekeeper so that I am tempted not to gamble it – and yet, what am I to do when the most tempting, advantageous gamble comes along?4
Preventing these changes might not be so benign as the examples given. What are we to think of the religious fanatic who knows that, if he were to make contact with the outside world, he would change his mind? With drinking we might be able to say that, since before the event we didn’t want it, and after the event we regretted it, there is an underlying desire to avoid drinking; but when it comes to a change in views after gaining more information, who is the person really? One could think of the binding as an act of violence by the man of today against the man of tomorrow, no less than if the religious fanatic were to ban Western propaganda in their little theocracy. Today, when semaglutide derivatives are able to change our desires with an injection, we might be able to say that our true desires are those which we want when thinking calmly about our future; but that formula does not adequately describe when we keep ourselves from learning what might change our cherished beliefs.
We cannot even bind ourselves fully to not binding ourselves. In some contexts, deliberate blindness to past behavior allows us to be better off. Montaigne is quoted in this context on how a man who seeks out an affair that his wife is committing is committing a grave error. The Romans, apparently, had the custom of sending out a man ahead of someone returning home, so that any lover would not be caught unawares with the wife. An interesting example from economics is a recent paper by Cooper, Homem-de-Mello, and Kleywegt (2014) shows how agents who have access only to information particular to their own firm’s demand can reach a consistently higher price than those who have access to information on the whole market. With access to all of the facts of the world, they can’t commit to restraining themselves from myopically grabbing as much as they can.
Elster also discusses emotion as a method of overcoming strategic inconsistency. Where in the beginning of the book he spoke of emotion as that passionate irrationality which overcomes our deeper plans, we can now think of irrationality as letting us keep our deeper plans. If people know that in the throes of wrath we care little for the harm to ourselves, it discourages others from over harming us at all. An intrinsic revulsion at the thought of betraying others, if known, lets others trust us with their money, and their hearts. I was struck here by how some of the ancient authors (Seneca, for example) spoke of anger as an all-consuming, and long-lasting, passion.5
I would like to venture some thoughts on what goodness is. Elster’s book, to me, is about all the ways in which humans bind themselves and others to reach outcomes which almost everyone would agree is better than the status quo. Goodness is the ability to know how the world should be, and to act to make it happen. Goodness is patience. Goodness is kindness. Goodness is the forgetting of quarrels. Goodness is not holding out for two pennies when one is enough. Goodness is behaving the way that you would want others to behave. Goodness is honoring the trust of others. Not all cooperation is goodness, of course; criminal confederates in the Mafia are, in cooperating with each other, striving to make the world a worse place for others. But goodness is loving others, with the faith that they will love you too.
With my lense of economics, I naturally think of the Myerson-Satterthwaite theorem. It is a generalization of Myerson’s earlier work on optimal auctions (1981) and on bargaining (1979), and extends to all bargains. If bargainers are going to be engaging in many trades, it is not optimal for the individual to make all mutually beneficial trades happen. You could get more by holding out for a better price, even if this means that trades don’t happen. Bradley Larsen shows how common these breakdowns in transactions are across a variety of contexts; on Ebay, for example, fully 35% of mutually beneficial transactions never happen.
Goodness, then, is saying what your true values are, even when the mechanism is not incentive compatible.
ii. Society
The natural analogy to one binding themself is for the government – us – to bind ourselves, the government. Thus, constitutions. The body politic will grow tempted by vice and time-inconsistent behavior, and we can keep ourselves from arriving at worse outcomes by making those principles which we all agree are important inviolate. We can stave off the temptation to abridge free speech and canceling elections by making these especially difficult to impose.
Now, the analogy is not one-to-one. Government is even more limited in binding itself than humans are, for there is nothing more powerful than it. To kick the cigarette habit, I might be able to go on a cruise for two weeks, and there is nothing one could do to get cigarettes on that ship; but the government, even if the constitution cannot be amended, can simply ignore it. To me, constitutions are more a way to coordinate around hard and fast rules for political opposition. Ulysses has been bound to the mast – bound by his own hands – but he can simply pull and break his ropes.
It is also unclear if we can truly speak of the government as having a will at all. It is impossible to guarantee that there is a democratic government with unlimited options and coherent preferences which results in an outcome favored by all (Arrow); there is certainly no system of voting which is immune to people misrepresenting their preferences (Gibbard-Satterthwaite). Tullock (1959) shows how, with selfish voting, there can be no stable majority in how to allocate funding. Imagine there are a hundred farmers, each of whom have a road leading to them whose maintainance is covered by them collectively, with a tax shared equally on all. In an ideal world, each person would adopt some standard of road quality and vote for the repair of each road whose quality is below that standard (what he refers to as “Kantian” voting). Each farmer, though, can strictly improve their position by always voting to repair their own road, and voting against all other roads. They get the benefits without any of the costs. What’s more, as the number of selfish voters increases, there is no equilibrium. Any coalition of 51 voters can be beaten by another coalition. For their to be any decision reached, someone must be in charge of setting the agenda, and it is possible (with an unlimited domain, and some assumptions about preferences not being identical) for the agenda-setter to achieve any outcome they desire. (There’s a paper on this – I will need to find the citation). We need irrationality on the part of legislators and voters in order for democracy to happen at all. I will not be entirely cynical about constitutions, though. It is plainly false to say that they have had no influence on the domain of political decision making; the mere fact that we have Supreme Court decisions which were unpopular, but enforced, is evidence enough that the Constitution.
In some ways, governments can be appropriately analogized to humans. We may pass laws as the result of ill-considered passion which we come to regret. For this reason, we adopt delays in the process of making laws, and may require multiple chambers to pass it. We vest an executive with the power to veto laws, who commonly has a longer term than the representatives passing the law, or have the chambers have different term length, so that they are not all reflecting the same momentary passion of the populace. Only a sustained desire for a particular policy can succeed.
And some of the ways in which we bind ourselves have very positive results. A central bank at the command of the executive is dangerous, for she might prevail upon them to act in ways that benefit her short-run political success, but endanger the future of the country as a whole. Entering into supra-national organizations like the EU is a way of binding oneself to free market policies at home, and to lower corruption. In exchange for access to markets, countries need to agree to not subsidize domestic firms and to allow competition. The carrot allows for the stick.
It is in his choice of examples that Elster shows his classicist bent. After proceeding through the National Assembly of France in 1789, and the American Constitutional Convention, he settles lovingly on ancient Athens. The people were prone outbursts of mob violence, where any constitutional constraint would be overriden. The Athenians did not want to outlaw this entirely, since the violence at times could be justified from their point of view. We learn of Lycidas from Herodotus, who was stoned to death – and his wife and children stoned to death too by the Athenian women – for simply suggesting that the Athenians accede to a Persian demand. That was justified, and cited approvingly. But there are other instances in which the Athenian people, laboring under a temporary passion, did things which they regretted, and so they created methods of constraining themselves. There were four main ways: reconsidering a decision (anaphepsis), separation of powers, allowing for time between votes, and allowing for violence against the proposers of motions.
The last could come retroactively. According to Xenophon, after a victory in a naval battle, the generals did not adequately rescue the sailors in the water on account of a storm. The people, in their wrath, voted for them to be accused collectively of treason, voted on as one, and put to death at one. Callixeinus was the chief instigator, terrorizing the Prytanes with his supporters. After they had killed them, though, the Athenians saw the fundamental injustice of the charge, and repented. Callixeinus was put on trial, and killed, same as the generals.
I found the debate between Cleon and Diodotus, regarding the treatment of the city of Mytilene on the island Lesbos, to be fascinating, even if morally alien to us. Mytilene, a member of the Delian League, had rebelled against the Athenians; when the rebellion was crushed, so furious were the Athenians that they voted the entire male population be put to death. According to Thuycidides, the morrow brought repentance, and the Mytilinean ambassadors pleaded that the matter be put to another vote.
Diodotus takes the view that haste and passion are inimical to good decision making. Implicitly, he is arguing against constitutions. As our information changes, so should our views. We ought not be bound to decisions which we regret, but should instead be able to choose the best policy in each time period. Thus, the law should be a continual negotiation, and no past legislative body should be able to bind the present legislative body anymore than it could bind itself.
Cleon argues for constitution. A democracy cannot sustain an empire, because it is too uncertain of what it wants to do. It must commit to one policy, and stick with it. In his view, to have chosen clemency for the subjugated Mytileneans would be superior to this worst of all policies, to choose retribution and then waver toward mercy. An act of uncertain mercy in one case would engender a hundred more rebellions.
A fundamental problem for democracies is that foreign countries must believe that its foreign policy will remain consistent from party to party. If they believe the next party will rip up the deal, they may never agree to it at all. Even if you disliked the nuclear deal with Iran, the Trump administration annulling it was a serious blow to our credibility in negotiations with other leaders to come. To deter, we must be believed.
iii. The Arts
In the last section, Elster turns to the arts. I cannot, in fairness, discuss the pictorial arts; I have no systematized theory of aesthetics, nor any deep knowledge. I can observe something, and think it pleasing, but I suspect I am missing out on some ineffable experience that other people feel. I can, however, speak with some authority on the subject of music, especially classical music.
I believe that constraints in the making of music serve partly to simplify the field of possibilities, such that any progress can be made in writing, and to make the music intelligible to someone listening only once. One mustn't forget that someone might hear even the most popular orchestral works only a few times in their life in their original form, and that to hear it at all depended upon a skilled pianist. Most people in the 1800s would not be familiar with Beethoven’s symphonies, but rather with the Liszt reductions for piano. A composer cannot write entirely free-form music, for it would pass through one ear and out the other. Neither, though, could they easily compose entirely free-form music. Instead, they compose a theme, and make variations on it. The form is a trellis on which the ideas can germinate.
We need the repetition of the ideas in order to understand them. The leitmotifs of Wagner – specific musical phrases representing particular persons and ideas – allow us to hum along in the audience, even if the subtle intents are lost on us. Likewise, a fugue could conceivably be replaced by counterpoint which has no repetition, but then we would lose the hook that makes the whole comprehensible. The phrases we come to love tide us over until we can memorize the whole.
Seemingly formless composers are simply those who make form so unobtrusive as to be undetectable to the casual observer, although it is still apparent to those familiar. They are like painters who strove to make their brush strokes undetectable. To bring an example from writing, Joyce’s Finnegans Wake seems to be a bizarre, alien work, until you read the drafts. Take the episode of St. Patrick and the Archdruid (who is an idealist philosopher a la Berkeley). Read closely and carefully the final version:
Tunc. Bymeby, bullocky vampas tappany bobs topside joss pidgin fella Balkelly, archdruid of islish chinchinjoss in the his heptachromatic sevenhued septicoloured roranyellgreenlindigan mantle finish he show along the his mister guest Patholic with alb belongahim the whose throat hum with of sametime all the his cassock groaner fellas of greysfriaryfamily he fast all time what time all him monkafellas with Same Patholic, quoniam, speeching, yeh not speeching noh man liberty is, he drink up words, scilicet, tomorrow till recover will not, all too many much illusiones through photoprismic velamina of hueful panepiphanal world spectacurum of Lord Joss, the of which zoantholitic furniture, from mineral through vegetal to animal, not appear to full up together fallen man than under but one photoreflection of the several iridals gradationes of solar light, that one which that part of it (furnit of heupanepi world) had shown itself (part of fur of huepanwor) unable to absorbere, whereas for numpa one puraduxed seer in seventh degree of wisdom of Entis-Onton he savvy inside true inwardness of reality, the Ding hvad in idself id est, all objects (of panepiwor) allside showed themselves in trues coloribus resplendent with sextuple gloria of light actually retained, untisintus, inside them (obs of epiwo). Rumnant Patholic, stareotypopticus, no catch all that preachybook, utpiam, tomorrow recover thing even is not, bymeby vampsybobsy tappanasbullocks topside joss pidginfella Bilkilly-Belkelly say patfella, ontesantes, twotime hemhaltshealing, with other words verbigratiagrading from murmurulentous till stridulocelerious in a hunghoranghoangoly tsinglontseng while his comprehendurient, with diminishing claractinism, augumentationed himself in caloripeia to vision so throughsighty, you anxioust melancholic, High Thats Hight Uberking Leary his fiery grassbelonghead all show colour of sorrelwood herbgreen, again, niggerblonker, of the his essixcoloured holmgrewnworsteds costume the his fellow saffron pettikilt look same hue of boiled spinasses, other thing, voluntary mutismuser, he not compyhandy the his golden twobreasttorc look justsamelike curlicabbis, moreafter, to pace negativisticists, verdant readyrainroof belongahim Exuber High Ober King Leary very dead, what he wish to say, spit of superexuberabundancy plenty laurel leaves, after that commander bulopent eyes of Most Highest Ardreetsar King same thing like thyme choppy upon parsley, alongsidethat, if pleasesir, nos displace tauttung, sowlofabishospastored, enamel Indian gem in maledictive fingerfondler of High High Siresultan Emperor all same like one fellow olive lentil, onthelongsidethat, by undesendas, kirikirikiring, violaceous warwon contusiones of facebuts of Highup Big Cockywocky Sublissimime Autocrat, for that with pure hueglut intensely saturated one, tinged uniformly, allaroundside upinandoutdown, very like you seecut chowchow of plentymuch sennacassia. Hump cumps Ebblybally! Sukkot?
I do not blame you for finding it difficult to understand! For here is the first draft, in which the facts of the story are stated simply and plainly.
The archdruid then explained the illusion of the colourful world, its furniture, animal, vegetable and mineral, appearing to fallen men under but one reflection of the several iridal gradations of solar light, that one which it had been unable to absorb while for the seer beholding reality, the thing as in itself it is, all objects showed themselves in their true colours, resplendent with the sextuple glory of the light actually contained within them. To eyes so unsealed King Leary’s fiery locks appeared of the colour of sorrel green, His Majesty’s saffron kilt of the hue of brewed spinach, the royal golden breasttorc of the tint of curly cabbage, the verdant mantle of the monarch as of the green of laurel boughs, the commanding azure eyes of a thyme and parsley aspect, the enamelled gem of the ruler’s ring as a rich lentil, the violet contusions of the prince’s feature tinged uniformly as with an infusion of sennacassia.
And now you can see at a stroke how it was created. The passage was successively adorned, until you can scarcely see its plain beginnings from the final result.6
Returning to music, composers like Gustav Mahler wrote things which alienated the audience before recordings, but which exceptionally reward relistening. Doing so reveals the structure underlying the work, which may not be apparent on first listen. In his 9th symphony, the first movement is a series of developments on a theme first advanced from the very first notes; but its scale is so enormous that one at first does not recognize the unity of its conception. Only upon relistening did I realize that the furious, baleful howl of the horns in the 18th minute is the same as the gentle, irregular heartbeat which opens the piece. Everything is made to seem as though effortless.
Ability need not lead to good music. Indeed, great ability can be a handicap, for it threatens to replace the creation of music with athletic feats which are not appealing without knowing the work which went into creating them. The famous caprices of Paganini are facile for their emphasis in showing form; the desire to show that it could be done came before knowing what exactly the artist wanted to say. Likewise, the piano work of Liszt is ridden with artificial demonstrations of prowess. Take the Liebestraume no. 3: the primary melodic material, which is extraordinary, is interposed with interminable chromatic cadenzas of no artistic merit whatsoever. Ability must be a tool to expression, and can never be mistaken for an end in itself.
And there are, of course, cases of great work which came, not even in spite of, but because of, the inadequate skill of the artists. What they have instead is sincerity, for that is an essential component of any great artistic work. Elster cites jazz singers in particular – clarity of tone, so essential to classical singing, is often a detriment to expressivity. Godspeed You! Black Emperor do not play music which has or needs much ability, but it is capable of attaining emotionally heights which technically adept but facile bands cannot.
I think this is, on the whole, a good book. The best part of it is the classical history underlining the economics; the worst of it are the aesthetic judgements of the third essay. Since it was a short and easy read, I recommend reading it.
Indeed, Krugman argues that modern evolutionary theory is entirely isomorphic to what neoclassical economics is doing!
The experimental evidence points toward something like hyperbolic discounting. The classic experiment to show this is to ask someone if they would prefer $100 today, or $105 today, and then ask if they would prefer $100 in a year, and $105 in a year and a day. People’s answers should be the same, yet they are very often not.
For myself, I wonder how much of hyperbolic discounting is simply a reaction to complexity. Ryan Oprea has a recent paper in the AER on the classic paradoxes of risk aversion. When he replaces the lotteries with complicated, but ultimately deterministic, games, he finds the exact same paradoxes. People overweighting some possibilities is simply a reflection of an overburdened mind. The research is so new that I do not believe that anyone has done it, and there may even be no one pursuing it at this time.
It is these sorts of contingencies which make optimal contracting difficult, and which lead to firms existing. Firms could simply be nexuses of contracts, a familiar form but not anything but a particular set of market transactions. It is the fact that future states of the world are indescribable which pushes us to form firms, which we can define (as with Grossman and Hart) as addressing who owns the residual. There are many unspecifiable improvements possible, so changing ownership to the people most able to achieve those gains benefits the whole world.
I wonder what a classicist would think of Pinker’s argument that the world has grown less violent over time.
Woah, and then people say good zoomer blogs don't exist! LMAO. You're doing our generation a valuable service.