July 28, 2018 Dear Reader

Chapter 29

Sweet Talk is Virtuous

Economics as an academic field has with language is that it ignores the language used by economic actors. Adam Smith, whose first job was teaching English composition to high-school age students in Scotland, and who wrote a little unpublished book on rhetoric, would be distressed at how uninterested in language his offspring have become. Smith was indeed, with a few partial exceptions in Austrian and old institutional economics, the first and alas almost the only economist to acknowledge that language, and therefore the ethical reflection he called the “impartial spectator,” matters to how the economy works.

Understand, though, that bringing language into thinking about the economy does not necessarily trash everything that economists have already discovered. I do not wish to bring comfort to humanists who detest mathematics and statistics, for example, or people of the left who detest everything about what they call “neoclassical” economics, by which they often mean, to be particular, my beloved Chicago School.

In 2002, for example, Bernhard Guerrien launched an attack on those of us of Chicago and elsewhere who believe that supply and demand curves are pretty neat stuff (Guerrien, 2002).  Guerrien notes that supply and demand curves assume a "given" price. This is of course correct. But he then argues that there is no conceivable source for the givenness except the patently absurd fiction of a Walrasian auctioneer, the fellow whom Léon Walras is thought to have imagined in 1874 (he didn’t actually: the phrase is a later attribution) as the person who adjusts excess supplies and demands in a market by literally shouting out new prices. The criticism is incorrect, and can be seen to be incorrect as soon as we bring language seriously into our view of the economy. 

What's incorrect about the criticism, an old one, actually, is that there is a source, an obvious one, for “given” prices, though neglected by the Walrasian-Samuelsonian economics that Guerrien and I join in criticizing.  The obvious source is also ignored by Marxist economics, neo-institutional economics, post-Keynesian economics, behavioral economics, law and economics, much of the rest of economics, and even I believe by economic sociologists.  The only economists who so much as mention it are the Austrians and the old institutionalists I mentioned, which is one reason I count myself a fellow traveler of these disdained little groups.
The missing source is conversation, rhetoric, language, sweet talk itself.  The price gets its givenness from the literal conversations that go on in markets.  I do not mean by "conversations" only the putting and taking of offers, surrounded otherwise, as has been assumed in economic theory since Jeremy Bentham, by stony silence.  To be sure, mere money offers are, Smith had noted, a variety of persuasive talk:  "The offering of a shilling, which to us appears to have so plain and simple a meaning, is in reality offering an argument to persuade one to do so and so as it is for his interest" (Smith, 1762-3/1978/1982), p. 352).  But people do not merely silently offer shillings and silently hand over haircuts.  People do not in fact behave in the manner supposed in “autistic” economics (as rebellious French economics students have called it), the manner that Guerrien assumes---as vending machines.  They talk, or as Arjo Klamer puts it, they converse.  

And in conversing they open each other to modifications of the price, it may be, and anyway they establish, as we say, the "going" price. Every market participant, wrote Smith, "is practicing oratory. . . . In this manner . . . [he] acquires a certain dexterity and address in managing their affairs, or in other words in managing of men [and women, dear Adam, if you please]; and this is altogether the practice of every man in the most ordinary affairs." The ordinary affair of economics itself, for example. The going idea in Samuelsonian economics, we advocates of a humanistic science of economics say with indignation, has long been the crazy idea that people do not converse. The Samuelsonians are mistaken.

Of course, in a large market or a large conservation a small voice is seldom heard. That’s what an economist means by “givenness,” and was the point of Albert Hirschman’s pregnant distinction among the three tactics one can use when faced with an unhappy social situation: exit, voice, or loyalty (Hirschman, 1970). There is little point in driving to an enormous California supermarket and initiating an aggressive conversation with the manager about the price of milk.  You had better wait until you are talking to your friend the local shopkeeper, perhaps, who might actually respond, persuading you in the ensuing conversation that nothing is to be done, because after all he is in turn a small voice in the market for milk.  Or you might, as an economist, wait until you are talking to the Milk Board, which sets the wholesale price of milk, though doubtless it does so after much talk with fellow Board members and with politicians and with Ministry of Agriculture functionaries.  You might change their minds, and so their talk, and so the price, a bit.  You certainly could change some weak minds if you were, to take an impossibly extreme example, the President the United States, say, and wanted to redefine the word "torture."
The situation in markets is identical to that of language.  No prudent person would initiate conversations with strangers on the bus about the definition of "givenness" in economic theory.  If she does she can expect them to edge away from her. She will instead wait until she is talking to other economists, at any rate to economists imagining in their conversations a post-autistic economics that is not so dogmatically of the Left that it objects to every idea that the cursed bourgeois economists have articulated.  We use the French word amour or the English word love without stopping to quarrel about their meanings, or insisting that love actually means "hate," or "light bulb," or "the train will arrive in six minutes."  That is, the on-going conversation of language---I note that Walras' colleague Saussure made this point a century ago---gives to us mere ordinary speakers of a language a set of distinctions serving to define what's on offer in French or English by way of sheep/mutton as against mouton. In the same way, Friedrich Hayek pointed out long ago, prices are “given” only in the sense that the search procedure of a market conversation coughs them up, out of bids and offers formed from “knowledge initially dispersed among all the people” (Hayek, 1945, p. 520). Hayek didn’t emphasize quite as much as he should have the conversation that does the trick.
Guerrien will perhaps reply that the going price, or meaning, is just an instance of bargaining.  But I think he and I agree that if bargaining in a strictly game-theoretic way is what we are talking about, then we should abandon all hope for a useable economics.  The Folk Theorem showed some time ago that in a properly infinite game and an assumption of prudence only you can get an infinite number of equilibria.  As some game theorists put it, “game equilibrium models of rational play lead to an outcome set where players can do almost anything and still be consistent with the theory. The prediction that individuals might do anything from a large set of feasible strategies is neither useful nor precise” (Ostrom et al., 1994, p. 322). Prudence-only game theory, without social agreements of solidarity and justice on how a conversation can change minds, has no implications. None at all: change the assumptions, change the equilibrium. And in every empirical test on offer, this or that set of prudence-only assumptions has failed. Unlike supply and demand curves.
A price is not set usually by silent bid-offer, move-countermove game bargaining, with its intrinsic paradoxes, an elderly example of which Guerrien has repeated.  Price is not set by an auctioneer in most markets---though I wonder what Guerrien would make of the Alsmeer flower market in Holland, with its Dutch-auction clocks ticking the price down; or of the remaining open-outcry pits at the former Chicago Board of Trade. Most prices get their meaning and in particular the givenness of their meaning from the economic conversation.  Just as amour has a more or less given meaning in French, or torture in English, modifiable at the edges by particularly persuasive talkers, or, to speak of the main source of actual linguistic change, by the games that teenagers play with words, so do dictionary-makers face a more or less given money price for their product. Larrousse cannot suddenly decide to charge 10,000 euros a copy for its big French-English dictionary, or even much above the going price. And therefore it lives with supply and demand curves. There is nothing mysterious or self-contradictory about the situation. 

I do not claim that we economists have already figured out how language and the economy intermesh. In fact, my point here is the opposite. The scientific task still remains to be done, yielding a humanistic economics, that is, an economics acknowledging humans as talking, singing, story-telling, ethical creatures. Until then the science of economics will be incomplete and paradoxical in the ways that Guerrien has noted.

Meanwhile, givenness is how we little folk in a large society face any piece of our culture, such as language itself or the going price of milk.  We only need to recognize that the economy is part of the culture, and of its conversations, to recognize that supply and demand curves do after all work, rather well.

* * * *

Let’s then get down into the details of a humanistic economics that recognizes talking people. Realize at the outset that the role of language in the economy---as is so obviously true of language in economics---is not merely the transmittal of preformed messages. Language persuades, and when the language speaks us we are changed.

The economist Jacob Marschak wrote a paper in 1968 exploring the “economics of inquiring, communicating, deciding,” but got no further in thinking about it than the model of transmittal and gathering of bits of information, such as telephone numbers and the price of hog bellies. “Data are gathered. They are communicated [note the word] to the decision-maker. He, on the basis of the message received, decides upon the action” (Marschak, 1968, p. 2). The only fruit of the sender-receiver metaphor thus deployed is to observe that “perfection is costly” (p. 3), in other words that encoding and decoding are costly (see his diagram on p. 5) and are therefore subject to a rational calculus of cost and benefit. At some point---the statisticians call it the “optimal stopping point”---it’s not worth inquiring any more into the price of 10-year old Toyota Avalons in good condition in Chicago. You probably aren’t going to find a better price, and anyway the additional search won’t yield enough of a better one to overmatch the extra search costs.

The same point had been made in 1961 by another economist, George Stigler, who developed the mathematics more elegantly than Marschak did, and stated the basic point more eloquently, too: “One should hardly have to tell academicians that information is a valuable resource: knowledge is power. And yet it occupies a slum dwelling in the town of economics. Mostly it is ignored” (Stigler, 1961/1968, p. 171).

Thanks to economists such as these---Kenneth Boulding, for example (1958, pp. 87-97); or Ronald Coase and his “transaction costs”; or George Akerlof and “asymmetric information”---the transmittal and gathering of bits of information is no longer ignored in economics. In fact it could be said to be one of the two main preoccupations of economic theory since the 1960s, the other being further explorations to and beyond the outer limit of reason of “rational” behavior assuming one has already acquired all profitable information, “non-cooperative game theory.”

But the metaphor of transmittal has narrow limits. When Thoreau was told by some technology-admirer that the extensions of the new telegraph now made it possible for “Maine to speak to Texas,” he replied, “But does Maine have anything to say to Texas?” The meaning of the message in economics has been left aside. Do we have anything to say?

Students of communication, rhetoric, linguistics, philosophy, and the like call “speaking to Texas” the “conduit metaphor.” In 1979 the linguist Michael Reddy gave fully 141 expressions of the conduit metaphor in English rhetorical practice, such as “You’ll have to try to get your real attitudes across to her better” (expression-type A, “implying that human language functions like a conduit,” number 1, p. 311) or “Her unhappy feelings fell on deaf ears” (type G, “implying that the [bits of information] may or may not find their way into the heads of living humans,” number 141, p. 320). The point is, as Reddy puts it, that “English has a preferred framework for conceptualizing communication, and can bias thought processes toward this framework, even though nothing more than common sense is necessary to devise a different, more accurate framework” (p. 285). He gives 45 expressions that when used alone, without the conduit metaphor lurking in the background, imply a quite different framework: for example, “How do you build readings for sentences like that?” Building readings is a cooperative enterprise in a common space, a cooperative game, not a pre-formed bit of information sent hurtling through the conduit, like signed checks whisking in the pneumatic tube to your drive-in bank teller. My personal metaphor, less felicitous than “the conduit” because it depends on a specifically American cultural reference, is the “Roto-Rooter theory of communication.” Communication in the Stigler-Marshak-et al. view is a matter of pipes between minds. If they get clogged , the sending of information costs more, and so it’s then worthwhile to call the Roto-Rooter man and have the pipes reamed out, re-establishing clean conduits in which “Your concepts come across beautifully” (Reddy, number 8).

“But wait a minute,” you might say. “What’s wrong with the conduit metaphor? Isn’t it true? Aren’t we in fact engaged in getting our ideas across and gathering information?”

No, we are not, not entirely. The conduit metaphor describes some part of economic language, such as the report that hog bellies were $0.9283 per pound at 2:30 Eastern time on June 28, 2007. Obviously some language is transmission as through a conduit, as when you give your friend your telephone number or when you inform an economically naïve audience that Ben Bernanke’s portfolio as Fed chief is after all only a few percent of existing world bonds. Humans do such “communication,” but so do birds signaling territorial limits, or fish in a school. Trademarks and brands are of course linguistic items, "signs" literally. Informational advertising would provide some of the data for a study of social language in this restricted sense, making use of the immense academic literature on marketing, for example. Semiology arises at that point---again in animals and plants as well. Flowering plants are signalers, and co-evolved with insects and birds and mammals: "Here I am, oh pollinating insect, oh seed-spreading bird or mammal."

And beyond the point of honest persuasion in “conveying” information comes the temptation to lie, “misinformation,” “manipulation,” which is the vulgar, newspaper meaning of “rhetoric.” This we humans have to a notable extent in common with other great apes---but also indeed with camouflaging plants and animals. Though advertising is, financially speaking, mainly informative or a bond as to quality, notoriously it is sometimes persuasive in dishonest ways. Precisely because signs and signals and advertising and rhetoric are sometimes not mere information, and meant instead to change ones mind, humans (and other great apes) are suspicious of rhetoric. It’s one reason that since the 17^th century the study of rhetoric has been devalued, as merely democratic beside the aristocratic glories of first-order predicate logic.

* * * *

But persuasion, as against informational messages sent through the conduit without rhetoric or judgment, is in fact a very large item in a modern economy. We economists have got to stop ignoring the fact, if it’s a fact. We’ve got to stop ignoring meaning and its making through words.

* * * *

In explaining the fact of 25 percent of national income being sweet talk a temptation of the modern economist is to try to model it in the style of Samuelson, as the outcome of still another adventure of the prudent person, Max U. The modern economist does so because it’s her only model. If something---love or justice or courage---does not fall within a utilitarian maximization subject to a resource constraint, she has nothing to say. But language, I am suggesting, unless reduced to bits of information, as it cannot be, cannot be so modeled.

The limits and patterns of human speech do of course limit and give pattern to the economy. Some conversations are impossible in humans. At the most abstract level, some sort of Chomskyan limits of deep structure might possibly apply, though it seems doubtful. Perhaps there are deals, orders, desires, plans that would be possible in a language of another species but are interestingly impossible, or at any rate difficult, in human language. Beings that were not differentiated individually, for example, would find orders naturally persuasive in a way that humans do not. Wittgenstein said that "to imagine a language is to imagine a form of life" (Wittgenstein, 1963, p. 19). He might as well have said that to imagine a form of life is to imagine a language. "It is easy," he remarked, "to imagine a language consisting only of orders and reports in battle" (Wittgenstein, 1963, p. 19). An army that is something other than a gang of Homeric heroes clashing one-on-one in single combat is a form of life that responds to particular orders issued by particular people. The phalanx on the left flank moves when the general speaks, as though it were an organism and not a collection of free citizens of Athens.

But the binding constraints are much more likely, it seems to me, to be matters of pragmatics and socio-linguistics than matters of syntax and vocabulary. I have a friend, a Dutch woman, who built a vacation home on a Greek island. She found that within Greek society it was impossible, simple impossible, for a woman to tell a male contractor what to do. Her contractor ignored her requests, and she was forced to hire another Greek, a man, to give the orders. Even that did not work perfectly: an order to have large waste pipes for the toilets was ignored, with the result that---as is common practice in rural Greece---her soiled toilet paper is not flushable. There's an economic effect.

The formal attempts to extract any interesting constraints that language places on economic behavior from sheer logic or even from an enriched logic of the rules tha linguists call conversational implicatures has not borne fruit, and seems unlikely to. The attempt of the game theorist Ariel Rubinstein to do so shows how little can be expected even from very canny ruminations on evolutionarily stable strategies or a supposition that the equation ψ = [φ(x,y)∩φ(y,z)∩T]→φ(x,z) is a tautology (Rubinstein, 2000).

But economics still has something to say. To Texas. The economist and rabbi Israel Kirzner put his finger on what a free society achieves, from which we can understand how meaningful language works in one. "It [is] highly desirable to choose among alternative social arrangements those modes of organization that minimize [ignorance of knowledge that can be absorbed without decision and search, by the sheer noticing of it]. . . that is, those modes of organization that generate the greatest volume of spontaneous, undeliberate learning" (Kirzner, 1979, p. 147, 145). His assertion runs against the love of explicitness in modern life, the proliferation of handbooks on leadership and of axiomatizations of thinking. Surely, the handbook-writer avers, we need to transmit through a conduit to the student's mind numerous bits of information, and if this can be centrally planned, all the better. Every schoolchild in France is on the same page at the same hour of the same day, thanks to the planners in Paris.

But real innovation, Kirzner is saying, entails real ignorance, that us, "knowledge about which nothing is known" (1979, p. 144). It can be put economically: known knowledge (shades of Donald Rumsfeld) earns its normal reward. If you know how to read a balance sheet you do not on that account alone become Warren Buffett, because so many other people know how to read a balance sheet. Unknown knowledge, on the other hand, generates supernormal profits. When sometime before 1211 an anonymous Florentine invented the idea of a double-entry balance sheet, then he, or his Italian imitators, could pick up the profit from the innovation, and did (Origo, 1957/1986, p. 109). Once the reading of balance sheets is widely known, however, the supernormal profits fell to zero.

It's still a good idea for people to learn to read balance sheets, engaging in "search" that has a known reward to the MBA graduate or law student who engages in it. The opportunity cost of such searching may be good for the society, as against a worthless search for, say, learning to read the stars astrologically. But it is not an innovation. National income does not actually fall, since learning to read balance sheets has a marginal product equal to its opportunity cost, at the margin, and therefore has intramarginal gains (“rents” economists call them, if not the “supernormal” profit of real entrepreneurship), whereas learning to read the stars does not. The intramarginal reward to routine learning sustains the national income. As a matter of fact, as an economist can persuade you in one of her maddening diagrams, it simply is the national income. But national income will not rise unless the innovation is Kirznerian.

"The ease of calculation provided by money," writes Kirzner, "is thus not merely a device for lowering transaction costs relevant to deliberate search," as the Samuelsonians claim (Kirzner, 1979, p. 150). "It represents a social arrangement with the ability to present existing overlooked opportunities in a form most easily recognized and noticed by spontaneous learners." Kirzner makes a parallel point in his writings on entrepreneurship.

Kirzner's analysis is correct so far as it goes. What is missing from it, however, is language. The alertness that Kirzner thinks of as the essence of entrepreneurship involves language in its fulfillment. Unfulfilled it’s just another bright idea. The necessary, next entrepreneurial step---which Kirzner does not treat---of persuading oneself, a banker, a supplier, an employee, a customer, oneself is rhetoric all the way down. In consequence a community of free speech briefly unique to Northwestern Europe after 1700 or so, for example, "represents a social arrangement with the ability to present existing overlooked opportunities in a form most easily recognized and noticed by spontaneous learners" (Kirzner, p. ).

The crucial point was discovered in 2007 by Sarah Millermaier, who argues in the way of Jürgen Habermas that communication is after all a cooperative game (Millermaier, 2007). A real conversation, in Habermas' words, "specifies which validity claim a speaker is raising with his utterance, how he is raising it, and for what" (1981 [1984/1987], p. 278). That is, a real conversation entails serious rhetoric. What Habermas calls "strategic" speech is on the contrary a reading through the speech to the "underlying" interests. It is speech meant to achieve a result external to the practice (to use, as Millermaier does, the language of still another student of these matters, Alasdair MacIntyre). Millermaier observes---and here with MacIntyre and myself---that the conversation must be ethical and the ethics must be of the virtues and therefore that what I am calling "real conversation" must draw on the seven principle virtues (McCloskey 2007).

Think of an academic discussion, perhaps one on how the way that language works in an economy adumbrates a humanistic science of economics going quite far beyond the prudence-only, Benthamite-Samuelsonian-Posnerian routine that economists have been grinding on for so long. Imagine that the main speaker is not trying earnestly to uncover the truth, say, or to learn from the audience by listening, really listening, but instead is focused entirely on some result external to the practice of serious scientific inquiry---getting a job offer that will raise her salary at home, perhaps; or demonstrating to the admiring audience how very intelligent she is. Imagine that the audience is similarly engaged in a non-cooperative game (the old Industrial Organization seminar held at the Law School of University of Chicago in the 1970s was like this when certain members were present, and others absent). Such a boys’ game may be fun to play. But it’s not serious conversation, not science. Except in those cases in which the science is run by boys.

If speech is merely strategic, a non-cooperative game, then the only virtue in play is prudence. Every attempt to characterize speech by a well-trained Samuelsonian economist is going to try to reduce it to such prudent tactics. Economics is after all the pure theory of prudence. It is natural to the rhetoric of economics since Bentham and especially since Samuelson to imagine that all behavior is reducible to that of the charmless, unloving, and above all calculating fellow, Max U.

Millermaier’s point is that such a reduction is corrupting of real conversation. It makes impossible the mutual formation of meaning which much of our economic life is about, and depends on. We engage in polite chatting around the water cooler and are able thereby to cooperate with our colleagues. If we engage in it obviously for that purpose, though, people catch on, and we find it more difficult to gain cooperation. An economistic way to make the point is a paper by Paul Ingram and Peter Roberts in the American Journal of Sociology in September, 2000, "Friendship Among Competitors in the Sydney Hotel Industry." They find that the friendships among competing hotel managers in the 40 Sydney hotels in their study generate about $2.25 million Australian more of gross revenues per year per hotel—for example, through recommendations of the competing hotel when fully booked—than would be generated by a hotel with friendless managers (p. 417). So far so good for Judge Posner. They add, however, "the critical caveat that the instrumental benefits of friendships are inextricably tied to the affective element," that is, you can't successfully fake friendship (p. 420; compare Mueller 1999, p. 39). The faithless ones get found out. Considering the depth of skill among primates in performing and detecting falsehood, this is not surprising. Both Prudence and Solidarity work. "Individuals who try to form and maintain friendships solely as a means to material gain will fail to evoke trust and reciprocity." That is, prudence only will not work, and so "those who would limit the intrusion of society into economy by . . . characterizing embedded relationships between buyers and suppliers as predictable outcomes of a repeated, non-cooperative game" are mistaken (Ingram and Roberts 2000, p. 418).

That’s another reason that prices and meanings cannot be sheer, non-cooperative games. It would be like insisting that married people only deal with each other instrumentally. As Millermaier observes, for another example, programs of corporate ethics that declare themselves as “using” values to achieve Max U’s goals will undermine the cooperative game that makes language and ethics possible.

The conundrum of language in the economy, then, cannot be solved within Max U models. To the extent that language is reduced to Max U it ceases to exhibit one defining characteristic of human language, which is, I hope you believe by now, not the mere transmission of information but the making of meaning and the imagining of novelties.

The mind, that ocean where each kind

Does straight its own resemblance find;

Yet it creates, transcending these,

Far other worlds and other seas,

Annihilating all that’s made

To a green thought in a green shade.

To put it another way, the Max U model fits smoothly with the conduit metaphor. But Max U does not fit at all with a rhetorical (or Wittgensteinian or Burkean or Austinian or Habermasian or MacIntyrish) theory of language. If these were just silly theories, amusing to the effete snobs in the Department of English but unworthy of the tough, masculine science of economics, and econowannabe sciences like political science or law and economics, then economics could go on ignoring them. But they are in fact the best thinking about what language is that the 20^th century offered. It would be unscientific to go on insisting that all we economists can talk about is our old, if unreliable, friend, the implacably silent Max U.

* * * *

Another example. Trust, of course, is one outcome of sweet talk. Economists have devoted a good deal of attention to trust over the past decades though they have usual tried to make it fit the Procrustean bed of a no-language, single-virtue, prudence-only view. Even if an economic actor does not herself talk, the economist gamely claims, she will nonetheless form opinions about the trustworthiness of others, looking merely as a prudent person at their actions. We say, “Actions speak louder than words.”

But the proverb is false if taken always to be true, as economists now do. Sometimes words speak louder than actions. Borrowing and lending take place only after a persuasive story has been told and believed. You do not lend to your brother-in-law if you do not believe his promises to repay, or if you do not share his vision of the millions to be earned in real estate. The more so with strangers. In the absence of signals of trustworthiness such as belonging to your own unpopular religion, say, or having a high rating from Dun and Bradstreet, no stranger gets a loan.

A classic paper in 1963 by the legal sociologist Stewart Macaulay studied firms that did business in Wisconsin. He confirmed what everyone in business knows, that business normally depends on a state of trust, not on explicit contracts to be enforced in courts. One large manufacturer of cardboard boxes looked into how many of its orders had no agreement on exact terms and conditions that would satisfy a lawyer looking for a "contract." The manufacturer found that in the mid 1950s the percentage ranged from 60 to 75 percent of the orders, in an industry in which an order canceled means you end up holding a lot of useless boxes shaped and printed to the particular customer's specifications (Macaulay, 1963, p. 196).

It drove the company lawyers crazy. One said, "Often businessmen do not feel they have `a contract'—rather they have ‘an order.’ They speak of ‘canceling the order’ rather than ‘breaching the contract’" (Macaulay, 1963, p. 197). Another lawyer declared that he was "sick of being told, ‘We can trust old Max,’ when the problem is not one of honesty but one of reaching an agreement that both sides understand"(Macaulay, 1963, p. 195).

The non-lawyer businessmen didn't see it that way. "You get the other man on the telephone and deal with the problem. You don't read legalistic contract clauses at each other if you ever want to do business again. One doesn't run to lawyers if he wants to stay in business because one must behave decently" (Macaulay, 1963, p. 198). One uses the courts only when someone defects. But few defect. There's a purely prudent reason, to be sure—that defecting is bad for business. But there's a just, faithful, loving ("good old Max") reason, too.

People want to be virtuous in business as elsewhere in their lives, and their virtues depend on relations developed through talking with good old Max. (Talking to Max U, by contrast, is valueless cheap talk.) Macaulay concluded that "Two norms are widely accepted. (1) Commitments are to be honored in almost all situations; one does not welsh on a deal. (2) One ought to produce a good product and stand behind it" (Macaulay, 1963, p. 199).

In 1912 before a House committee on the money trust J. P. Morgan was being questioned by a hostile Samuel Untermyer:

Untermyer: Is not commercial credit based primarily upon money or property?

Morgan: No sir; the first thing is character.

Untermyer: Before money or property?

Morgan: Before money or property or anything else. Money cannot buy it. . . because a man I do not trust could not get money from me on all the bonds in Christendom.

Of course. If you want to be frightfully sophisticated about people's real motives and claim that these are not the rules of bourgeois life, that capitalists are a pack of liars and thieves, you will need to explain why you get indignant when the rules are violated, and why in your daily transactions you assume they will be obeyed. Breaches of contract and negligent torts are worth worrying about, but they are not the central business of business, which is talking people into cooperation to get the job done.

Trust, of course, runs stock markets. The rumor of the Street determines the price of stock, at any rate within limits of fundamentals---though the very fundamentals, such as imagined futures sales and returns, or the fairy dust of the Federal Reserve Board, are themselves often sweet talk. Trust and friendship, therefore, make possible speculative bubbles, from the tulip mania of the 1630s to the dot-com boom of the 1990s. The very fact of capitalism's speculative instability, in other words, argues for an entirely new prevalence of belief in strangers. "Credit" is from creditus, "believed." Each of the hundred-odd quotations in the Oxford English Dictionary illustrating the noun and the verb date from after 1541, and most of the commercial quotations from the 16^th century are suspicious of it. An act of 34-35 Henry VIII (that is, 1542) noted that “sundry persons consume the substance obtained by credit of other men.” Shame on them. Contrast the neutral language of Locke in 1691: credit is merely “the expectation of money within some limited time.” A shift in talk had taken place, 1542-1691, and a shift in the ideological support for capitalism (McCloskey, forthcoming).

A business cycle based on pyramids of credit was therefore impossible in the distrustful 16^th century and before. The macro-economy could in earlier times rise and fall, of course, but from harvest booms and busts, wars and pograms, not from credit booms and busts. In those olden days God's hand, not human beliefs, made for aggregate ups and downs. Medieval and early modern people trusted only allies, and had wise doubts even concerning some of them: “How smooth and even they do bear themselves!/ As if allegiance in their bosoms sat,/ Crowned with faith and constant loyalty” (Henry V, II, ii, lines 3ff). Pre-moderns had to keep faith with God and with their lords temporal. Late moderns keep faith with the market and with their friends, and build upon credit.

On this theory the episodes of disorder and unemployment in capitalism from the 1630s in Holland, and from 1720 in Northern Europe generally, arose from the virtues of capitalism, not from its vices, from its trustworthiness, not from its greed. To be more exact: the business cycle arose from trustworthiness breaking down suddenly in an environment of quite normal human greed for abnormal gain, the accursed love for gold which has characterized human beings since the Fall. What is novel in capitalism is the faithful trust, generated by talk. Why do people ride on airplanes at great expense to have sit-downs with bankers or customers or Mafia godfathers? To generate the trust to do business with.

* * * *

Another example. "A gap persists," writes the game theorist Judith Mehta, "between accounts of behavior framed by rational choice theory and experimental evidence of how people actually behave in a bargaining situation" (1993, p. 85). The way people frame the story they are in, the meaning of the story as much as the plain facts, changes the outcome. A "bargaining situation" would be, for example, that of two people who meet in the middle of the Sahara, one with food only, the other with water only. At what ratios will they trade? The solutions are obviously multiple. If Ms. Jones, with her water jug, is a canny bargainer, while Mr. Brown, the food guy, is a bit of a simpleton, you get one solution, favorable to Jones. If the skills at talking and thinking are reversed, you get another. If Jones and Brown are Rawlsian, they share the water and the food exactly equally, because behind a veil of ignorance as to what position they are in, that is the best solution, supposing (as Rawls somewhat arbitrarily did) they are both risk averse. . . and if one is not much taller than the other; if one does not have stored fat and the other does; if they can reach reflective equilibrium; if they have somehow become before the story begins ethical actors; if they are Moslems; if they are Buddhists; if, if, if. The two may wish to be fair. But, Mehta observes, the word "'fair' does not have some singular, objective meaning" (p. 94). We make the meaning, through talk.

As Mehta puts it, "real individuals ascribe particular and shifting identities to themselves and their opponents in a bargaining situation (for example, as ‘friends’ or ‘non-friends,’ or as partners in a relationship). . . . They adopt a particular set of expectations and behavioral responses contingent on the meaning they ascribe" (p. 93). These are roles we assign in stories we tell. An example is the role---literally, the theoretical, fairy-tale role---played by the chairman of the Federal Reserve Board in "setting" the interest rate. There is no evidence that Ben Bernanke actually "sets" the world interest rate. The notion that he does so is impossible, though a wildly popular tale, considering the bit of information that the portfolio he has influence over is a tiny portion of the loanable funds available in the world's capital markets. But the exchanges dance, at least in the short run, to Ben’s tune, because of the heroic story that people tell about the meetings of the Fed open market committee.

* * * *

One can go on like this. The best tactic, though, is to go back to the rhetoric of economics itself, looking hard at the metaphors economists use and noting the language games supposed in them. Take consumption. Imagining having ice cream to consume entails a language of thought (a point that the philosopher Jerry Fodor makes); that is, choice entails a language, an internal debate (as Aristotle pointed out), even a rhetoric. This is true of animals, and certainly of Crusoe on his island alone. He said so. The human ability to imagine is merely a more extreme form of choice-making common to all living things. But it is extreme. A bird imagines (we suppose) a nest of a species-particular kind. But only language-using humans, so far as we can gather from watching whales and elephants and chimps a little, can imagine a thousand different forms of Crusoe's cave and compound, those other worlds and other seas.

The very choice of technique, the tough-minded consideration of “production functions,” involves language. How things are made with recipes will entail talk. "Take two eggs. . . ." The books of recipes that Paul Samuelson characterized as "the production function" are informational, and this is as far as economics gets in thinking about them. It’s the conduit metaphor again communication as bits of information. But books of recipes are persuasive, too. "Hmm: that looks good." A chemical engineer must be persuaded that this or that technique he has heard of does in fact work as advertised, and indeed trade magazines directed at him are quite elaborate exercises in persuasions-for-a-sophisticated-audience, sweet talk of a professional sort. Look for example at the Journal of the American Medical Association, its articles, its editorials, its ads.

Likewise working. The metaphor of a production function must always involve getting people to do the job---take the eggs, adopt the technique, arrive on time. Organizing other humans obviously involves language, about which popular books about management obsess. About a third of the business books on the airport rack are devoted to rhetoric. Count it the next time you’re in an airport. Management, I noted, is the great realm of sweet talk in a free society---not so in a society of status and instant obedience backed by swords. And not so in a society of utter routine, the steady state, Groundhog Day, even if the workers are otherwise free. If every question of what to do is already answered, then people just do it, without the persuasive, "John, we want you to go on the road next month to consult with our big client in Milwaukee. It's an important assignment."

The management need not be suited and tied. Any human cooperation in a task requires language, at any rate for a new, non-routine task. A master electrician needs language to tell his journeyman to push the wire further behind the wall. The catcher signals the next pitch, the pitcher shakes off the sign, the catcher tries another. They are speaking, and persuading.

Social clotting can stop persuasion, well short of any purely syntactic inability to express this or that. If a worker resents every order, unwilling to imagine himself a member of a sacred Team or unwilling to try to see the economic point of the order (perhaps because he does not give a hoot whether the customer is well served, or whether the Team wins), then cooperation breaks down. A society in which low status people---women, blacks, untouchables, young people---are not listened to by high status people will forego a gain from cooperation. The point is similar to one defense of classic democratic theory, namely, that the more opinions that are expressed the better can the society choose the best ones. A mechanical form of the argument appears in the voting schemes of Condorcet. A less mechanical version is the advocacy of free speech by John Stuart Mill, and Lord Bryce's phrase, adopted enthusiastically by the economist Frank Knight, "government by discussion." The economist Scott Page has brilliantly discussed diversity in such terms (Page 2007). The question is how big the loss is from government by politburo as against government by discussion. The experiments in the 20^th century with communism suggests that the loss is big, in treasure and in spirit.

* * * *

One more consequence of a linguistic economics would be the admission of ethics. Prudence is a virtue, I have noted, but it is not the only virtue. One can make a case that the virtues are seven: prudence, temperance, justice, courage, love, hope, and faith (McCloskey, 2006). George Akerlof has recently argued that macroeconomics depends on norms and motivations more complex than buy low/sell high---that an economics that thinks of people as normless and prudence-only will get the course of prices, incomes, unemployment wrong. He concludes that a more scientific economics than the one we now have “would observe decision makers as closely as possible, with the express intent of characterizing their motivation, and would use such characterization as the basis for modeling economic structures” (Akerlof, 2007, p. 56).

An economics that does not examine meanings expressed in words will have a hard time doing Akerlof’s “observing” and “characterizing.” Ostrom et al. note that their experiments do not distinguish exactly which ethical imperative is added to prudence only, only that some additional norm must be working as well, since their subjects do not act as a prudence-only Max U-er would. "When an [anonymous] defector is called a 'scumbucket,' is the reproach being used because someone is breaking a promise, is being uncooperative, or is taking advantage of others who are keeping a promise?" That is, is the scumbucket seen as bad because he is unfaithful to an identity of promise-keeping, or because he is unloving, or because he is unjust?

But whatever exactly the sin, the experiments by Ostrom et al. and dozens of other recent experimenters, quite aside from the testimony of law and literature since the invention of writing, show that people don't like to be sinful and selfish. To put their point in Ostrom et al.’s own, game theoretic terms, ethical considerations beyond prudence only in effect "trim the branches of the decision tree" that violate justice or love, say. The experimental findings, and the evidence "from the field," overwhelmingly attest to such trimming.

Yes, I can get away with stealing a book from Sandmeyer's Bookstore next to my building in Chicago. After all, the owner trusts me, and often turns his back. But I wouldn't think of it. Stealing from him would outrage my sense of myself, my impartial spectator. I do not always "trim" even the monetarially unprofitable branches. If Sandmeyer's charges a dollar more for a book than does Barnes and Noble I nonetheless gladly buy it at Sandmeyer's, to support an independent bookstore against a goliath, to honor my commitment to a civilized neighborhood, to express my love for the owners. People do. It's cool to be cynical about human motivations, remarking sagely that one should follow the money. Take it easy, we say in Chicago, but take it. Cool though such analyses may be, they are nothing like a full account of human behavior.

* * * *

My love for economics is not I hope in question. I've practiced since 1964 its core disciplines, "price theory" Chicago-style (if not always with Chicago politics) and its characteristic fascination with numbers (if seldom in the bankrupt style of statistical "significance"). I've been a transportation economist and a quantitative economic historian. But I want to go a lot further, for the good of the order. Adam Smith was a professor of moral philosophy, not a student of prudence only. Economics since him, in Marxism and in Samuelsonianism, has become the exclusive study of Max U, that unattractive character---indeed, literally an inhuman, sociopathic fellow, who never stoops to exercise voice. He is a character whom Jeremy Bentham invented, Karl Marx politicized, Paul Samuelson mathematized, Gary Becker applied to everything, and Richard Posner adduces to get that nonsense-upon-still “justice” out of jurisprudence. I want economics to stop obsessing about Max U and become "humanistic" in a particular, academic sense. I want it to take seriously, as old Adam did, "the humanities," and become what the French revealingly call a “science humaine."

What this would mean is that we economists would deal with fully human characters---real ones like Madame Bovary and Jesus of Nazareth. It would mean that we would acknowledge in our models (realizing at last that a "model" is not always just the same thing as "a Max U formulation") all the human virtues: prudence, yes, but also justice, temperance, courage, faith, hope, and love. And the corresponding vices, alas.

A good first step is to take seriously the economics of language, right? Am I persuasive?

Let’s have a conversation about it, a real one.