«Chapter 14: STYLE AND ETHNICITY: THE EVOLUTION OF SYMBOLIC TRAITS “. what are the advantages which we propose by that great purpose of human life ...»
arbitrary signal of what sort of person is best to imitate will arise, and each of us will be motivated to go along merely because everyone else thinks it is the right indicator. Like peacock feathers, the result is beauty without function.
To see how the runaway process works in a simplified case, let's use another example of language evolution from Labov (1972). In New York City, middle class speakers have historically not pronounced the r in many words like fourth and floor. However, middle class speakers are very sensitive to the status gradations of language, and prefer to speak what they regard as the higher class dialect. It is a fact that higher class speakers in New York tend to use more r than middle and lower class New Yorkers. r use is a variable that people are more aware of than the dialect variants on Martha's Vineyard we discussed above. By having a preference for upper class dialect, and using r use as one of the indicators to select language models, middle class New Yorkers are tending to use more and more r. This should all have come to a halt when middle class and upper class speech contained the same amount of r. It hasn't because of a curious mechanism Labov calls “hypercorrection.” Although Middle class speakers have enough awareness of r to use it as an indicator, they do not actually have very good control over how much r to use. It is interesting how Evolution of Symbolic Traits 14-253 imperfectly we are aware of our speech, even in the case of a fairly conspicuous and publicly stigmatized bit of behavior like dropping r. The most “advanced” middle class speakers tend to use even more r than upper class people, the phenomenon of hypercorrection.
Presumably, most middle class people don't actually have upper class language models, merely slightly higher status middle class people. No matter how much r middle class people use, they “hear” the higher status people, whose dialect they prefer, using more. Everyone up and down the status hierarchy uses more r in the next generation, and so more r is genuinely spoken. The class-based correlation between preference for r and amount used remains. The next generation repeats the whole indirect bias/hypercorrection process, and r use increases again.
In these kinds of situations, even the upper class tends to follow along, pushed by the middle class, so the system doesn't simply overshoot a little bit and then stop or come back down to upper class norms. (Middle class perceptions acting as preferences are more important here than what the upper class actually does; there are too few of them to matter much.) There is every indication that this system will run off to complete use of r. Perhaps New Yorkers will start to roll their r like Spanish speakers! Interestingly, in England, the middle class perceives r-less speech to be higher status, and English speech is running in the opposite direction to that in New York.
The linguistic example is simpler than most presumed cases of the run away process because there are no strong adaptive forces acting to restrain the purely symbolic dialect indicator, and because the psychology of hypercorrection “artificially” maintains both a correlation between preference and indicator and creates a mean preference that remains higher than indicator until the dialect change goes to completion.
Even in language, where the range of selectively neutral variation is so enormous, the run away process presses the envelope. Some languages favor rather difficult tongue maneuvers to speak some words. For example the Amerindian language family Salishan is characterized by very complex strings of consonants in words. It is conceivable that Salishian languages are complexified to the point of real difficulty by exaggeration for stylistic effect. English has two painfully exaggerated features. One is its huge vocabulary. It is said that the retention of many French derived synonyms for Old English words was originally a result of lawyers retaining more synonyms to generate ever more complex legalese. The second is spelling. English is one of the few languages without regularized spelling rules.
We seem to resist regular spelling because mastering spelling is considered a mark of intelligence. Objectively, a spelling bee would seem to be a form of child abuse, yet the national winners of such irrational contests are juvenile celebrities!
14-254 Evolution of Symbolic Traits Pure run away models are controversial. The controversy among theorists (see Pomiankowski, 1988) is over whether when more realism is added, cases like the Ponapaens prestige system based on giant yams can respond the way r when the added realism of selection against the wasteful growing of costly items for display is introduced, and when there is no hypercorrection effect to produce a displacement and maintain a correlation.
When an element of natural selection or direct bias is added, the correlation between the preference and the indicator tends to collapse in some models.
The handicap hypothesis proposes that the exaggeration effect is rules by real adaptive advantages. Zahavi (1977) has argued that males can signal their overall fitness by managing a serious handicap successfully. The gaudy tail feathers of a male peacock are essentially saying to females “look, my genes must be good, or I wouldn't be able to obtain food and avoid predators carrying around all these lovely (OK, useless too) feathers. If you mate with me rather than drab old Joe over there, your offspring will benefit from my excellent genes.” The complex social life of humans is likely to involve signaling all sorts of things, and it is easy to imagine that Zahavi's hypothesis extends to cultural traits.
If the ongoing theoretical debate settles in favor of the run away process, then we would interpret Ponapaen giant yams (and American giant cars) as a situation that started out as an adaptive case of indirect bias that got away. Those who grew larger-than-average yams (or drove slightly better cars) were perhaps once better farmers (businessmen). However, once the size of yams had been elevated to the status of an important indicator trait, the run away process might take over. Soon people began to grow special giant yams just to attract the attention of people who were trying to use yam size as a way to learn to be better farmers. Then the role model choosers stopped caring if large yams were really an index of farming skill or not, because once “everyone knows” that yam size is an index of people's general skill and wisdom, and that prize yams earn the respect of authorities, the actual correlation of the skill of growing yams with any non-socially defined skill or quality is beside the point. The slave- of-fashion mechanism is off and running!
If the handicap hypothesis wins, we will have to give an honest signal interpretation to the same cases of exaggeration. At the beginning of the evolutionary process, it may be easy to fake a handicap, say by having some colorful feathers but hiding them except when displaying to females. In such a case, there is no serious handicap due to the anti- camouflage of bright feathers. The handicap has to be exaggerated enough so that it is an unfakable honest signal. A male pheasant or peacock must really expose himself to being seen by predators and get away in spite of it. Any Ponapaen could fake being a good farmer by devoting some extra attention to a little patch of regular yams. Like any reasonably successEvolution of Symbolic Traits 14-255 ful businessman can buy a nice model Toyota, Chevy or Ford. But when it is a flat out competition to consistently grow really huge ones, you call forth maximum the horticultural effort and wizardry you possess, and the size of your yam is likely to reflect real your skill.
However, such extreme devotion of time and talents to this one task may mean that the actual food production of the best farmer is only a little better than that of low-skill individuals who stick to real business. Like businesses along a suburban strip, everyone would be better off if they could agree to display only a small, cheap, but honest advertisement. But cheap, honest, signs are easily subverted by aggressive, unscrupulous advertisers, and a sort of arms race follows. The competition-driven exaggeration to prevent fakers from taking advantage only stops when each bird, farmer or business dissipates vast resources displaying a sign whose size and expense are pretty well correlated with the size, skill and health of the business/bird/farmer, because everybody who could afford a more elaborate signal would be driven to do so.
The difference between the pure run away and the handicap hypothesis is that in the case of the handicap, the exaggerated display is maintained because elaborate displayers still have the best genes/culture from the ordinary fitness point of view. Unlike the pure run away hypothesis, it is not just a matter of everybody else’s attention to style that motivates you to pay attention to Lexus drivers in preference to drivers of ordinary Toyotas. Style will be correlated with better ordinary adaptive traits; Lexus drivers will really tend to be better business people. We imitate them because we think they have better business skills, as well as because we know no one will ever imitate us unless we can display an unfakable signal of being good at business. A Lexus is hard to fake as a signal because it is costly. Only good business people can afford them.
It is not clear that the handicap hypothesis is much less pathological than the pure run away hypothesis. In the extreme case, practically all of an individual's advantages due to having superior genes or culture may be cannibalized to support the costly signal. If I try to reserve any of my cultural superiority to actually try to live longer or teach better, a rival with slightly less advantage, but slightly more willingness to display a more severe handicap, will attract more imitators. Every ounce of a superior Ponapean farmer’s extra skill may be poured into his giant, useless yams.
On the other hand, the famous evolutionary biologist W. D. Hamilton and his student Marlene Zuk (Hamilton and Zuk, 1982) have proposed that unfakable signals are usually not much more expensive than is required by observers to make an accurate assessment.
They think that the brightness and sheen of feathers and the red, blood suffused comb of a rooster are a bird's medical report to potential mates. If a chicken carries a large load of 14-256 Evolution of Symbolic Traits blood parasites, like malaria, it will be anemic and its comb more dull than a health bird that is resistant to malaria. If it is infested with lice, its feathers will chewed up and dull. On this hypothesis, honest signaling is possible with a minimum of costly, competitive, exaggeration required to generate unfakability. A Hamilton-Zuk signal is unfakable because it gives pretty direct visibility to the underlying biology. Attempts to evaluate the two hypotheses by biologists have been controversial so far. As far as cultural signals are concerned, the work has hardly started. A Lexus seems pretty expensive relative its transportation function to us, but perhaps you can think of a way to support Hamilton/Zuk for such cultural traits.
The exaggeration effect in the run away and handicap situations is most extreme in the case where individuals contribute no real resources to their offspring (or imitators).
Thus, male characters are most exaggerated in polygynous species like the chicken-like birds, where males play no role in rearing the young and spend all their effort trying to attract as many matings as possible. In many songbirds, where males and females both sit on nests and feed young, males and females differ little in coloration, etc. Likewise, in cultural prestige systems, the most extreme display behavior seems to occur in roles that are active in horizontal and oblique transmission. Media stars, who have a lot of money and the ability to reach the masses, often spend massively on display. Rich businesspeople who plan on leaving the family fortune to the kids seldom bother with really extreme showiness. If Dad is really willing to contribute directly to his family, it is important to Mom and the kids that he doesn't spend it all on fast cars and fancy whiskey.
In human cultures, roles effective in horizontal and oblique cultural transmission with little contribution of resources to imitators are very common. Casual friendships are a common example. There is ample scope for both men and women to behave like the polygynous males of classical female choice sexual selection. Our devotion to symbol systems that have apparently been exaggerated is not so hard to understand!
You are busy acquiring an expensive signal of your cultural worth to display to potential employers, your college diploma. There are two schools of opinion about what diplomas are all about, and they convey two extremes of the quality signaling idea well. The most common, championed by most professors, university administrators, etc., is that your diploma is an accurate index of what you know and how well you know it. Your diploma, transcripts, etc. are relative cheap and unfakable signs of your real skills. Many of these real skills you acquired in classes like this, and such skills justify the high cost of a University degree. Employers, mates, and friends value the diploma itself for its signalling your possession of the skills, but the sheepskin itself is a nearly costless way advertizing your talEvolution of Symbolic Traits 14-257 ents. This is like Hamilton's and Zuk's hypothesis.