Murder, Reciprocity, and Trust CHAPTER 3 ways than any other species.

Among the melancholy achievements

49 of

Homo sapiens is not only to be more intelligeIU.-than any other species

I

Murder, Reciprocity, and Trust

THE MURDEROUSNESS

OF MAN

Human social organization has been able, as we have seen, to exploit the advantages of large groups because of exchange ~etw~en unrelated individuals. But we still need to understand how this widespread reliance on exchange has been possible. It is virtually unknown in the rest of the animal kingdom, and it involves important risks. Only rarely do two individuals make a simultaneous exchange of goods or favors of. a known value. Much more commonly there is a favor extended by one m return for a compensating favor at a later date. You give me meat from your mammoth in exchange for a promi~e of meat in the. future; you make me a loan in exchange for a promised sequence of interest payments. How can you be sure I will keep my promise? Even when exchange is simultaneous it often involves a risk: in exchange for y~ur potatoes I have given you vodka, but the vodka may be spiked with methylated spirit; the motorbike I have given you may ?reak down. as soon as you drive it, even as I am struggling to repair the washing machine you gave me in return." . When individuals are close relatives this element of risk does not necessarily prevent fruitful exchange. Natural .selectio~ has favored g.en.etic mutations that encourage helping of relatives, which has two distinct advantages. First, relatives are more likely than nonrelatives to k~ep their promises, and secondly, it matters less if they do no~. A ~utatlOn that, on balance, makes its bearer more likely to help rela~lves Will f~vor the propagation of copies of itself in future gene.ratlOns provided enough benefit can be given to relatives for any particular cost t~ th.e giver. Indeed, although a sibling shares on average. only h~lf of. an individual's genes, this still makes for as close a genetic relatI~nshlp as the individual has with its own child. This does not eliminate rivalry among relatives - far from it - but it has favored, throughout the animal kingdom valuable and sometimes complex systems of interrelative exchange. In' contrast, helping nonrelatives is valuabl in voluri nary terms only when the favor is returned. But th in nriv \ to -II : I nr many; and, regrettably, human being hear in IHOI(' iCI II Illd 1\11' r u
A, rhe old ommunisr jol had it , "W~ I' 11'lid III ,111, 11I,1
lit

but also to have slaughtered members of its own species more vigorously, systematically, and cruelly than any other in nature. In his novel Crime and Punishment, Dostoyevsky depicted his character Raskolnikov as haunted by the enormity of the murder he had committed/ and many readers have seen the novel as support for the view that human beings have a complete and instinctive repugnance for murder, a repugnance that is mysteriously absent only in a tiny minority of psychopaths. It is a moving vision, but one that is sadly incompatible with what we know of the pressures that have shaped our evolution (as well as with much of modern history). For there are good reasons to think not only that natural selection has favored a tendency to kill other members of the same species but also that.!he coincid~ murderousness and intelligence is not accidental. On the contrary, the selectio for muraerou~ss and the selection rot intelligence are mutually reinforcing. The more murderous the species, the greater the selective benefit of intelligence to individual members, and the more intelligent the species, the greater the selective benefit of murderousness to individual members. The first point is reasonably easy to understand. Intelligence already confers a selective advantage when the main hazards you face are the hazards of nature. But the selective advantage is even stronger when the hazards include aggression from other individuals who are on average just as intelligent as you. So any evolution toward murderousness among our ancestors would have increased the speed at which intelligence evolved. The second point is a little more subtle. What are the selective pressures in favor of murderousness, and why should they be stronger in an intelligent species? The pressures in favor of murderousness arise from a simple fact: two unrelated individuals are rivals, both for resources and - if they are males - for the sexual favors of females. (Females are not sexual rivals in the same sense, since a male's impregnation of one female does not prevent him from impregnating another very shortly afterwards, though they may well be rivals for the resources that the male controls.) The fact that murder is much more commonly carried ut by men than women, even allowing for such factors as differences in physical strength and access to weapons, suggests that sexual rivalry ha been much the tr n r factor of the two in the evolution of human vi I n e.' Killin an unr I, t d 111I11b r f the same sex and species elimin t xual rival.' Thi in 'idrnr, lIy, 111 a likely explanation for the clisturbil1A tt'll 1('11' 01 Villllllt" 10 II' sso i, t d with a xual thrill; it is not, r '/'1'('111111 I I Itllldl,.1 ,II I( 1011 () 1 si'l minority but an

1'111'11,1 IIII'"Y

II ."

Murder, Reciprocity, and Trust ;hapter 3 ... gentlemen in England now abed Shall think themselves accurs'd they were not here, And hold their manhoods cheap while any speaks That fought with us upon St. Crispin's Day.'

51

Square Alfred Eisenstaedt's T' 'f . ' 3 1 The spOIls 0 war: VJ Day '" imes F Igure 'h' f Getty Imagesrrime Life Pictures. famous p oto or 1/,.

, 'ed sexual access that comes from u evolved anticipatlon of the mcr~as I I I n r inforced down rhc " ' f ri Is It IS one t 1 at H1 fl' cessful elimmatlon 0 nvars. ff I f r (rom univ r al, but U II the part 0 111 $-, a I d encY,on agesbyaten I I' 'ildl III rhus wh; 1lW i nt t make a dl,ff ren, :- 0 f:) I(1WI\" dl pi III' 1II'w w('11 wlwl\ Ii, pI, Y d pr( W SS 111 '()\1~ st, () 1\\1111111 I vllh tlH' 'I'y Ihl\l: hi III H\l' , h'nry V 1"1 h IS 110111 II

It is well known that once a certain characteristic becomes a basis for sexual preference, such preferences can be self-reinforcing. This is a tendency that has been adduced to explain such runaway evolutionary phenomena as the peacock's tail and the large antlers of some species of deer. The fact that females in future generations will be attracted by some characteristic, even a wholly arbitrary one, increases the adaptive benefit to any female in the current generation of seeking a mate that has that characteristic. As it happens, there is some evidence from preindustrial societies that males who kill other males tend to have larger numbers of children than males who do not - though this evidence is ketchy and controversial.' But even if a tendency for females to be exually drawn to fighters had no direct adaptive benefit (say, if the reater ability of violent males to gain food from others were offset by their increased likelihood of an early death), once such a tendency had b come established it could become indirectly adaptive through the greater likelihood that fighters would find mates. Even if nice guys don't finish last in the line for food, they may finish last in the line for mates. Nice guys may have other attractions (they may share more with their f 'males, for one thing), but if they have sons who can't do well in the ontest for females, that already counts against them as mates." Why, then, should the growing intelligence of the forerunners of l lomo sapiens sapiens have increased these selective pressures in favor of murderousness? In a species where contests are decided mainly by hrute force, a male can eliminate a sexual rival simply by forcing him pity ically to submit. But the more intelligent the rival, the more likely it I Ihat, having submitted now, he will find a cunning way to return to hi , xual pursuit later on. So eliminating permanently the rival who h.I been temporarily defeated is a strategy that confers much more seh Iiv benefit in an intelligent species. Two sources of evidence reinforce this evolutionary argument that human beings, especially males, are likely to be strongly disposed to kill III h 'I', unrelated individuals in the appropriate circumstances. One uur " is the behavior of other primates, particularly other apes.' This nk-nc needs to be interpreted with caution, since there is great variI' III primate behavior, even between such closely related species as 1,,1\'1 ,nz and bonobos, and this variety indicates that social and 111I~\i I fa tors can have a strong impact on the incidence of violence. II \ nvirorimental flexibility in itself should come as no surprise: inti, II is 0 ntral : rp,Unl '111 f thi bo k that institutional arrange-

52

Chapter 3 Murder, Reciprocity, and Trust 53

, d the ability of human beings to live ments have enormously Increase h ld otherwise be disposed to ' 1 ng those t ey wou b without VIOence amo , , he wild has nevertheless yielded so erfight. But careful observation In t, d b fear of reprisals many other ing discoveries. When uncon,straIne y" to kill individuals who are '11 100t opporturunes primates systematic a yexp f icide b unrelated males, for instance, not related to themselves. In anti hin anzees gorillas, and langurs has been regularly observed amo~g c ~~ lions)~ Among bonobos it is (as well as in some other mdamma s, sue a cases are kno~; M1t this ' I I on an no documen ted hei , certain y ess comm ,f 1 k cooperatively to ensure t err Inappears to be because ,ema es word' les not because the males d nst marau mg ma , d fants are protecte agai h Among chimpanzees, relate ' ' , lly trustwort y. , themselves are mtnnsica I h vi I t unprovoked attacks against I late to aunc VIOen , males regu ar y cooper mbers of other troops, even when such atisolated and defenseless me S h incidents were not known ield f d other resources. uc h tacks yie no 00 or Goodall and her collaborators (~n fact, t, e before the work of ~~ ularized the idea that intraspecies writings of Konrad I.: P that is now known from field violence was largely ritualize~'1 a view ong chimpanzees is particularly , b ist k n) The VIOence am I f studies ' to ,e nus, a e. rs to a Iarge exten t between groups of ma es ,0 revealing SInce It occu 'I tion This kind of behavior ' d wi h t parncu ar provoca I . unequal Size, an Wit ou f fora ing parties and looks disturbreflects the random encounters 0, bgt een groups of human males. ' 'I t s of aggression e w d ingly sum ar to pat, ern, h species most closely relate to Violence, in short, IS endemic -ns hng t fte this I'Sbecause of behaviour ' h it happens ess 0 en, man. In species were 1 d ' d not because of instincts that h hived to eter It an , patterns t at a e evo ,h f' of reprisals What@ver ItS funable WIt out ear . would b causes, p~ace , , damental pure VIOlence In pn~a t es, and especially in the great apes, cannot be described as p~~holog;ca\he innate murderousness of Homo A second source of eVI enche, or ts of contemporary nonindus' , th ograp lC accoun I sapiens consists In en, lar myth) are extreme y , , f hich (contrary to popu , , trial SOCIetIes, many 0 w '11 nds these reports and It IS violent." Considerable controbversydstI , slurrOeUvary across nonindustrial ' I h I I of 0 serve VIOenc undeniab e t at eve s '11 imperfectly understood. It may be societies for reasons that are sn ~erYII I societies are somewhat more ' h t imple agncu tura true, for Instance, t a s ( ble conjecture if only because warlike than hunter-gatherer~ ~t r~::~~~!ternativelY, this apparent tenthey have more resources to g icultural societies simply leave more dency may reflect, the fact that, ag~c~orm of torched huts and pillaged traces of the fighting they do (In, t e nthropologi t t vi it in unseth ) that they are easier f a or , , store ouses, or /1' f ,. I'll .c i oh "'V' I varranon d) , h the exp anation ", tied time. But w ateve.r I " r 1'1"f",1 It III.r,lI. wid h ha ha id I indu tria icn W " th ,t'llI a rive h Id over hun nn II' IIII' "I' , mar pr I I 11111 11.11' 111'111' ROll." U I

wrote ab e savages in the eighteenth century, has now been Convincingl dite . Anthropologist Carol Ember wrote a pioneering article in , s about Hunter-Gatherers," which showed that nearly two-thirds of hunter-gatherer groups for which records existed waged w:rratIeast every two years." Ethnographic accounts of high lev s 10 ence, erween in IVI s and between groups, exist for hunter-gatherers as different as the Akoa, the Bushmen, the Tasmanians, and the Yanomamo. Among preindustrial agriculturists, regular and deadly warfare has now been documented for societies once thought to be peace-loving, such as the Pueblo Indians of the American Southwest. II Many such ethnographies could be cited: a striking example is the book Bl~eir A;fYu~ent, in which anthropologist Mervyn Meggitt records th~les of violence among the Mae Enga people of the central highlands of Papua New Cuinea." Once again the message is sobering: where there are no institutional restraints on it, systematic killing of unrelated individuals is so common among human beings that, awful though it is, it cannot be described as exceptional, pathological, or disturbed. In short, and whatever the precise evolutionary origins of such behavior, the potential costs of being cheated have become dangerously high for human beings dealing with nonrelatives. Not only may you never be repaid for the mammoth meat you gave me, but you may be murdered in return. Many itinerant traders making tentative first Contacts with an unfamiliar tribe have suffered just such a fate. In the circumstances it is n tonishing that systematic exchange among nonrelatives should have 'volved at all, let alone that it should have become the foundation of the fantastically complex social and economic life we know today. Remarkably, trust in nonrelatives has become an established fact of () iallife. When I go into a shop, a person who has never seen me before will hand over valuable goods to me in exchange for a scribble on a piece of paper. When there is a knock on the door, I am prepared to allow into Illy home a man I have never met, who is wearing the uniform of a local dorn stic appliance store. Why do I infer that his purpose is to repair the washing machine instead of to kill me and rape the women of the house? , low has this degree of trust become possible? Buyers do sometimes pass " I I hecks; murderous rapists do sometimes pose as repairmen _ but it It q'ln n rarely enough not to threaten the general trustworthiness of

II",",

ial rr'n"orion,.

What is the e~.

:L<2 <l
, l

Coo!,

III

1111111(

h' foifly on(id'lIl thn: Oil" III '('Slors 'volv d t tru t familiar 1.llivl's IwlOl't, III Y 1111(' 10 "live' 1111 I ill 11'l111\("s. Wh 11 tw indi-

-( t~.JL-!+ \
54 Chapter 3
.7~~V"'"

Murder, Reciprocity, and Tri. viduals expect to see each other often in t?e future, th~y have an additional incentive to keep agreements: cheating might bnng a ~hor~-term benefit but it would jeopardize the possibility of cooperation m the future.' Provided the benefits of future cooperation are sufficie.ntly large and sufficiently certain, and the future in question is not too dls~ant, the individuals may resist the temptation to cheat for sh?rt-term g~m. Such cooperative tendencies probably evo~ved ~n two p-a.1ll ways. FI~st,. they are a by-product of the evolution of mtelhgence-JAs human social mte~ligence developed, individuals could increasingly ~akulate. tha.t then long-term interest lay in keeping rather than breaking certain kinds ?f agreement, namely, those with individuals they expected to see agam repeate<lly in the relatively near future and from. whose good':"lll they expected to benefid The second is thr?ug? se.lect~on. for. what IS sometimes called "redp(ocity," namely, an mstmctrve inclination to do unto others as they have already done unto you. If others have treated you well, you treat them well in return, but if they have ~urt you, .you hurt them back. An eye for an eye, certainly, but also a ?Ift for a g.lft. Both of these motives for cooperation have been Important m human evolution, and both continue to play an important part in human motivation today. At first sight the distinction between them may se~m rather unimportant: what difference does it make whether the motive for cooperation is calculation or instinctive reciprocity, when m~st ?f the people with whom you deal are those whom you ~av~ dealt With I~ the past and whom you hope to continue to deal With in the future. And even if you knew the difference in principle, ho,:" coul~ you tell m practice? But in fact the distinction matters, because It provides the key to how humans have evolved to trust complete strangers, those they have never seen before. Studies of human cooperative behavior in many different contexts ha e clearly established that the hope of future cooperation. plays an . portant role in reinforcing cooperation in the present. It IS known that people cooperate more when they interact frequently, when they have the means of telling whether others have cheated, and when th ains from cooperation, as opposed to going it alone, are larg.e. Thes actors have proved important in such diverse cont~xts as ml.htary engagements care of the environment, and the effectiveness With which , local political institutions respond to the needs of t hi' popu anon, 11 e There used to be, however, much controversy about whether people also have an instinct for reciprocity (a reaction to how they have b 11 treated in the past independently of th pro, p 'ts for future c~lJabor(l. tion). It was thought to be impos ibl to Ii, t'll! IIII'I 'on motiv frol~1 the other wh n aim t all human int '1':1 IItHI IIIVlllvl' I '\l11I .. s rrnns: tion from pa t involv m I1tS to (111111" 11111 Iplll till

55

"

Now a set of ingenious laboratory experiments b rns and colleagues at the University of Zurich has shown that individua s receiving generous treatment from others they will not meet again nevertheless respond generously to them in return." For instance, in an experiment that mimicked behavior in labor markets, subjects took the part of workers or firms and were randomly and anonymously matched with each other (the. partici~ants in the experiment were students playing at computer termmals). Firms could choose to pay a minimum wage or a wage above the minimum; workers could choose to make a minimum investment (a monetary equivalent of "effort") or a level above the minimum. After each turn participants were rematched, and none knew whether the partners with whom they were matched were partners with whom they had been matched before. It became clear as the experiment proceeded that workers who had been ge~erously treate~ by their employers were likely to make high levels ~f mve~tment. ThiS mvestment brought them no benefit, since any firm With which they were matched in the future would not know how they had previously behaved. So their behavior must have been due to ," os J r ciprocity and not to calculation. Nevertheless, firms in turn gradually ()o.!- ::;::::I arned to benefit from this, as the cost of paying high wages to workers was more than outweighed by the value of the extra investment the L'ff~'i~ undertook as a result. So firms that paid high wages were more profit, :',ble than t~ose that did not -.- not the conven~ional capitalist view. 15 ~here re~am unresolved questions ab?~t the Wider applicability of re= .r ~J1ts obtamed. under laboratory conditIOns, though the findings have b II1ce been rephcated so many times by 0 rchers that there is no " longer any serious doubt about their obustnes (they are not due to hating or misunderstanding by studen example, and real money I nt stake). But these findings suggest not only that cultures of cooperatIOn. may persist through reciprocity but also that grou s in which coopI "\tIve habits have developed may be more economica c ss ul

D U/;/

~I
>

. IC mo nar ow - n res e e avior is the norm. - T:, . I' trndoxically, too narrow a nose for profit may be bad not only for the --7'. W/~ 11111ut also for the bottom line. b ~ Sim,ilar experiments by ~ehr and others have also established a sys-, . . , tl : 11:\ t". ~ ndency for l~dlVlduals to repay unkind behavior with unkind h, II IVI .1 ven when this bnngs no benefit to themselves.16 In one experijV! 1IIt'"~ with two subjects, th~ first was asked to propose a division of a It I" 11,( I tw n the two, while the cond could SImply accept or reject lilt ,1)1'01 o,sc I. In th vent of r j tion n ither received anything. Once , ,1111 J .rs w r rand rnly (In I nnonymously matched on a one-off -l?l44fi~ sui It I I , so (/1 'I' iould I ' no possill' 1('11'(il (rom pr nt behavior in ....f)~,o hlill il1l1l1'l1'illl-\ (lit 111'(' 'O()IWI'IIIIIII, It W1 ,I'll' rbnr th nd IT

IIITI

Ib

or

I':I~'

C,O,(

56

Chapter 3 Murder, Reciprocity, and Trust 57 your

subjects would systematically reject proposals they thought gave them too small a share, even if by rejecting the proposal they would receive nothing at all. Such behavior is sometimes known as "cutting off your nose to spite your face" - a description that implies it is irrational. But rational or not, it is clearly widespread in human nature. Knowing this, the first subjects would ensure that they gave enough to the second subjects to keep the risks of rejection low, although this rarely implied equal division, especially when the sums of money at stake were substantial. Yet another set of experiments conducted in groups suggests that the presence of reciprocators who not only respond in kind to cooperative behavior but also help to punish noncooperation in others, even at some cost to themselves, may be critical to ensuring that cooperative habits become established. Where the only motive for cooperation is the past generosity of others and there are no sanctions for cheating, an initially cooperative group culture may gradually unravel as individuals succumb to temptation; but when it is known that aggrieved parties can take revenge, and will do so even at personal cost, the incidence of cheating in the experiments shows a sharp decline." Why is reciprocity important for the ability to trust strangers? First, it explains why the complex web of trust that underlies modern social life does not unravel as soon as unscrupulous individuals test its strength. Economist Kaushik Basu describes a simple problem." You take a taxi ride in a large, unfamiliar city, and when you have reached your destination you pay the driver the amount you owe. You have benefited from the ride and you will never see the driver again, so why do you bother? (This is the kind of question that sometimes gets economists a bad name.),f19 The answer, very probably, is that you just do: you feel the instinct of reciprocity, either in the form of a wish to behave decently to a person who has just behaved decently to you, or because you believe you have a moral duty to do so (these two are not so very different, since the latter can be described as a wish to behave "cor, rectly" toward someone who has just behaved correctly toward you), But now suppose you feel no such instinct, and instead you ask your self, "What's in it for me?" You will probably reason that if you refuser] to pay, the driver would make a great fuss, at least embarrassing you ill front of bystanders, possibly getting you involved in a fight, and Pl'l haps even leading to trouble with the police. It's just not worth III" bother for such a small sum. So it seems that YOLI don't have to appru]
Will DU"o"1in '1'111'S/1I1 \I IIf Ilill/o,wphy recounts lIill! Schopcnhnucr used I<) Iri1v(\ II "PIIII 1111' "dill III ill(' SllP'l of 0 ,'CSIIlIII,'1I1 menl 1'0 cncourngc rhe waiters, hut tlWIl 1I'lIl1lYI II III ill, I lid,
the philosopher

to reciprocity behavior.

as a motive: plain self-interest

is enough

to explain

But wait. The driver could just as easily make a fuss even if you had paid, threatening you with embarrassment, a fight, and the police in order to make you pay up a second time. None of the other parties Would be any the wiser. The driver will not see you ever again either, so why does he refrain from trying to wring a second payment out of you? Once again the answer might be that the driver is moved by an instinct of reciprocity - or he might just be afraid that you would react violently if he attempted to cheat you in this way. But why should he expect you to react violently? However far down the line we go to prove that your behavior or his is just the consequence of enlightened self-interest, at some point we still need to appeal to the fact that people behave differcntly if they feel they have been cheated than if they do not. Since no, body else observed the transaction taking place between the two of you, ' no third-party enforcement mechanism can explain why that transaction affects how you subsequently behave. Even when third-party enforcement mechanisms (such as the courts) role in strengthening the web of trust, reciprocity is the glue that makes these mechanisms credible. Why does the judge make any IIItempt to decide the case on its merits instead of simply finding in !VOr of the party who can pay her the most? Again, perhaps she just docs. Or perhaps she is being self-interested, reasoning that if she takes t bribe she will suffer for it. But why should she think that, unless some IIIhcr people (those who report her bribe-taking and her colleagues who IIIYl~stigate the allegations) also behave in ways that are sensitive to the 1I1I1'insicfacts of the case? Every person in the web who reasons selfIlIlt'l'cstedly can do so only on the assumption that someone else, some-I wlu-re, will not be acting according to self-interest alene. I( r FOI'l1'lalmechanisms for enforcing agreements~on 1111 instinctive and uncalculating It' tend.eacytO repay kindn~~; '~ith IlltlllcsS1rud lilIkinaftess with reveng2what such mechanisms can do,

do playa

.1 U

1bl

rl/(
/tMtJ7

'I.J'

h"w,'y",'; is to ensure that a little reciprocity goe.s a long way. Modern Iii 'id Ilfc depends on a large number of ingeruous arrangements for
IlIltll'dng reciprocity with self-interest. When we meet and transact lilt ~11'nngcrs, we do not always need to ask ourselves endless ques"t nhour their intrinsic trustworthinesao long as there are enforce11'111 IIH'l'honisl11s in place, and so lonb- as there is a good enough lilt I' Ihnl' somcone somewhere is intrinsically trustworthy, we may I 1I1'"d 10 worry whcchcr the strangcr at Our door or at the checkout III Nllop is genuincly motivated by rCCiP~'Oity or merely by selfI, Jo'llllt", will do (oJ' tlw pur'pose in hnnd.l-)/(A-~--wlJVl' ~\. '"' II ,!IIC'NI/OIlS Ihol would St'('111Pill'O/lold /, n "('nsonnbly stolJc and "-

tk6

t /

* Though not just economists:

r('7'; WI ~ ~ I" <::

IJ
t")

b4.'~' ~

58

Chapter 3

C~

y~
.

Murde Reciprocity, and Tw"

'

59

well-ordered society may be quite reasonable and even essential in a poor and violent one. Reciprocity has also mattered in the history of humanity because it has enabled hunter-gatherer bands to take the first cautious steps toward conducting exchange with strangers (such contacts occurred, as we have seen, well before the adoption of agriculture). An itinerant trader making the first contact with an isolated band whose only motivation was self-interest would almost certainly have his goods stolen and would be lucky to escape with his life. After all, the band could hardly reason that offering him goods in return would be of benefit in the long run, since it would have no reason to expect him to come back again soon or often, given that this was the first time the band had ever seen him. It has surely been reciprocity that, prehistorically, tipped the balance between hostility to strangers and a cautious willingness to deal with them. Often that reciprocity will have been betrayed, as many of the Native Americans who dealt with the first European visitors discovered to their terrible cost. But their case, though tragic, made ultimately less difference to the future of human society than the reverse - namely, the discovery that a willingness to trust others could produce important benefits to both sides. Had it not been so, modern society as we know it could never have evolved.

formed very much better on the test. This led Cosmides and Tooby to conclude that our reasoning abilities are sensitive to context in ways that would have been beneficial for our ability to spot cheats during our evolutionary history. 21 But calculation is not a virtue that necessarily inspires confidence in others. Suppose I know you have a shrewd nose for your own best interest, and nothing ~ut. Then I may be cautious about cheating you, but I shall also be cautious about trusting you. I shall not make spontaneous gestures of friendship, for I know that these will make no difference to your ~ubsequent' treatment of me, which will be based solely upon calculation of your future interest. Pure calculation, in other words, exercises trust wisely but does little to inspire it. Reciprocity, almost by definition, is less wise. It inspires individuals to be generous to others purely to repay past generosity, even when this bnngs no future benefit. It can also trap individuals in cycles of revenge generated by past wrongs. Based on emotion that is triggered by outside ~vents rather t~an th,e conscious decisions of the individual, reciprocity ~s open to manipulation by others. But when it comes to inspiring trust in others, Its msensinviry to calculation is precisely its strength." If I know that my present generosity to you will incline you to help me in the future, re,gardless of your interests at the time, I shall be more likely to take the nsk of helpmg you. Your disposition for reciprocity makes yo~ a more credible partner than someone who has no such disposition. ThIS character, in short, gives you a power of commitment beyond the reach of the most sophisticated calculation. It is a power that the calculating can appreciate even though they cannot aspire to it - calculators would rather trust reciprocators than trust other calculators like themJ selves. or Il' I'U. Th~se considerations suggest that individuals who can simultaneously ~ exercise trust shrewdly and inspire trust in others need to have some 7'disposition for calculation in their dealings with others - but not too ~ mU,ch, o,r no one will trust them. They also need some disposition fo~ .J hi ~ reciprocity - but not too much, or others will exploit them, and the _ memory of past, wrongs will cast too long a shadow on their lives. They need a way to SIgnal to others that they have the element of reciprocity that makes them trustworthy, and they need to signal it in a way that any p~rely calculating person could not convincingly mimic. The psychologisrs Mlc~ael Owren and Jo-Anne Bachorowski have recently proposed an mge~IOus theory according to which smiling and laughter may have evolved m human beings as just such signals." Both smiling and laughter ,are human capacities for which only the most rudimentary f rrn eXI t In orh r P cies. Both appear to signal emotions associated with a Iii ing (01' otll('J', rind a willingn to behave generously toward

SMILING,

LAUGHTER AND THE NEED

TO SIGNAL TRUST

In short, calculation and reciprocity have complementary virtues. It is worth asking a little more precisely what these virtues are. Calculation enables an individual to trust only those others she can afford, on best evidence, to trust, and to do no more for others than is strictly in her own best interest. It's an essential ability for an individual facing a variable natural environment and a complex social environment. It is likely that the increasing size of hominid brains over the last six or seven million years reflected the selective pressures arising from a more diverse ecology and the growing size of social groups. Only individuals who could calculate well, who could judge how their behavior should be adapted to fit different circumstances, who could spot a cheat, were likely to survive to reproduce. And there is good evidence that human beings have senses very finely tuned to signals of the trustworthiness of those around them. In one famous experiment, the psychologists Leda Cosmides and John Tooby showed that subjects often had difficulty ~g a logical puzz~ich required them to identify playing cards that failed to conform to a rule of play. However, when th v ry ame logical puzzle was reformulated as a problem of identifying I -opl who had failed to conform to a rule of ocial behavior, t hI Ilhit- t per-

i_

Is

..
60 Chapter 3

~~~

..J

..

" -~
Murder, Reciprocity, and Trust \)v.,; 61

"':"''':.'0 ",\

them-what Psyc~ogist~U "pOsitive a ect." Both appear to trigger such feelings in oth rs as well. Owren and B horowski suggest that the capacity of smilin and laughter to act as re 'able signals of positive affect, and theref e of trustworthiness, would ve made them highly adaptive for the . dividuals that could smile an laugh. Any genetic mutations capabilities would theref e have tended to spread. iven their reliabili as signals of trustwor\l{lipess, evolution would also ave tended to favor a tendency to respo~armly to them in turn." -c: However, any signal that makes other people rhink I am trustworthy is one it would be highly useful for me to be able That way I could make people trust me, and do favors for me, ~ut incurring the cost of doing favors for them in return unless it suited me to do so. So, Owren and Bachorowski suggest, no sooner had smiling become reasonably well established as a reliable signal of trustworthiness than it also became adaptive to be able to counterfeit smiles. Smiles that are under deliberate control are known to use a different set of neural circuitry than spontaneous smiles (the latter are also known as "Duchenne" smiles, after the nineteenth-century psychologist who first wrote about the difference). Not everyone can fake smiles successfully _ indeed, politicians are predominantly drawn from among those human beings who can. Many people (myself included) are effectively barred from political careers by their inability to produce convincing smiles to order for the camera. But enough people can do so to suggest that the evolution of smile mimicry has proceeded quite far in the hu-

G\)

toi~

~ ~

man species. But almost nobody can fake laughter convincingly. Laboratory studies show that many people are unable to discriminate reliably between spontaneous smiles and those produced by good actors. But virtually everyone can tell the difference between spontaneous laughter and the deliberate lau hter of even very talented actors. And deliberate laughter provo es much less positive a feet in those t at hear it than does spontaneous laughter. These facts lead Owren and Bachorowski to suggest that laughter probably evolved later than smiling (no doubt in response to the mimicry of the first prehistoric politicians). The fact that smiling was losing its reliability, because so many people could fake it, made it valuable to have another, better signal of positive affect. And the possibility that laughter evolved later would explain why the ability to fake convincingly has not yet had time to evolve. A telling piece of evidence in support of the signaling theory of laughter is the way in which, across all kinds of cultures in the world, people who have made a business deal with each other tend to seal the deal by having a drink together. Drinking alcohol notoriously affects people's

judgment. In fact alcohol is a depressant that not only makes people feel all stimuli less strongly but particularly diminishes people's sensitivity to pains, including future pains." (The reason why alcohol provokes car accidents is not primarily that it slows down people's reaction times but that it makes them reckless, through diminished sensitivity to future dangers.) In short, if people entering into a business relationship needed to keep a clear head in order to work out carefully how much they could afford to trust their new partners, having a drink together would be the worst possible way to seal a deal." But alcohol is also, famously, a disinhibitor. Most importantly, it makes many people laugh. In all cultures, many businesspeople spend evenings exchanging jokes that, to begin with, virtually nobody finds funny, but at which everyone, toward the end of the evening, is laughing uproariously. At the same time that it disables people's capacity for exercising trust wisely, alcohol is enabling people to inspire trust by stimulating an excellent signal of positive affect, namely laughter, that is not under direct voluntary control." Mimicry should not always be considered, though, as tending to weaken or undermine trust. One of the first kinds of social interaction of which babies are capable at just a few weeks of age is to return the smiles of those who have smiled at them. All of us learn at the very start of life that smiles are the sign of the love and goodwill of our relatives, particularly our parents. The first traders that successfully made contact with hunter-gatherer bands and persuaded them to exchange instead of to fight must have tried very hard to smile convincingly in order to trigger a reciprocity response. Without some degree of successful mimicry such exchanges might never have got going. Here, as in many other contexts, trusting strangers is a process of conscious and unconscious mimicry of the way in which we trust our family and friends.

So How

DID RECIPROCITY

EVOLVE?

Researchers are still puzzling, though, over how exactly human beings evolved the kind of reciprocity that mimicry exploits." The fact that, in the laboratory, people are willing to incur personal costs to reward those who have treated them well and punish those who have treated hem badly even when these people will never knowingly see them again, has led some researchers to suppose that reciprocity is an evolutionary "mistake. "29 It would have been more adaptive, according to this view, ro evolve a tendency to show reciprocity only toward those you are likely to see again (we could call people with such a tendency "opportunisrs")." r lowcvcr, nskillg why reciprociry evolved may be a little like INking why humnn mlll!'N hovt, dl'vl,lolwd n desire to have sex with

62

Chapter 3 Murder, Reciprocity, and Trust 63

beautiful females who are currently using contracepti~n. Obviously an evolutionary mistake (if only all mistakes were so enJo~able). But not such a surprising mistake, since what evolved was a. desire to have sex with beautiful females, period. Reliable contraception arnved ~n the scene far too late in evolutionary history to have created any kind of adaptive pressure for a more discriminating desire, directed o?ly towar? those currently not taking contraception. Perhaps the evolution of reciprocity is no more of a puzzle than this.. . Since Professor Ernst Fehr and his fellow expenmenters have arn~ed on the scene even later in evolutionary history than th~ contraceptive pill it is evident that whatever else reciprocity may be, It ~annot helpfull~ be described as a tendency to return in kind the beh.aviOr of others whom the experimenter has ensured we shall never knowm~ly see again. It is a tendency to return in kind the behavior of others, penod. So were our ancestors often enough exposed to people whom they had reason to believe they would never see again for there to have been som~ real /l . selective pressure against reciprocity over ~voluti.onary time? ThIS re. ~ains a hotly debated question." One pOSSIbIlIty IS that they were not, \ "" . \\ that most people who were generous to our ancestors were people they { J :;.~" had at least some reason to expect to see again in the future. Although . ') our ancestors were certainly highly wary of strangers, most of whom ~ ( ~- 1~ would have been hostile, if not lethal, friendly strangers may have b~en \) cr-.~,y, a great rarity, certainly unusual enough for there. to be no great adaptive r- \) essure in favor of cheating them. Indeed, fnendly strangers can be tho ht of as successful mimics of our genuine friends, successful b~/ cause f their comparative rarity until very recent times. A seco~d POSSIbility. is hat there was indeed some real selective pressure against r~Cl\~ \ procity at n individual level, but t~at this was offset by the adaptive ,ff_ benefits to roups that displayed reciprocity (such ~roups ~I?~t ~e bet\\'-" ~ r placed to trade with others, for. instanc~).32 A third possibIltt~ IS that selec . ressure against reciprocity and m favor of opport~nIsm has been opera . g for too short a time to have become dominant m ~uman populationse x erimental evidence su ests that a ro. omon of subjects disp ay reciprocity, not that they all doJ In short, reclprocat?rs mId opportunists rnexisl. FInally, as I suggested above, opportunists may do better than reciprocators when they get a chance to cheat, but unless they can hide the fact that they are opportunists, they may be given fewer opportunities to cheat in the first pl~c~. They may eve3~fac collective ostracism from the reciprocators within their gro.up. Th question remains open; all of these explanation may contain a~ .1 ~ ment of truth. Fut we can safely agre that w ow the a t ~lshll1g c mpl xity of m- dern ociety to th PI' \ 'f) '( I) III I 'ost. m . I)' ,

history wh.o risked slaughter at the hands opportUnIties for mutually beneficial trade.

of strangers

to explore the

RECIPROCITY

AND REVENGE

r~

'Y \
Y

Reciprocity can ~f COurse be dangerous, and not just when it is manipulated. Certamly, It helps to establish cooperation and exchange. This is partly because of the tendency to repay kindness with kindness, and partly becaus~ the tendency to repay unkindness with revenge makes people more ltkely to keep their promises. But this second characteristic also mea.ns that once promises are broken, the participants can become trapped in a. cycl~ of revenge. Testimonies to this from history, anthropology, and Im~gmative literature are legion. Modern tragedy has often located dramatic compulsion within human beings' Own iron subjection to the laws of revenge, which endows the action with a sense of inevitability in works such as Shakespeare's Romeo and Juliet and Gabriel Garcia Marquez's Chronicle of a Death Foretold. The codes of honor of the Mafia culture of Sicily have come to seem strangely dysfunctional in the modern world, but they have a comprehensible rationale as the work of sociol.ogist . I'-gO Gambettq has shown US.35Gambetta 'emphaSIzes the way m wh ch' a a ran;lfies originally took upon themselves the role of providing som the underpinnings of trust in a complex ~~dern socle~y, because the Italian state proved incapable of doing so in SI~tly at the nrne of Italy's unification in the mid-nineteenth century. But this role came at a heavy cost: the same human motivations that can reinforce a culture of trust can also entrench a culture of distrust. In icily's case, trust of insiders has reinforced distrust of outsiders a cultural handicap that has cost the region dear as it seeks to integrate itself into modern Italy and the modern world. Fortunately, the disposition to trust in members of a certain group more than others has often been a help, not a hindrance, in widening ~he range of people on whom we can rely. An unknown trader coming II1t Our VIllage may suff~r from being an outsider to our group, but he may benefit from belonging to some other group or tribe with some of who e members we have previously dealt or to a more distant branch of Our wn tribe; he may even come with tokens or letters of recommendalion. 't rhi day, immigrants to North America often seek out members o~ rh ~r. wn thnic community; migrants from the countryside to Indl(11l ItJ Fe 11 contact mernb r of th ir own caste; the stranger at my 111101' stan Is a b tt J" han of adrnis i 11 if h is wearing the uniform of .1I'('puilil 'oml any. 'I'll, f',/'Idunl illl'l'I'lIioll
0
111l'" '1"1111\ S ()

tor

m n

ur an

stors, a

II, s

1'0 II\()

('

111111(111'1

Ih

J")

,(

II"

("USI'into

I, ,.

r r

gional,

64

Chapter 3

Murder, Reciprocity, and Trust

65

national, or even global cultures of trust, punctuated though it has been by many episodes of reversal, is at the heart of the history of modern life. It is important not to romanticize this process: when.I ~ay I can trust a stranger, I do not mean that I like him, have any cunosity ab?ut him as a person, or care in any deep sense about what. happens to him, The point is that I do not need to like or care about him m order to ~e able to deal confidently and reliably with him. Some people ha~e .seen m this fact a chilling, even dehumanizing quality of modern SOCletIeSand have yearned for the times when those who mattered to us were those e could we knew, people whom we might hate or love but to whom not feel indifferent. The economistJames Buchana~has descnbed how a purchase of fruit at a roadside stall can take p!aceln s~ite of th.e fact that neither of the people concerned has any particular interest in the well-being of the other: they are "able to ... transac.t exchanges efficiently because both parties agree on the propert~ nght~ relevant. to them." In reply the economist Samuel Bowles descnbes this as creatmg "a psychological environment of anonymity, indifference to others, mobility, lack of commitment, autonomy" and concludes that "we learn to function in these environments, and in doing so we become someone we might not have become in a different setting."36 People vary in how much this troubles them (Bowles is pointing out that anonymous markets have an important effect, though not necessarily one w.e should, on balance, regret). But it is absolutely clear that a warmer timbre .of human interaction in all our encounters is quite incompatible WIth the degree of complex interaction upon which most of us now depend. Indeed, it is precisely the most tenuous, the most anonymous, of our links to the outside world that do most to connect us to new opportunities. The powerful implications of this simpl.e fact were dra~n som.e years ago by the sociologist Mark Granovetter in a famous article. entitled "The Strength of Weak Ties." Granovetter began by reportmg a study of the social networks that helped unemployed people to ~nd jobs. Personal contacts are often a more important means of finding work than formal institutions such as employment exchanges. Yet Granovetter found that people were much less likely to find work through close friends (their "strong ties") than through casual acquamtances (their "weak ties"), even though close friends should in princ~ple have more reason to help them. The reason is that your close fnends are likely to know many of the same people that you do; the! are therefore less likely to be able to bring you genuinely new inform~tlOn and opporrunities." Casual acquaintances, in contrast, are more I~kely to act as a bridge between otherwise closed groups and rh oursid ~or1d. Granovetter's article and other work in pir d by ill \IIIV(' I rovid d a w 1m unt rw ight to th I ia lrlin ill II illllli y.l! whi .h hn spo-

:v

ke~ of the alienation induced by modern living without noting how this IS linked to the very conditions that give it energy and creativity. . Of the impersonal institutions that have enabled this gradual integranon of local cultures of trust into a wider culture governing relations between strangers, none has been more central to the process than the ~nstitution of property rights. Property rights are a set of rules governmg who has the right to manage the various valuable resources in our environment, to enjoy their fruits, and to dispose of them to others. Without the assurance that the resources you theoretically own now will be protected from marauders until the time comes for you to repay me, no amount of trust in your good intentions may induce me to be generous to you today. Another way to express this is that trust cannot be purely bilateral: trust between any two people rests on a web of trust between each of them and the others with whom they also deal. For this reason, as we shall see in later chapters, all societies have needed the maintenance of defense, civil order, and some degree of consensus about what the social rules decree. Some historians have even argued that the ability to establish such property rights, and the social consensus that enables them to be enforced, is the single most cogent explanation for why some countries (such as Britain and the Netherlands) were able to industrialize and grow faster than others at a critical period in earl modern history." As with other institutions, property rights rest on delicate balance between reciprocity and self-interest, and different soci eties have placed different emphasis on the formal and the informal (0 consensual) components of that balance. A growing literature on wha has come to be called "social capital" examines the many subtle factor that explain why trust ~ars to have become more securely embedded in some societies than in others." What all stable societies have in common, though, is that the balance between reciprocity and self-interest holds even when unscrupulous individuals test its strength. It is time to look more closely at some of the social institutions that have been responsible for this extraordinary spread of the willingness to trust strangers. The hallmark of the most successful of these institutions i their ability to entrench a culture of trust with a minimum of explicit nforcement. For instance, what distinguishes safe cities from dany rous, crime-ridden ones? Jane Jacobs has written that "the public I ace - the street peace and sidewalk peace - of cities is not kept prirnarily by the police, necessary as police are. It is kept primarily by an mtncate, almost unconscious, network of voluntary controls and stand: rd ~mong the people therr;;selves, and en~orced by t~e people themIv . Indeed, h ar u d, nee a street IS well equipped to handl I fa n 'r r n it h(1$ I 01 h n ood, ffective demarcation between priv.u . (lncl public NI n '(' III I II 1 11,si upply of activity and eyes, the
(~

."
..l\

66

Chapter 3

. . . . ~ In other words cities need an institution ore strangers t h e merrier. '. I I' m keeoi doubtless backed by formal sanctIOns but arge y ino f peace- eepmg, . . . h b t ble m ' d to day activity And that insntunon . as to e sa, tis f orma Im Its ay- -. the sense that people react to its presence by behaving more, no es, cooperatively. B fi hall We shall look in more detail at cities in chapter 7.. u: rst we s look at one of the most remarkable trust-creating msnmnons of all, that no sane person would take seriously for a mon:ent w~re It not or the fact that everyone else in society is normally qurte willing to do so.

u. -~

CHAPTER

Money and Human Relationships

I
III
1

ore

MONEY

AND BARTER

lt is money.

11111,\

I II

II

III

In northwestern Russia a man and a woman are bartering goods, each offering something they have made against something they want. This is the way goods have been exchanged for most of the human past, and it epitomizes the traditional culture of face-to-face interactions. What are they exchanging? Animal skins, it turns out, against rather simple shoes. You could be forgiven for imagining the scene as taking place in a forest clearing, or at the side of a muddy track, with horses occasionally stamping their feet and launching clouds of breath into the winter air. In fact, the woman is sitting in a heated office, and the man is nowhere to be seen. They are talking on the telephone. lt is February 1992. The consignments whose exchange they are negotiating are far too large to fit on a horse and must be sent for several hundred miles across the frozen landscape by container truck. Almost everything about the transaction speaks of the industrial age: the animal skins have been scalded with chemicals, the shoes have been cut and stitched by machine, the goods will be accompanied by invoices and bills of lading, the man and the woman are dressed in machine-made clothes and work by the glare of strip-lights. Only one thing is preindustrial: they refuse to use money. Inflation is currently running at 2,000 per cent per year, and the money would lose its value faster than they could pass it on. Now travel forward to 2001. Throughout the decade that has elapsed ince the end of communism, barter has taken the place of money for a r markable proportion of transactions in a society that in most other r pects has been seeking to model itself on the market economies of the indu trialized West. These are not just deals between individuals trading h L1 ehold goods or personal services (a well-known phenomenon under '( mrnunism). Firms are trading commodities with each other instead of 11':."\ling them for money, and doing so at all stages and all scales of I rod I ti n. No ood ar t large or too sophisticated to be part of a hnrr r hangc; tanl s, nil') Inn ngin , and oil and gas refinery equipmcur have nil f ntur I ill \I h d -nls. tin I 'fst lndilll\ 110 I 1111111'1'1\ 0 'it'l '( n fun ti n wirh Lit 111 n y