The Exploitation of Women in Evolutionary Perspective

In the previous issue of Critique of Anthropology, I tried to lay a philosophical and theoretical foundation for a scientific analysis of social evolution. Essentially, I suggested an empirical approach to Marxist anthropology, and I urged the application of criteria of radical materialism in the evaluation of our most prominent concepts and theories. In particular, I criticized certain libertarian notions


The Exploitation of Women in Evolutionary Perspective
John H Moore, University of Oklahoma In the previous issue of Critique of Anthropology, I tried to lay a philosophical and theoretical foundation for a scientific analysis of social evolution. Essentially, I suggested an empirical approach to Marxist anthropology, and I urged the application of criteria of radical materialism in the evaluation of our most prominent concepts and theories. In particular, I criticized certain libertarian notions among Marxists, especially Engels and Hindess and Hirst (1975).
In this second article, I wish to return to a criticism of libertarian ideas, not polemically or in the abstract as before, but by reference to ethnographic data, following my own suggestions on the utility of an empirical approach.
The general problem I am considering here is whether women are exploited by men in hunting-collecting societies. In contrast to the libertarians, I believe that they are, and have been throughout history and prehistory. If we can establish this fact, I believe we have taken a large step toward the creation of a scientific theory of social evolution, along lines suggested in my previous article.
The ethnographic evidence Despite the fact that I have advanced my theoretical arguments in some detail in Part I, I do not believe that 'pure theorists' will necessarily be convinced by them. As I have mentioned, their reliance on intuition and their rejection of data makes their position unassailable because incontrovertible. And so now I will necessarily adopt the tactic of ignoring their ontological position and I will proceed immediately to an empirical critique of their theories.
In approaching the data, the initial problem is one of samplingdeciding which societies I shall examine to determine whether exploitation exists.
At present, I am assembling material for a longer work in which I intend to use many different sampling techniques. For purposes of this article, however, I will use only two. The first of these is the riskiest one I have been able to devise -I will use as my examples those societies where libertarian and other writers have alleged that there exists no exploitation. That is, I will challenge their assertions using their own illustrations. The other sample I will examine, the 'large sample', is more structured and comprehensive, as I shall explain.
The small sample In reading the allegations of libertarian writers, it is often difficult to maintain a focus on the issue of exploitation. This is because the libertarians themselves are often more interested in oppression, egalitarianism, coercion -the issues of human freedom which have always obsessed anarchists. These issues, however, are not unrelated to the issue of exploitation, as Lenin explains in The State and Revolution. For what is the purpose of oppression and coercion, he says (1960, pp3)8-13), other than to expedite exploitation. In a stratified class society, it is the state which enforces exploitive economic relations. But there are similar institutions in 'primitive' society, as we shall see.
In the following paragraphs, however, I will contest the allegations of the libertarians as formulated by themselves, rather than by reference to the rigid Marxist definition of exploitation, which I will develop as a preface to the large sample. For the sake of brevity, I will discuss only Kropotkin's assertions about equality among men, and only Diamond's assertions about the role of women. I will use only a few of their more prominent examples, a subjective sampling procedure which I do not recommend and which I will not use in the large sample to follow. I will then add a few paragraphs criticizing Hihdess and Hirst's use of Meillassoux's Gouro ethnography. The point I wish to make in this section, however, is that even the libertarians' own examples will not support their allegations, if we look at the whole ethnographic context of their sources.
In his book Mutual Aid, Kropotkin made many interrelated assertions about the internal character of 'primitive' societies, although the pervasive general characteristic that he asserts is that these societies were egalitarian, communistic, and democratic in their social and economic processes. The following phrases are used by Kropotkin to describe these tribes: (The Bushmen) used to hunt in common, and divided the spoil without any quarreling... (A Hottentot) cannot eat alone, and, however hungry, he calls those who pass by to share his food... (The Fuegians) share everything in common, and treat their old people very well, no quarrel disturbed the peace... (The Papuans) live under primitive communism, without any chief... They work in common ... like all savages, they are fond of dancing... When (an Eskimo) man has grown rich, he convokes the folk of his clan to a great festival, and after much eating, distributes among them all his fortune... The Dayaks usually have but one wife, and treat her well... They show great respect for their wives and are fond of their children... (Kropotkin, 1902, pp92-9, 109). And so on.
In evaluating Kropotkin's assertions, it seems initially very peculiar that he would have chosen the Dayaks of Borneo as an illustration of an egalitarian society, since his sources clearly describe them as having sultans, rajahs,, nobles, chiefs, slaves, and 'slave-debtors'.
Kropotkin's primary source, The Head-Hunters of Borneo by Carl Bock, describes this in great detail, even mentioning a 'festival' during which forty slave-debtors were tortured to death for the amusement of the gathered nobility (1882, p219).
As democrats, Kropotkin's example of the Hottentots does not fare much better. As early as 1695 Johannes Grevenbroek described the Hottentot chief (1933, p251) as 'armed with the power of life and death, and of administering justice ... when he emerges from his kraal, as from a palace, and allows himself to be seen in public, they strew rugs and mats before his feet to do him reverence.' I One of Kropotkin's sources, Kolben (Schapera, 1930, p280) also describes the operation of the Hottentot gerontocracy, which symbolized its dominance by urinating on young men and calling them women. They also forced young men to undergo ritual mutilation. Grevenbroek reports (1933, p209): 'Should any of our Hottentots refuse to subject his male members to the sacrificial blade or lancet or operating knife of the priest, preferring to preserve his genital organs perfect in the shape and number provided by nature rather than submit himself to agonies of pain and partial castration, this enemy of all amputation is insulted .. , shut out from all fellowship and inheritance, and shunned as if blasted by the lightning of heaven. ' Among the Hottentots, according to the same sources cited by Kropotkin, we find such exploitive institutions as forced labour, rank privileges, 'extortion' by chiefs, and primogenitive privileges . as well.
One of Kropotkin's other illustrations, the Australian aborigines, provides probably the most extreme example of gerontocracy among tribal peoples. The men of forty and over monopolized wives and food, swappingdaughters as nubile marriage partners. Like the Hottentots, they subjected young men (and women) to a long series of ritual mutilationscircumcision, subincision, scarification, all of which served to reinforce their political control. Kropotkin's sources, Fison and Howitt, were well acquainted with Australian gerontocracy.
In Kamilaroi and Kurnai (1880, p355), they wrote: 'It is worthwhile to consider what is the nature of the monopoly and by whom exercised. It is the monopoly of women in partial exclusion of the other clansmen... It is exercised by the elder men to the exclusion of the younger men... The perpetuation of this monopoly is encouraged by those interested in it having sisters or daughters to exchange with each other for wives, and is aided by the custom of betrothal while girls are even mere infants. ' We will return to the example of Australian aborigines in the large sample.
Turning now to Diamond, we should first examine his comprehensive ethnological assertions about those societies which he unblushingly labels as 'primitive'. In a 1971 essay for the Partisan Review, Diamond generalized about such societies (1971, p176), apparently the same fifteen societies he mentioned in an earlier (1968) essay: 'There is a predominantly natural division of labour, the person engages in a variety of tasks, and no significant disparity between mental and manual labour exists. ' As he states further, there are no 'exploitive political structures'. Let us examine the validity of this assertion for the Hottentots and Australian aborigines, as well as for the Plains Indians, all of which are used by Diamond as illustrations of 'primitive' societies.
Concerning the general condition of Hottentot women, we have first of all, the words of Kropotkin's source, Johannes Grevenbroek (1933, p195): 'Their women, who are wonderfully complaisant, and no less chaste, they abuse like cattle or slaves, making them bear loads upon their backs, and they keep them under a harsh and rigid discipline. From ankle to knee their legs are bound with a close series of loops and thongs, and on no other ground and for no other reason, as it seems to me, than that they may be deprived of the opportunity of running away.' As for the Hottentot 'division of labour', it can scarely be called that, since the women did nearly everything that can be classified as 'labour'. According to Olfert Dapper, another early (1668) Dutch observer, the women gathered vegetables for the 'daily provender' while carrying all the children along, 'to the greatest hindrance of their movements' (Dapper, 1933, p55). William Ten Rhyne, in 1686, adds that the women also did the milking, made butter, did all the cooking, made pots, and built the huts (Ten Rhyne, 1933, p129). He adds that 'the men look after the huts and the herds, or else are occupied in war'.
One might ask what reward Hottentot women gained for all their work. Ten Rhyne says (1933, p125) that 'they are not allowed beef, nor fresh milk, but mutton occasionally. So much do they despise the women. ' I From Grevenbroek we learn (1933, p263) that 'the daughters are forever debarred from inheriting from a father of brother, or from any other source. And indeed by the most ancient law of nations they are regarded as the end of one family and the beginning of another, and are made of small account.' Australian aborigine women had an even tougher go. A recent survey of early ethnographies indicates (Josselyn Moore, 1974) that Australian women supplied up to 80% of the food needed by the group, while doing all the child care. Their rewards and privileges were about the same as those of Hottentot women. They were pawns for the marriage machinations of the gerontocracy, as mentioned before, they were subject to genital mutilation, like the young men, and they were put to death if they pried into any ritual secrets of the men. According to Walter Roth, the 19th-century Australian who was designated 'Protector of Aboriginals', an unfaithful wife could be maimed by having her hamstrings cut, could be gang-raped, or flayed alive, while her lover was only exiled for a few months (Roth, 1906, p6).
A woman could even be called upon to forfeit her life for a murder committed by her brother He adds, 'The husband has the right to loan, exchange, sell or divorce his wife, who has no reciprocal powers: he can kill her if he likes... ' I Speaking of rape as a means of control, we can now consider the case of the Plains Indians of North America. We learn from George Bird Grinnell, who is quoted by Diamond (1974, p156), that a Cheyenne man could put his wife 'on the prairie' if she committed adultery or was disobedient, or simply to change his luck in war or hunting (Llewellyn and Hoebel, 1941, pp202-11;Grinnell, 1962 and1915).
A woman put 'on the prairie' was gang-raped by the members of her husband's soldier society, and sometimes beaten and killed. Women treated in this way were 'bad wives', women who did not live up to their responsibilities in Cheyenne society. So what were these responsibilities ?
Red Eagle, an informant for Biren Bonnerjea in 1934, described the . traditional division of labour as follows (Bonnerjea, 1935, pp135-6): Tlowadays since the white people have come to our country a man does all the hard work, but it was different before that; a woman used to do all the housework. When we moved from one place to another as we did very frequently -she would take down the tipis, carry . them and all other household things to the place we were going, and when we arrived at our destination, she would not only set up the tipi but see to all other household affairs, such as cooking, gathering wood, and so on. She went into the woods, cut down trees, and brought the wood on her back to the tipi. A man could not do these things because others would laugh at him if he did so; these were women's work. A man did nothing else but look after the horses... He would also go hunting, procure game, and when needed, go to war.' And if a Cheyenne woman was not enthusiastic about her wifely role, she could be beaten, unilaterally divorced, gang-raped, or even put to death! My last example from the 'small sample' is the Gouro of Africa's Ivory Coast, as described by Meillassoux (1964). This monograph is important both because it is the basis of Terray's Marxism and 'Primitive' Societies (1972), and because it serves Hindess and Hirst as a prominent example of a non-exploitive society (1975, pp45 -56) .
To begin with, we should note that, at a theoretical level, Hindess and Hirst seem to be of two minds regarding the role of 'exploitation' as a definitional quality of 'primitive communism'. Before discussing the Gouro they say (1975, p43): 'Primitive communism on the other hand is characterized by a very limited development of productive forces and a limited division of labour. If there are no classes there is no surplus-product sufficient to maintain a class of nonlabourers, together with their unproductive functionaries. ' I But on the other hand they later define the 'primitive communist mode of production' in terms of 'the determinate mode of appropriation of surplus-labour... ' (1975, p59). So do they say that exploitation exists among the Gouro or not? I believe they say not, because of two passages. First they say, generalizing about 'primitive communist' societies, 'The distribution of the product follows ideological criteria. ' Within the context of their discussion, they are clearly implying that if an individual voluntarily gives up surplus product for an 'ideological reason', where there is no coercion, then there is no exploitation.
More specifically, concerning the Gouro, they give us the following instructive passage (1975, p67): 'The control by the elders over the conditions of labour and over certain of the conditions of reproduction of the productive community is a necessary effect of the dominance of the complex redistribution variant of the primitive communist mechanism of appropriation of surplus-labour. Thus even if the elders in a lineage society perform little or no productive labour they do not necessarily constitute a class. Exploitation, in the sense of the appropriation of surplus-labour by a class, cannot be deduced from the co-ordinating and regulating position of the elders.' If I may paraphrase, Hindess and Hirst are telling us that the control by elders is necessary, and that this appropriation of surplus is justified because they do the work of coordinating the productive process. But is this not always what ruling groups tell us -that their intelligence and insight, their mental labour, is essential to the system? And it is also a regular occurrence that rulers can convince the workers of this, so that workers, for 'ideological' reasons, voluntarily give up a portion of their production. But whether there is a slave mentality present or not, when a non-worker consumes the production of workers, that is exploitation J And if coercion is used only occasionally, it still constitutes a constant threat.
-Looking at Meillassoux's ethnography in the original, rather than considering only Hindess and Hirst's excerpts, we can see more clearly the several kinds of exploitation and oppression which are an integral part of Gouro life. Chapter 5 contains a detailed description of the 'hierarchie masculin_e' which dominates Gouro society. Elsewhere, we find that women do not own the fields or the grain they grow (p141), and that they are delegated to the dullest, dirtiest work which must be done in the Gouro economy (pp209-11). Similarly, junior males work for their seniors, and most have no hope of ever rising to the position of vieillard so that they are never repaid as Generally speaking, I believe that those who have made libertarian assertions about life in 'primitive communist' societies have been at least charitable (Hsu, 1968, p35) and at worst dishonest in presenting their ethnographic data. Also, the readers of their arguments have so far had no assurance that the libertarians were approaching the data with any scientific safeguards against a thoroughgoing subjectivism in selecting only those examples and illustrations which might support their allegations.
In discussing the small sample, I have tried to show that even the libertarians' own examples will not support their arguments. In the analysis of a larger sample, to follow, I will show that approaching the data with a considered rigour not only answers some questions with authority, but also raises new questions of utmost importance to students of social evolution. In fact, as I shall show, I believe that there are a few examples of egalitarian societies, but they are not the ones offered us by the libertarians.

The large sample
The definition of exploitation that I am using in this analysis is the . one developed by Marx, from which recent theorists have not deviated significantly for a century. Marx gives us several lengthy definitions in Capital, both for degree of exploitation, surplus labour/necessary labour, and for absolute exploitation, which is the amount of labourvalue expropriated from the producer (Marx, 1906, pp235-353).
Mandel also provides us with a good discussion of exploitation, as well as a succinct definitionexploitation is a 'grabbing by one part of human society of the social surplus product which has been produced by the rest of this same society' (Mandel, 1962, p89).
To this definition I should add, and make explicit in response to Hindess and Hirst, that exploitation need not create a class of entirely leisured or otherwise unproductive people. Partial exploitation is also quite usual, especially among tribal peoples. That is, if one group in a society labours for an average of four hours per day, but consumes the production of ten hours' labour per day, while a more productive group labours for twelve hours per day, but consumes the production of only six hours, then exploitation, though partial, is still taking place. Marx discusses this at great length in Capital, although he does not use the term 'partial exploitation'.
In examining the ethnographic data, I will only be concerned with absolute exploitation. The reason for this is that there are so few ethnographies which contain enough detail about labour-time to calculate a ratio or degree or rate of exploitation. Mostly what we find are general statements about groups of producers (women, cohorts, clansmen) who must give up part of their product, their 'surplus', to a dominant group. Often this takes the form of some exploiters for having been exploited in their youth (pp172-3). kind of 'reciprocal' relationship, a situation which could more ' accurately be described as 'unequal exchange' -the producer , exchanging his services or material product for some kind of ethereal non-product, a cure, a privilege, protection, etc. Often, too, the unequal exchange is legitimated by ideologies or taboos reinforcing the unequal division of labour -for example, men are 'religiously forbidden' to do certain kinds of work (Oberg, 1937, p83;Tanner, 1944, p685), women are 'forbidden' certain foods and privileges (Merker, 1910, p251;Lips, 1947, p423). So this is the kind of data we are looking for in the ethnographies.
The societies in the ethnographic 'large sample' are selected to represent that stage of human evolution before horticulture, before stratification, and before the development of the state. I employ the principle of uniformitarianism in selecting recent societies to represent prehistoric ones, asserting that they are analogous unless there is some material archaeological reason for saying they are notsome artifact, architectural feature, or settlement pattern which shows that Magdaleneans, for example, have no modern analogue. Specifically, I intend that the societies in the sample should represent the period from the late Middle Paleolithic to the middle Upper Paleolithic, roughly 50,000 -10,000 BC in Europe and the Near East. This is the period after Neandertal Man, but before intensive horticulture had developed.
To select societies for analysis, I began with the hunting and gathering societies of George Murdock's World Ethnographic Sample, a sample which was developed, in part, to avoid the dilemmas of Galton's problem. * Following Lee, I then selected from the sample those societies which are listed as having a 100% dependency on hunting, gathering, and fishing (Lee, 1968, pp41-2). Then I selected from that list the societies which are represented in the Human Relations Area Files (HRAF). I did this so that my sample and analysis could be easily replicated and checked by other researchers.
To enter the files, I used two subject categories which I hoped would give me some measure of the exploitation of women by men, sexual exploitation, and the exploitation of men by other men, what Morton Fried might call 'exploitation by rank' (Fried, 1967, chapter 4). In the Outline of Cultural Materials (Nmrdock et, al. , 1961), category 462 (p57) is entitled 'Division of Labor by Sex' and category 561 (p76) is called 'Age Stratification'. Under topic 462 I hoped to find generalizations about relative amounts of productive labour performed by men and women, and under topic 561 I hoped to . discover types of exploitation among men short of outright social class distinctions, which is category 565. I later decided that * 'Galton's problem' is how to distinguish between diffusion and independent invention among neighbouring societies. See Kroeber category 462 was more successful than 561 for my purposes, but I shall return to this matter in my discussion of results.
I began with a list of 19 societies (see Table 1), but I had to throw out two North American entries, the Tubatulabal and the Yokuts, for lack of information. As for the rest, I discovered that I had data about sexual exploitation for all 17 of them, and about exploitation or oppression by rank for nine of them. I soon found, too, that I had to discriminate between strong absolute statements, such as 'It appeared to me that in marriage the men get the best of the bargain' (Batchelor 1927, p15) and milder evidence, such as long lists of women's work and short lists of men's work. Therefore the categories of Table 1 appear as they do. and also Cooper's (1917) generalizations contradict the observations of themselves and others has led me to disregard the Ona and the Yahgan for present purposes. Otherwise the results are as in the table.
For lack of space, I am not listing here the complete bibliographic references for the several hundred ethnographic sources I have consulted. Instead, I include in an appendix the Human Relations Aiea Files codes for the references, followed by the page numbers of the salient information. Lack of space also prevents me from quoting all the relevant passages, which I would like to do, but interested researchers can consult the HRAF to replicate and evaluate my findings.

Discussion
Generally speaking, my survey has shown that 11 of the 15 hunting and gathering societies in the sample exhibit sexual exploitation. There is no example of a society where the women exploit the men. If egalitarian relationships were 'normal', one would expect as many deviations toward female dominance as toward male dominance, but this is not the case.
At this point, it is useful to examine the general characteristics of the societies in the sample which exhibit sexual exploitation, bearing in mind that we hope these societies are analogous to societies in the Paleolithic. First of all, we should note that, without exception, the four 'agalitarian' societies occupy ecological zones -the Andaman Islands, the Arctic coast, the Great Basin, and the Amazon forestwhich were not inhabited until after the zones represented by the exploitive societies were fully occupied. Logically, this would imply either that the marginal zones have encouraged a development of a non-exploitative division of labour, or that egalitarian society persisted in marginal areas after it had disappeared in ecological zones where production was more efficient. Let us look at these two models in more detail.
Model 1 suggests that human society has been sexually exploitative as long as there have been hominids, since the late Miocene, and that egalitarian society is a recent adaptation to the special demands of marginal zones. Supporting evidence for this model from the field of primatology is the nascent exploitation of females present among 'proto-hominid' analogues, such as baboons and macaques, where males displace females in prime feeding spots which the females have found, or snatch food away which has been prepared by females (Itani 1958, Altmann andAltmann 1970). That is, there is a suggestion of evolutionary continuity from this kind of nascent sexual exploitation to the institutionalized exploitation of hominid females in the Pleistocene.
Concerning Model 1, I would further suggest that there are special but different material reasons, or 'objective conditions', which encourage non-exploitive sexual relations among the Andamanese, Copper Eskimos, Washo and Siriono. With the Eskimos and Siriono, there is a scarcity of food available for women to collect, so that they can produce no surplus to be expropriated. The ecological situation of the Washo is such that there is very little division of labour, and therefore less opportunity for sexual exploitation. In the Andaman Islands, the emphasis is apparently on the exploitation of young men by older men, the young men being the most productive members of that society because of their pig-hunting.
Model 2 suggests as a valid generalization that the labour demands of Paleolithic production were so heavy that there was no surplus, and no exploitation, and that this lack of surplus has continued among modern societies in marginal areas. Ethnographic evidence contradicts this, however, since the conclusion of Lee and others is that the vast majority of hunters and gatherers (from which I would except . the Siriono and the Copper Eskimos) tend to be very efficient in their production, and they produce a surplus which they 'consume' as leisure time (Lee and Devore 1968, pp83 -95). Other species of hominids also exhibit efficient ecological adaptation, and likewise do not spend much time, relatively, in looking for and preparing food .
Model 2 can be saved, however, by asserting that egalitarian 'primitive communist' societies of the Paleolithic were absolutely unique and that the societies in the first two columns of Table 1 are not comparable to them. But what material evidence is there for marking a difference between these societies and Paleolithic societies, and alleging the existence of a unique stage of social evolution for which there are no modern analogues? How far back in time must we go to find societies whose material remains (tools, settlement patterns, artwork) differ significantly from recent 'stone age' peoples? We must go to the Chelle-Acheulean, at least, and perhaps before that if we count Australian aborigines as having an 'Acheulean' stone technology (Gould 1969).
Clearly, I prefer Model 1, because it does not require positive evidence which we don't have. Also, I believe that Model 2 is the implicit model of the libertarians, especially including Kropotkin, Diamond, Cantine, and Hindess and Hirst. It has no supporting material evidence, which is why the libertarians reject the rules of evidence. Agreement with Model 2, I submit, does not depend on examining evidence, but on arousing sentiments endorsing a 'Golden Age' as a stage in human history, after the manner of the Greeks and medieval Christian mystics (Lovejoy, 1971).
Concerning this 'Golden Age' character of some evolutionary theories, I would call particular attention to the excellent essays included in Rayna Reiter's anthology Toward an Anthropology of Women (1975). Putting my assertions within the context of the issues raised in that book, I would agree with Gough and Leacock that we have no evidence for the prior existence of 'matriarchal' society, as some 'Golden Age' feminists have alleged. But this is not to say that male dominance, even if it has always existed and even if it is still universal, is either desirable or inevitable. Like other forms of oppression and exploitation which appear to have a biological basis, sexism is something to be overcome by practical struggle.
Turning now to the subject of the exploitation of men by men, we can see that category 561 of the HRAF does not give us good entry into the relevant issues. But we can discern several apparent patterns of rank privilege which suggest comprehensive theories of evolution, especially concerning the question of how a society evolves from simple sexual exploitation to more complicated patterns.
Archaeologists and biologists have perhaps used the principle best .
among modern researchers, but the basic idea, as applied to the anthropological problem of social evolution, is that simple social patterns appear before their permutations. For example, a simple clan system is logically prior to a system which involves clans plus phratries or moieties. A simple pairing of spouses is logically antecedent to a pairing with a ceremony and a brideprice. The principle is, of course, debatable, and no doubt there are exceptions to it, but I believe it is fundamentally sound.
In many of the societies of the large sample we observe sexual exploitation with no operative gerontocracy (Bushmen, Gilyak, Montagnais), and in some we can see a strong gerontocracy as well as sexual exploitation (Tlingit, Aranda, Tiwi). But there is only one society, the Andamanese, where there is gerontocracy without sexual exploitation. The principle of seriation, then, suggests that simple sexual exploitation is historically prior to the development of gerontocracy. That is, if we assume that our recent ethnographic sample accurately represents prehistoric societies in various stages of evolutionary developmentan anthropological application of the principle of uniformitarianism -then we should be able to seriate them logically in accordance with their internal structure, just as if we were seriating pottery, or as Ribeiro has seriated civilizations. Regarding our own ethnographic sample again, it is important to note next that the gerontocratic societies -Murngin, Dorobo, Tlingit, Tiwi and perhaps Pygmies, do not institutionalize the exploitation of men, as we have defined it. Rather, these societies institutionalize the exclusion of some men from the privilege of exploiting women! That is the older men, especially in the Australian examples, control access to women and thereby increase their absolute exploitation, the amount of production they receive from women, without increasing the rate of exploitation of individual women. Hart and Pilling, writing about the Tiwi, even maintain that polygyny increases the efficiency of production (hard and Pilling 1960, p36): 'This production unit to reach maximum economic efficiency, required the vast majority of all females to be concentrated in the households of a very small number of husbands; namely the very oldest men. As a necessary correlate, men under twenty-eight had no wives at all and very few . men under forty had any wife except elderly and physically very unattractive widows. The efficient economic organization thus obviously created a moral and social problem -the problem of how to keep the unmarried young men away from the young women. ' I . At the ideological level, it is interesting to note how frequently the status of women has provided a model for defining the status of young men. Kolben tells us that Hottentot youths were urinated on by older men and called 'women'. Australian aborigine ceremonies are also extremely concerned with the femaleness of the male initiates (Spencer 1914, ch.3). Even in our own society, 'woman' is a common sexist epithet for a clumsy young man. Putting some evolutionary dynamics on these observations, then, we can theorize that the subjugation of women provides a cultural model for the subjugation of men. Intertribally, it is significant that the Iroquois followed this same example and required the defeated Delawares to pay tribute and wear dresses (Miller 1974).
In the Paleolithic, then, with young men deprived of women and women's production, the stage was set for the exploitation of men, just as the rigid stratification of feudal society set the stage for the proletarianization of the feudal peasantry. As long as societies were in a hunting and gathering stage, however, the young men had very little to offer by way of production to be expropriated as 'surplus'.
Our ethnographies indicate that men in most such societies cannot even feed themselves, much less produce a surplus to exchange with an older man for a wife. Young men in such societies are dependent on their family of orientation for food and shelter until the time that they are fully initiated and can marry.
The domestication of plants and animals, however, changed all that. In sedentary communities, a young man could offer the surplus horticultural production of his family of orientation as brideprice, or he could offer himself for groom service. Groom service could be institutionalized in sedentary communities because it was now possible for an older man to accumulate other people's labour in the form of houses, fences, non-portable tools, etc.
The data I have examined does not 'prove' the evolutionary theory I have just outlined, the data has only suggested the theory. But it is an attractive, comprehensive, and concise theory and, given the validity of the principles of uniformitarianism and seriation, it is possible to prove it using recent ethnographic data.
Another perspective that must soon be taken on social evolutionary theory is an archaeological one. In reading the evolutionary theories of ethnologists such as Diamond, Fried, or Service (1975), I often find myself asking in frustration: What societies are they talking about -where and how long ago? Do they mean Paleo-Indians, the residents of the Upper Cave at Choukoutien, Dolni Vestonice, Molodova, Shanidar -who exactly? Once again the pure theorists have not been very specific, and they are probably not willing to be. Terry Counihan has given us a neat phrase for condemning this lack of specificity, this sort of 'weak-kneed attempt to avoid criticism' (Counihan 1976, p77).
Among archaeologists, however, the approach I am suggesting is called 'ethnographic analogy', and it is a regular part even of bourgeois archaeological synthesis (Binford and Binford, 1968). Many archaeologists regularly look at recent societies for clues in interpreting archaeological evidence. For example, to get a significant interpretation of Danubian longhouses (Clark 1969, pp126-47) it is considered useful to consult Iroquois longhouses. And if one is interested in reconstructing Danubian social behaviour, an analysis of Iroquois ethnography is certainly relevant. This kind of empirical approach I find refreshing in contrast to the evolutionary pure theorists' customary appeals to libertarian sentiment or to Marxist orthodoxy as the final arbiters of theoretical dispute.
Conclusion I hope that I have left any number of loose ends dangling in the analysis just completed, and that my suggestions and assertions will provoke the comments and criticisms of my colleagues. For dialectics is not merely the process of human evolution, it is also the process of scientific inquiry into human evolution. Too often, Marxists pursue their debates from positions of unassailable orthodoxy, without ever finding a test situation or a natural experiment to resolve the debate.
It is my opinion that a commitment to analyzing ethnographic data empirically will enable us to resolve certain issues and pursue our researches with collective assumptions as well as collective goals. But unless we share criteria for evaluating our theories, we are condemned to perpetuate our narrow orthodoxies. It is for this reason that I have been particularly critical of pure theory, for it allows the free play of intuition, with no possibility of scientific progress. Pure theorists can locate themselves anywhere in the theoretical world that they care to, for whatever reasons, and remain immune from criticism.
Whatever the dangers of the empirical approach, it at least gives us a common body of data for analysis, and it offers the possibility of collective progress toward elaborating and sharpening our evolutionary theories. And whatever the merits of my particular theories about the roles of men and women in hunting-gathering societies, I hope that I have at least shown how such questions can be considered within the framework of an empirical, materialist, rational social science.