8 pages ANTH 165 paper

Origins_of_Agriculture_----_The_Earliest_Farming_Demography_As_Cause_And_Consequence_.pdf

Home >Social Science homework help >8 pages ANTH 165 paper

The Earliest Farming: Demography as Cause and Consequence

BENNET BRONSON

INTRODUCTION: DENSITY AND HUSBANDRY

I propose to discuss here the association between agriculture and popu- lation size in very early times in the millennia that separate the Palaeolithic from the first appearance of cities and states. That some such association exists, and that the association is in part causal, can hardly be doubted. But its precise nature remains elusive, the subject of many vexed and convoluted debates from Malthus' and Ricardo's time down to the present day.

Can the apparent explosion of population in the time of the early states be explained entirely by the Neolithic Revolution? Obviously not: the revolution precedes the explosion by several thousand years. Can the one partially explain the other, by saying that food production (i.e. agriculture) is a PRE-CONDITION though not a sufficient explanation for demographic expansion? Perhaps. Such explanations are a staple of elementary text- books in anthropology (and nowadays even history) and may well be limitedly valid. But they suffer from a number of practical defects. They are no longer as productive of important new hypotheses and stimulating research as they were in the 1930's and 1940's .Moreover, they are not so subtle as to inspire anyone with respectful surprise. Such defects may contribute to the relative eclipse of agriculture-as-cause formulations among modern theorists.

On the other hand, the idea that the causality is reversed, with expansion in population supplying the motive force behind agricultural progress, is presently enjoying a modest vogue. One of the principal proponents of this position is the agricultural economist Ester Boserup (1965), who has

Origins of Agriculture, edited by Charles A. Reed, De Gruyter, Inc., 1978. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/sjsu/detail.action?docID=3044546. Created from sjsu on 2021-01-16 19:43:25.

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

2 4 BENNET BRONSON

elaborated a typology of agricultural stages arranged in order of in- creasing intensity (by wnich she means increased frequency of land use) which evolve from one to another under the influence of the exogenous variable, population density. An attractive feature of the Boserup formulation is that it does not simply assume that more intensive farming appears because increasingly dense populations need it but instead provides a mechanism to explain the changeover. This mechanism depends on the reasonable assumptions that (1) the technology for some kind of agricultural intensification is readily available to most peoples, and (2) that the average farmer is inhibited from employing this tech- nology by the fact that the more intensive systems are also the most labor demanding in terms of output per man-hour. Hence, all agricultural regimes, even if intrinsically quite intensifiable, will remain in the most extensive state possible until the farmers are forced to change through the pressure of population and an increasing scarcity of land.

This straightforward but novel view of agricultural evolution has been received by some (P. Smith and Young 1972; P. Smith 1972) with enthusiasm and has even been extended back into the prehistoric period (Cohen, this volume) on the grounds that hunting and gathering is still more economical of labor than the most extensive forms of true farming — thus, the very existence of agriculture is seen as a response to demo- graphic factors. Others (Bronson 1972) have questioned the applicability of the Boserup model as originally presented, pointing out the lack of empirical evidence for the central proposition that extensiveness and labor efficiency are really correlated and suggesting that there are strong theoretical reasons for doubting that agriculture actually did evolve along a single track as Boserup proposes.

However, the details of this and similar models of agricultural change are not immediately germane. My concern here is to discuss the abstract issue of demographic explanations of subsistence systems, with reference not to recent alterations in farming methods but to the origins of farming itself. Can it be said that population pressure is a sufficient, or necessary, or even plausible precondition to the Neolithic Revolution? More impor- tantly, is it an explanation? Do we gain anything in the way of theoretical rigor or predictive power by postulating a single demographic prime mover to explain all the manifold subsistence choices made by early man?

To answer these questions, I must redefine (or select among the defini- tions of) several concepts and reconstruct several models of subsistence economics and ancient demographics. Such concepts and models will be found to occupy the major portion of this paper.

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 25

DEFINITIONS AND RESTRICTIONS

Several terms in the following pages are likely to cause confusion unless the meanings assigned to them here are described. Among these are "efficiency," "intensiveness," and "permanence" as applied to subsistence regimes, and the more general terms, "agriculture," "cultivation," and "domestication."

"Efficiency" in this paper can be understood in two ways. When prefixed by the word, "labor," it refers to the success of a given subsistence method in minimizing the number of man-hours required for each unit of production. Prefixed by the word, "land," on the other hand, it refers to a related but sometimes opposed kind of success, the extent to which the subsistence method minimizes the quantity of land required for each unit of production. One could also evaluate efficiency according to other criteria (for instance, by social utility or effectiveness of capital utilization) but land- and labor-efficiency are what will mainly concern us here.

The concepts of "intensiveness" and "permanence" also have potential for causing misunderstanding. Both are applied to land use, but while the former is essentially a synonym of land-efficiency, the latter refers only to the relative frequency with which a plot of land is exploited. We know of subsistence regimes which are at once permanent and extensive (e.g. medieval plow farming — see Homans 1970 and Slicher van Bath 1963) and others which are intensive in spite of their impermanence (e.g. Ibo swiddening [Morgan 1955] which has a higher carrying capacity than many permanent regimes). The confusion between intensiveness and permanence is built into a good deal of the traditional terminology with which non-Western agriculture is described. Such terms as "shifting agriculture" have therefore been used sparingly here. They impute an excessive importance to simple permanence of field location, a datum which has only a limited relevance to demographic questions. We are far more interested in the land- and labor-efficiency of a regime than in whether its fields are in the same location from one year to the next.

"Agriculture" and "horticulture" are here treated as synonymous. Although some students of the subject have seen an evolutionary gap between agris and hortus, regarding one as an attribute of advanced societies and the other as intrinsically primitive, such an attitude is faintly ethnocentric. Many demographically successful modern peoples (e.g. the Javanese, Terra 1954) gain a major portion of their livelihoods from gardens, from plots of land too small and messy to be called, by our clean-cropping Western standards, "fields." Yet these plots may be centrally important from an economic point of view and, moreover, may be cultivated with a very high degree of skill, if not of hardware tech-

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

2 6 BENNET BRONSON

nology. My own feeling is that the owners of these plots are as much agriculturalists as any Ukrainian peasant or Nebraskan factory farmer. To treat horticulture as an essentially distinct system is misleading; it is equally misleading to talk as though horticulture is inferior from the standpoint of subsistence or necessarily early in an evolutionary sense. Some ancient farmers — wheat growers in the Near East, for example — undoubtedly possessed plots a Westerner would call a field as soon as they began to cultivate a staple crop. But others — root-croppers in South America and Africa, mixed farmers in eastern Asia — began with small gardenlike plots and have continued to depend on them down to the present day.

Agriculture is not, on the other hand, used synonymously with "culti- vation," nor does either term necessarily mean "domestication." In the following pages, the term agriculture is reserved for contexts of sub- stantial dependence on plants grown by humans, while cultivation denotes only that a useful species has been deliberately caused to repro- duce by man. All agriculturalists are indeed cultivators, but a cultivator need not always be an agriculturalist; he may be just a gatherer (or a factory worker) who occasionally puts a seed or cutting into the ground with the expectation of using the result. This distinction is unorthodox but useful. One consequence of it is that cultivation is seen to be more elementary and perhaps older than agriculture, a theme which will be expanded in a later section.

"Domestication" is used here in the strictly biological rather than partly cultural sense, referring not to taming, growing or other patterns of regular human utilization but instead to the genetic effects that some- times accompany that utilization. The term is arbitrarily restricted to effects produced specifically by human use. Even though one can easily conceive that plants might become adapted to, and undergo genotypic changes because of, preferential utilization by cows, and even though one might plausibly call a plant adapted for growth in a field favored by cow manure a "bovine domesticate," our present interest is focused on the causal interactions between the natural environment and man.

Even within this limited sphere, under other circumstances it might be necessary to make a still narrower restriction in the meaning of domestication, confining it to meaning the effects produced specifically by cultivation and thus excluding the inherited phenotypic changes that frequently have attended the adaptation of weeds to human, but not necessarily subsistence-connected, habitats. Luckily, distinguishing between weeds and useful domesticates is not necessary to what follows. All that matters is to establish a conceptual separation between cultivating

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 27

and domesticating and to observe that neither can be assumed invariably to accompany the other. Cultivation (and even agriculture) without domestication is perfectly conceivable; there is no reason why even repeated cropping should necessarily always produce a phenotypically distinctive population. Likewise, a process much like domestication, and perhaps cytologically and morphologically indistinguishable from it, can be assumed to have occurred in numerous species of weed and perhaps in some selectively utilized wild species as well; hence, quasi domestication without cultivation is also possible.

The conceptual distinction thus has a practical consequence. If we have no evidence but remains of plants, we cannot demonstrate conclusively that cultivation did or did not exist. The plant remains can of course indicate probabilities. I myself am inclined to feel that the abundant presence of domesticated characteristics yields a fairly strong presump- tion of cultivation and that their absence is indecisive, indicating no more than that the site in question MAY have been inhabited by pure gatherers. But in either case, plant remains by themselves are insufficient. Acceptable proof or disproof of cultivation requires the use of several additional lines of evidence.

A last comment should be made relative to cultivation and agriculture. Here, both terms are confined to plant growing. Animal husbandry is indeed an integral part of many agricultural systems and the histories of the domestication of plants and animals in many areas are inextricably intertwined. Nonetheless, I have excluded animals from the following discussion. The reason is simple: I have not yet sorted out in my own mind how herding is related to population growth or whether, except insofar as traction power and manure are necessary to agriculture and scavengers to public health, it is related to population growth at all. Certainly, herding seems an inefficient way of getting protein and a most wasteful source of calories. In some environments it may be adaptive enough, but in others the idea of replacing the efficient wild fauna with domesticated animals seems demographic madness. If a causal relation- ship exists between population growth and herding, and if the adoption of herding is not due to quite different motives, that relationship is subtle and complex indeed.

With these distinctions and definitions in hand we can now return to the central subject, the association between population and agriculture. We must necessarily consider three sets of models before any conclusions are reached. The next three sections, accordingly, treat (1) the beginnings of cultivation, (2) the beginnings of staple agriculture, and (3) the history of population density.

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

2 8 BENNET BRONSON

BECOMING A CULTIVATOR

The proposition to be presented here is that the beginning of cultivation — that is, of the habit of deliberately growing useful plants — was neither a unique nor a revolutionary event. It probably happened repeatedly in different places, starting at a very early date. Its causes may have been comparatively trivial. And, for a period perhaps as long as ten or more millennia, it may have had few discernible social or genetic effects. The proposition is supported by the following arguments.

To begin with, cultivation is not in essence either a complex idea or one difficult to develop. True farming — committing one's resources to the establishment of an artificial ecosystem to yield a staple food supply — may be filled with subtle risks and calculations, but small-scale non- staple cultivating is elementary, so much so that it is not beyond the inventive reach of almost any human being. We can be quite sure that activities resembling cultivation go far back into the Palaeolithic. By the time a modest degree of intelligence had appeared in the human stock — certainly by the late Pleistocene if not before — extensive and in some cases massive interference with the habitat of certain selected species must have already begun. Even non-human predators (e.g. cows) are often observed to feed with discrimination, singling out a small number of species for special attention. But when the predators are intelligent and use fire, the potential for sustained, focused, and drastic selective pressure is clearly increased by several orders of magnitude. Through field fires lit by humans and intelligent concentration on selected food sources, numerous species must have been virtually exterminated long before the famous extinctions of big game during the terminal Pleistocene. Numerous others must have begun their adaptation to microhabitats influenced by humans such as refuse piles and fire clearings, and thus started to become quasi domesticates. It should be remembered that domestication as defined above is not necessarily a consequence of cultivation. Moreover a few species must have been deliberately favored by man. Many recent gatherers are reported to intervene extensively in the life cycles of wild species, going so far as to replant them (wild yams among the Andaman Islanders, wild rice among the Great Lakes Indians) or even to irrigate them (among the Paiute). Ancient gatherers surely were also given to this sort of intervention. One can easily imagine that a Neanderthaler had the foresight to spare a fruit tree growing near a regular camping spot, or that an Upper Palaeolithic sapiens sapiens had the intelligence to remove weeds from a bed of useful perennials.

It seems most realistic therefore to envision the process of human

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 29

adaptation in the late Pleistocene as forming a continuum of selective exploitation, intervention, near-cultivation and quasi domestication. Somewhere in this continuum the first act of deliberate cultivation must have occurred, without fanfare, or important consequences, or awareness that anything new had been done. The contemporaries of the pioneer among all cultivators were surely as aware as he or she that seeds sprout and planted cuttings become new plants. Accidental planting and subsequent utilization must already have occurred numberless times. The only new aspect of the situation was the element of deliberation, the deci- sion to plant a seed or cutting with the intention of using the result.

We may assume that this first of cultigens had the following charac- teristics: (1) it was of a kind necessary or strongly desirable in the eyes of a group with a rather simple lifestyle; (2) it was in short supply within collecting range of this group's usual camping places; (3) it was not a major staple — if it had been, then planting a few individual plants would not have solved the problem of scarcity while planting a whole field full would probably have seemed to the group a dubious investment of their labor; they could far more easily have moved to an entirely new area; and (4) the plant may have been perishable, or rare everywhere in the region, or distributed in what the ecologists call a "fine-grained" fashion: that is, spread evenly over the landscape rather than in widely separated but easily harvested patches. This last set of characteristics would make resupply difficult even if the group should resort to the strategy of detaching a large part of its labor force to concentrate on long range foraging expeditions. If the plant is hard enough to procure even under those conditions then its labor-cost will be unacceptably high. The group will have no choice but to do without or to learn to cultivate.

Under the assumptions that this protocrop was highly desirable, quantitatively unimportant in the everyday diet, locally scarce, and diffi- cult to keep in adequate supply even when areas outside the local zone were exploited, one might venture an a priori description of the plant. It should be native to a fine-grained environment (like a tropical forest) or to an environment of low species and individual density (like a desert). It should have an annual habit and other traits that will make it likely to die off under careless exploitation (unlike a fruit tree or a grass). And it should contain some substance rarer than standard proteins, fats, sugars, and starches — perhaps an ester flavoring, an alkaloid stimulant, a glycoside poison, a fiber, or a dye. The theoretically ideal protocrop would be a non-staple plant with several important potential uses, such as flax, hemp, areca nut, turmeric, or the fruit banana. And empirically speaking, it is of interest that plants with these qualities are quite often found

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

3 0 BENNET BRONSON

archaeologically in protoagricultural contexts — chile and agave in Mexico at Tamaulipas in the Infiernillo Phase (Mangelsdorf, MacNeish, and Willey 1964: 430) and at Tehuacän during the El Riego phase (C. E. Smith 1967: 232); nuts of Piper and areca in the lowest levels at Spirit Cave, Thailand (Gorman 1973: 100); and cotton and Lagenaria in early South America (Pickersgill and Heiser, this volume).

But detailed speculative models of this kind are a luxury at this early stage of prehistoric research. What matters more at present is to produce general models, and such a model can be abstracted from the preceding paragraphs. The probability of an early hunting-and-gathering group becoming cultivators is seemingly controlled by only four sets of factors: 1. Pre-existing technical knowledge — that is, familiarity with certain aspects of plant reproduction. 2. Sufficient rationality to be capable of acting for the sake of remotely rather than immediately anticipated gains. 3. A moderately strong locational constraint, which may be either positive or negative. It may be either (a) a focus of attraction, perhaps a natural resource that is difficult to transport and constantly used (e.g. a water source or a concentrated supply of a staple food) or a cultural resource with the same qualities (a defensible locale or, conceivably, a shrine); or (b) a circumscribing zone of negative attraction, rendered marginal by such factors as environmental poverty, climatic discomfort, military danger, or disease. 4. A botanical commodity which is both highly desirable and scarce, scarcity being defined in terms of the labor cost of collection when the collecting group is under a locational constraint.

Seen through the glass of such a model, the probability of early cultivation in any area might seem quite high. Certainly knowledge and rationality can be assumed to exist in some degree even in remote pre- historic times, and the coincidence of locational constraint with scarcity must be a well-nigh universal condition. We might therefore conclude without further ado that the inception of cultivation should itself be a near-universal. But first a few comments on the role of population density are in order.

The main effect of the model in this regard is to reduce the role of demographic pressure to that of one among several factors producing scarcity. Perhaps the commodity in question has become scarce simply because the increase of population has outrun the ability of the local habitat to maintain the commodity in steady supply. In such a case, demography is one of two producers of scarcity, the other being the always-necessary factor of locational constraint — if no constraint

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 31

exists, and the group is free to wander anywhere in search of what it needs then "scarcity" can hardly exist. But the commodity may be un- obtainable for reasons other than straightforward population growth. Perhaps a small and non-increasing population has eaten all the com- modity up over the years it has remained in a certain locality. Here, demographics remains a factor but in a rather less decisive way. Con- ceivably the commodity may never have existed in adequate quantities within foraging range of the place where the population is constrained to live. The population might have migrated to that place and brought their knowledge of the scarce plant with them, or might have acquired a taste for a previously unknown plant through chance discovery or trade. A case in point is the interest in and subsequent cultivation of tobacco among the Northwest Coast Indians during the eighteenth and nineteenth centuries. Tobacco can be said to have become scarce among the Kwakiutl as soon as they discovered its existence, but in "scarcity" of this kind demographics plays no role at all.

A last point to be considered is that resource scarcity (and for that matter, population pressure) is a highly subjective matter as far as causa- tion in human societies is concerned. Whether a commodity has really become scarce and whether it really is necessary to survival are not entirely relevant when we seek to explain actual human decisions and actions. As modern specialists on agricultural development have begun to emphasize (e.g. Found 1971) what counts most in subsistence decisions is PERCEPTION. If a technique is perceived to be laborious then it will be resisted even if, from the standpoint of an outside observer, it is con- venient and economical. And if a commodity is perceived to be scarce, even though it may in actuality be abundant enough, then appropriate action will be taken. Possibly the original cultivator decided to plant his crop because he wrongly evaluated the difficulty of finding the plant growing wild.

But if we ignore this problem of perception for the moment, we can arrive at four interim conclusions. First, cultivation of an elementary kind should be extremely old; there is no reason why it should not have come into being quite far back in the Pleistocene. Second, this rudimentary cultivation need not have had any decisive genetic effects on the plants involved. If only a few individuals were grown at once and especially if the parts of the plant utilized were not the flowers or the seeds, a proto- cultigen might be indistinguishable from its wild congeners. Third, early cultivation need not have had much effect on human populations. Perhaps it enabled a few groups to lead more comfortable lives and encouraged some to slow their wanderings, but it may not have con-

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

3 2 BENNET BRONSON

tributed directly to any kind of archaeologically discernible increase in population. And fourth, increase in density of population need not have played a decisive causal role. Although a plant may have occasionally come into cultivation as a response to demographically induced scarcity, there are many alternative routes to that result.

All these conclusions are of course predicated on the notion that cultivation and agriculture, a substantial dependence on cultivated plants, are quite distinct institutions. As will be seen in the following sections, the interconnections between staple agriculture and demography are of a rather different kind.

BECOMING A FARMER

At this point we should inquire why the appearance of staple crop farming was delayed so long. Even if we reject the almost unprovable possibility of a Pleistocene origin for casual plant tending, we must still account for the fact that full-scale dependence on agriculture lags a surprising distance behind the known beginnings of cultivation. In Mexico, Peru and Southeast Asia, although perhaps not in the Near East, this time-lag seems to amount to at least several thousand years. What is the reason for such a long delay?

Several explanations can be invented. One of the most attractive is a hypothesis based on Boserup's model (see above) of agricultural develop- ment — that just as "extensive" agriculture is less labor demanding and therefore preferable to "intensive" agriculture, so gathering is still more economical of labor than agriculture itself. There is even some empirical evidence for such a hypothesis. Sahlins (1972: 1-39) has pointed out that many hunters and gatherers, contrary to what once was generally believed, are comparatively affluent. Both the Hadza (Woodburn 1968, 1972) and the !Kung Bushmen (Lee 1972a, 1972b) are said by their ethnographers to lead an easy life, devoting no more than a few hours a day to subsistence activities even (in the case of the !Kung) in distinctly marginal environ- ments. Thus one can argue that the apparent reluctance of early gatherers and casual cultivators to convert to true farming may have been due to a simple lack of incentive. Before the appearance of the incentives of the later prehistoric period — denser populations, markets, perhaps govern- ment persuasion — remaining a gatherer may have been the economically rational course.

However, the labor-saving explanation is difficult to accept as a uni- versally applicable rule. As I have argued elsewhere (Bronson 1972), a

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 33

great many other factors enter into decisions concerning subsistence besides labor-efficiency: considerations of security, of prestige, of com- fort, of health. For instance, nomadic gathering usually seems to exact a rather high price in natural and induced mortality among the very young and very old. It also limits substantially the possibilities of owning weatherproof dwellings, of developing non-subsistence technologies, and of storing food against times of scarcity. One is not convinced that the desire to do as little work as possible will invariably offset such con- siderations as these. It is far from certain, in fact, that gathering is always less work than some kinds of farming. Numerous food-production regimes, both shifting and permanent, require no more than a few hours' work each day in order to keep a family in food; most known hunter- gatherers (including the Hadza and !Kung) work at least this much, particularly when the labor cost of trekking from camp to camp is counted in. The labor-efficiency of gathering is indeed a factor to be considered, but it is not adequate as a full explanation for the apparent fact that substantial dependence on cultivation appeared so tardily.

An alternative explanation is that time was required for productive and trustworthy staple crops to evolve, and that the delay in the appearance of fanning was thus due to a built-in lag in genetic possibility. But this explanation seems weak. Except for a few especially intractable species (perhaps maize), few staple crops can have needed more than a century or two of human attention to reach an adequate level of productivity.

A third explanation, which might be called the "naive-demographic" model, depends heavily on the idea that the development of farming was a straightforward adaptive response to the development of large, dense populations. Thus, it could be argued, true agriculture did not appear earlier simply because it was not needed until the time it did appear. But there are a number of serious objections to such a baldly eufunctional proposition, among them the fact that demographic development on a local scale is inherently too fast-moving to explain a series of events that extends over several millennia. As will be pointed out in a succeeding section, if demographic necessity were the only cause, agriculture would have appeared much more quickly than it did.

The fourth explanation that suggests itself has to do with the minimiza- tion of risk. When the casually cultivating hunter-gatherer turned to farming, he may not necessarily have had to work harder, and he may have obtained a number of benefits from the settled life that then was possible. But it is undeniable that he took a considerable gamble. He committed a substantial amount of labor to a course of action from which he could receive no immediate return. Indeed, in those days of pristine

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

3 4 BENNET BRONSON

farming when no one had successful agriculturalist neighbors to observe, he could have reasonably doubted that he would receive any return at all. Even nowadays crops frequently fail, and still more frequently return no profit on the labor and capital expended, in spite of some nine or ten millennia of agronomic experience. Back in the days when farming began, the risk must have seemed and been very great indeed. While other factors may have contributed, simple caution is an almost adequate explanation for the reluctance of early cultivators to engage in full-scale farming.

The problem that remains is to find a model to explain why agriculture came to exist at all, why men everywhere, perhaps through judicious use of infanticide, war, and other fertility-controlling measures, did not remain casually cultivating hunter-gatherers down to the present day. The model that seems most useful is described below.

In its most generalized form, this model has a good deal of similarity to the one presented in the preceding section for the probability of becoming a cultivator. Again one must postulate a locational constraint and a scarcity of an important commodity. But here the commodity must be essential rather than simply desirable — that is, a staple food. And the question becomes more acute of why the proto-farmers stayed put when faced with this scarcity rather than just moving on. The risk they took by staying and attempting to grow the commodity was, as has already been pointed out, considerable. We must therefore assume that the locational constraint was very strong.

A number of more detailed submodels can be generated by considering the possible nature of this constraint.

The first submodel is a classically simple one — an island or otherwise circumscribed environment from which, for reasons of military danger, epidemiology, or sheer physical impossibility, the inhabitants cannot migrate. Within such an area it is plausible that population densities will increase quite quickly beyond the point where a hunting and gathering way of life can be sustained. In later times, such densely populated enclaves have been observed often to produce strikingly land-intensive agricultural systems, even in the midst of regions where most subsistence is of a very extensive kind. Numerous examples of what Clark and Haswell (1967: 50) call "societies under siege" occur in East and West Africa, Central and Southeast Asia, the Pacific, and the New World (see also Bronson 1972: 216). Since these isolated enclaves are often rather idio- syncratic in terms of the intensive farming technologies they use (e.g. the Haya — Allan 1965), one concludes that many of these technologies have evolved in situ in response to the fact that no one could migrate out when land grew scarce. But if constraints on out-migration can thus

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 35

render inefficient farming systems efficient, then why could they not at an earlier date make a casual cultivator become a full-scale agricul- turalist? Many of the same constraining forces were as operative in the early Neolithic as in recent times. It seems plausible that they could have had similar effects.

A rather more complex submodel is generated when the constraining forces are considered to be centripetal and positive: when, for instance, a population is drawn to a given place by the abundant presence of a second staple commodity different in kind from the one which is becoming scarce. A fishing lagoon on an otherwise unproductive coast would meet these requirements, as would a water source in a generally waterless region. The attractions of an abundant supply of protein or water might easily counterbalance the disadvantages of a shortage of a starchy staple in the eyes of a hunting-and-gathering group, causing them to attempt to raise that staple rather than move on to another place.

It will be observed that this two-staple model is a generalized version of two well-known theories of the origin of food production. The idea that the first agriculturalists may have been fishermen was originally suggested by Sauer (1952: 23) and has been subsequently taken up by several more recent authorities (e.g. Adams 1966 : 40-41). The latter sources emphasize the importance of sedentarization as a factor in the decision to plant a staple crop; here, the conflict between two separate locationally fixed staples is assigned the central role. Sedentarization undoubtedly predisposes to agriculture but is not a necessary precondition in this model's terms. A conflict between the need for fish and the need for grain could result in the adoption of agriculture even in the absence of a settled life. The fishermen could use the vicinity of the lagoon only seasonally, planting a (necessarily pest-resistant) crop and then continuing on a gathering circuit for the remainder of the year.

The water-source-centered version of the model rather resembles the somewhat discredited "oasis theory" of agricultural origins, whereby the first domestication was assumed to have occurred within oases isolated by increasing regional desiccation. The main difference between this model and the oasis theory in its more highly elaborated form (e.g. Peake 1928) lies in the way the future farmers are assumed to get into the oasis in the first place. While Peake and Pumpelly postulated that the farmers had to be trapped there by a vast climatic change, here no catastrophe is necessary. Many nomadic pre-farming groups must have stayed within oases for long enough to consume most of the food supply inside the watered area and within the exploitable zone surrounding the oasis. That the group would always attempt to farm rather than move on

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

3 6 BENNET BRONSON

to another oasis is of course unlikely. But, given a sufficient scarcity of water elsewhere and perhaps a reluctance to split the group into smaller units, it is entirely plausible that agriculture would sometimes have been the result.

A last submodel worth considering is the most diffuse and indeter- minate of all. Let us assume that most of the conditions laid down previously do not always hold — that under some circumstances agri- culture is less risky and easier than collecting, that no locational con- straint exists, and that the desired staple commodity, although in short supply, is not necessary to survival. A wandering band of gatherers in an almost deserted rainforest will serve as an example. What is to keep them from cutting down a few trees and planting a moderately large crop of, say, manioc? The labor investment need not have been excessive. If they girdled the trees they would have had to do little cutting and if the forest was deserted, and hence primary, the undergrowth would have been minimal. A quarter hectare of cleared area might have needed no more than two weeks' work and could have produced, in the case of manioc, enough calories to live on for a year. Moreover, since manioc has few natural enemies, the members of the band would not have been obliged to wait around until harvest time; they could have gone off and gathered wild foods elsewhere in the forest while the crop took care of itself. The band thus took no risk, made little commitment, and enjoyed a greatly increased level of security — if the supply of other staples failed, it could always have fallen back on the manioc, which can be expected to remain in edible condition in the ground for several years. Whether such farming as this is theoretically significant — whether it would ordinarily lead to any kind of sociocultural or demographic progress — may seem question- able. But that it is farming cannot be denied. Agriculture in some in- stances can have evolved for reasons which are both unrecoverable and trivial.

In summary, one can produce a number of quite disparate models of agricultural origins, varied according to the constraints and commodities assumed to be necessary. I myself see little to choose between them. Any could have happened. If we assume that agriculture was independently "invented" often enough, then all of these causal sequences should have unrolled at least once somewhere in the world.

As for the role of growth of population, this clearly varies from case to case. In the model of the population under siege, it is always present but not, as will be pointed out shortly, as a truly independent variable. In the two-staple model, increasing density may or may not be present, and demographic causation is entirely absent from the model of the part-

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 37

time forest-farmers. But before demographic issues can be dealt with properly two observations on that subject must first be made.

THE NATURE OF POPULATION PRESSURE AND INCREASE

The two observations in question have to do with (1) the a priori prob- ability of being able to project demographic growth curves into the past and so to make assumptions about the size of ancient populations, and (2) what does and does not constitute demographic pressure.

Increase Curves and Frame-Dependence

It is usual, when discussing the influence of demography on societal and economic development, to consider that long-term population growth is represented by the familiar exponential curve (Figure 1).

A curve much like Figure 1 would, with somewhat varying parameters, be accepted by most specialists as a fair model of worldwide demographic trends between the Palaeolithic and the present. It would be accepted validly; that international population growth has actually followed such a curve is not open to doubt. But its usefulness is quite another matter. There are reasons for questioning whether the exponential-curve model has any explanatory relevance to early socioeconomic evolution.

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

3 8 BENNET BRONSON

The main reason is that socioeconomic events do not (or did not until recently) happen on a worldwide scale. They take place instead within restricted blocks of area measuring at most a few hundred miles on a side, and have their roots in causes which operate within a similarly reduced frame. If we are interested in demographic causation then densities of continental populations are of no interest to us; such data are meaningless abstractions. And if we come to consider the probable history of popula- tions within restricted regions and localities, the exponential-curve model becomes unsatisfactory as a predictor of demographic density.

Empirically speaking, it is difficult to find a single example of a regional or local population before the era of modern medicine known to have followed a steady pattern of exponential increase for longer than a few centuries. Virtually every population of this kind for which we have long- term documentary records can be shown to have undergone substantial fluctuations. If we consider only the period before A.D. 1800, taking the diffusion of the Jenner vaccine as the cut-off point for the beginning of demographically effective medicine, we find that the late eighteenth century rarely marks the known apogee of any regional population. Northern Europe may be an exception, but in most regions the premedical peak was reached long before 1800 and was followed by a considerable decline afterwards. Aztec Mexico, Byzantine Anatolia and Egypt, pre- Mongol Persia, and perhaps Sung China and Roman Italy and northern Africa are examples of such early peaks. And in areas smaller than regions and nations, the short-term fluctuations must completely overwhelm any secular trend toward gradual increase. Seen within a frame of this size, the exponential curve cannot be expected to resemble the actual histories of populations except in a small fraction of cases.

The theoretical explanation for the "frame-dependence" of demographic models is obvious and need not occupy much of our time. Human populations are capable of intrinsically high rates of increase — even under premedical conditions — of a doubling rate of less than fifty years. The inhabitants of a given locality should therefore be able to fill it solidly with human bodies within the space of one or two millennia. The prob- ability that a population will actually sustain such an increase rate over a large area is of course vanishingly small, but as the spatial frame shrinks the probabilities change. If the frame is a region of 10,000 square kilo- meters, possibly this regional population has at some time in its history remained free from excessive mortality for long enough to produce a substantial population boom. And if the frame is a locality measuring only 500 square kilometers in size, the probability approaches certainty. Given a moderate reduction in mortality, the likelihood of in-migration,

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 39

and the absence of controls over fertility that is almost universal among modern peoples, we may assume that almost all 500 square-kilometer local populations have undergone a number of extreme fluctuations during the last ten millennia. The actual population curve for a locality of such a size would probably resemble Figure 2 more closely than Figure 1.

Figure 2. Model of population increase on a local scale

One interest of this indeterminate population model is that it frees us of the need to find mechanisms by which low densities of continental populations can be assumed to exert significant pressures on resources of land and labor. There is no need, for instance, to postulate that pre- Neolithic gatherers were driven to adopt a major subsistence change because of crowding at high relative densities of several persons per hundred square kilometers. If high absolute local densities are needed for a hypothesis, then they can be assumed to have existed almost anywhere and at any date.

But this conclusion has a corollary, and the corollary is of equal interest and importance: high densities of population do not invariably lead to the adoption of agriculture. Density-induced resource scarcity must have occurred in numerous localities during the late Pleistocene; even though these densities must sometimes have been considerable, not once are they known to have resulted in the large-scale cultivation of staple crops. In the early Holocene, such densities must have occurred at numerous times and places again; yet in only five or ten small regions can they be shown to have led to farming. Demographic pressure is thus a most inefficient

υ CO ο ο ο

ο ο ο ο

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

4 0 BENNET BRONSON

cause. When response follows the presumed stimulus only once in each ten thousand trials, one is justified in doubting the adequacy of that stimulus as an explanation.

Density versus Pseudo-Density

The second observation that should be made has to do with the nature of population density. From the standpoint of possible socioeconomic consequences, what does and does not constitute a "dense" population? No difficulties arise if we envision a classical situation of an increasing number of inhabitants, fixed renewable resources, an area finite in size, and a static exploitative technology; the population becomes dense and begins to experience scarcity at or rather below the point where the rate of consumption equals the rate of renewal of resources. A slight compli- cation that also causes no real difficulties appears when consumers and resources are distributed unevenly within the area. A fine-grained distribution of resources can be expected to result in a lowered threshold of scarcity and population density. The same effect should follow if the consumers are distributed in a coarse-grained fashion. The maximum carrying capacity of a locality is reached only when resources are clustered into easily exploitable nodes, and when the exploiters are spread out as evenly as possible.

On the other hand, there are some sorts of complications which cause real difficulties for the concepts of density and scarcity. One is the probability that any resource which is not necessary for survival can be exhausted eventually by a bare handful of consumers, just as long as these consumers are sufficiently omnivorous, determined, and improvident. Into this category fall almost all individual species of plants and animals. The consumers can always fall back on other species once the preferred ones grow scarce or have been exterminated. The category also includes all non-staple species as a class; a few consumers can exterminate these without suffering any consequences except perhaps for a certain regret at the disappearance of a favored condiment. Hence, we cannot always glibly say that some sorts of scarcity are due to population pressures; they may be due to simple overconsumption. Population pressure is not a meaningful concept except when referred to a critical class of resources, so critical that increasing scarcity can be presumed to bring Malthusian demographic checks into operation.

Another serious complication arises when we consider more carefully the subject of population distribution within the local area. Let us imagine,

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 41

for instance, an underpopulated valley inhabited by a number of house- holds which are relatively dispersed but focused around a single non- subsistence feature in the center, such as a shrine or defensible hilltop. The resulting settlement-pattern might resemble Figure 3. The issue here involves the problems of supply faced by households located at differing distances from the focal feature. The household marked Β on Figure 3 is out at the edge of the settled area; its inhabitants may be far from the population focus but are otherwise in an advantageous position, closer to their fields, to wild resources, and to most other things necessary to the household economy. As a consequence, B's resupply costs, measured in time and effort of transportation, are relatively low. Household A, on the

Figure 3. Pseudo-density caused by locational factors

other hand, is in a high-cost location. It may gain some advantages from its proximity to the center, but few of these advantages are economic. It is further from most resources and must regularly expend more time and labor in obtaining them, in spite of the fact that the valley contains adequate land and that no resource is scarce in an absolute sense.

The model being constructed here is only a restatement of the familiar "Isolated State" model of Von Thiinen (Chisolm 1967; Chayanov 1966), whereby locational considerations — factors of distance and ease of transportation — are said to be decisive in optimizing the choice of crops

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

4 2 BENNET BRONSON

and farming techniques in a marketized regional economy. However, this particular version of the Von Thiinen model is being used to point up a somewhat different conclusion: that even in an unmarketized pure subsistence economy, locational variations will still produce cost differen- tials and hence exert the same kind of economic pressure as does genuine scarcity. From an operational point of view, cost (in this case, labor cost) is the only meaningful measure of scarcity. That a given resource is actually common somewhere within the valley makes no difference to the inhabitants of Household A. For them, the resource is hard to get, and they are therefore under economic pressure, a kind of pressure which is most difficult to distinguish from the pressure caused by overpopulation.

Analogous forms of "pseudo-populational" pressure can be presumed to exist at some level in all societies for whom the choice of a place to live is not dictated by the location of a single food resource. If the location of the settlement, whether temporary or permanent, is chosen partly on the basis of defensibility, sociability, or the presence of a second critical resource, then some members of the society will be under appre- ciable pseudo-pressure. Depending on the keenness with which this pressure is felt, those members will be more or less receptive to the idea of new subsistence alternatives.

The last complication for the concept of population pressure has already been discussed, the fact that increased density must first pass through the filter of cultural perception before it is likely to have any socioeconomic effect. In human as distinguished from animal populations, pressure, scarcity, and stress are to a considerable extent states of mind. A group which feels itself in need of Lebensraum may take steps to solve the problem even though, by a more objective measure, the shortage of living space is largely imaginary. Likewise, the scarcity of a resource is not measured in actual labor cost but instead in terms of PERCEIVED labor, and this will clearly depend on a whole host of variables besides caloric expenditure and man-hours worked. In all probability such perceptual factors will usually tend to lower thresholds of pressure and scarcity rather than raise them. But how much these thresholds will be lowered in any particular case is impossible to predict. Thus it follows that an appreciable percentage of ancient subsistence changes will not be explainable by objective economic and demographic factors. Repugnant though it is to our nomothetic instincts, we must consider the possibility that some changes, including some instances of the inception of agri- culture and cultivation, may have been caused by a perceptual mistake.

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 43

T H E LIMITS OF D E M O G R A P H I C E X P L A N A T I O N

None of the foregoing is meant to deny the validity of some demographic explanations. Unquestionably population pressure has been significant and sometimes decisive in documented cases of recent alterations of subsistence patterns. The twentieth century intensification of farming by the Ibo (Netting 1969) is quite clearly a more or less direct effect of the recent population boom in Eastern Nigeria, and a whole series of historic shifts in English agriculture have been convincingly tied by Slicher van Bath (1963) to price fluctuations and, through this intermediary market mechanism, to long-term national demographic changes. But there are also numerous countervailing examples. Many of the pre-modern agri- cultural innovations in Tokugawa Japan (T. Smith 1968) seem to have been accomplished through administrative fiat, because of a concern for increased productivity on the part of landlords and tax collectors. Similar incentives to agronomic change are also known to have been present in eighteenth-century England and in Rome at the time of Virgil. Many modern changes of subsistence in Africa are better interpreted as responses to market development than to increase in population — witness the appearance of land-extensive commercial agriculture among the Gishu (Allan 1965) and Kofyar (Netting 1968). A twentieth-century farmer in New Jersey or Kent selects techniques and crops with regard only to input costs and output prices; he (and, one imagines, his counterparts on the outskirts of any ancient city) farms in a singularly labor- and land- intensive fashion because land is dear, transport to urban markets cheap, and prices for perishable produce high. Whether the total population within the city's hinterland is dense or sparse makes no difference to his choice of farm technology. If he is close to the city, even though that city may be in the midst of a fertile and uninhabited wasteland, he will be an intensive farmer.

Numerous other examples could be cited but there is no need. It is absurd to maintain that, in the modern and recent world, simple demo- graphic density is invariably the prime mover of subsistence change. It may be important in many cases and decisive in some, but too many other factors affecting subsistence exist — market forces, administrative controls, limitations on information, differences in perception — for one to conclude that population pressure alone is an adequate explanation for the majority of ethnographically and historically known instances of the intensification of subsistence methods.

Perhaps it may seem that the early days of farming represent a more pristine and simpler pattern, when the primacy of population pressure

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

4 4 BENNET BRONSON

should emerge more clearly. However, as the preceding pages have tried to show, this commonsense expectation encounters a number of theoretical difficulties.

The inception of cultivation (as distinguished from full-fledged farming) would seem to have a most tenuous a priori connection with increase in population. This follows directly from the postulate that the first cultivated plants need not have been staple crops. If they were not staples, or were not treated as staples, then they can hardly have begun to be grown because the growers were faced with imminent starvation. One of the archetypal instances of pre-agricultural cultivation is the El Riego phase at Tehuacän, where the ordinary diet, on the evidence of coprolites, is said by MacNeish (1972: 71) to have contained between 0 and 6 percent of cultivated plants. Now, this quantity of food may have made a consider- able difference to the comfort and even nutrition of the ancient Tehua- canos. But it did not save anyone from starving to death. One cannot believe that the Tehuacanos began cultivating in order to obtain 6 percent more of a staple, or because they sensed that a decline of 6 percent in gathering output meant future disaster. Whyever they began, there was no perceptible wolf at their door.

The beginnings of agriculture, on the other hand, may have had a firmer relationship with demographic factors. If the knowledge of cultivation was already widespread, it is entirely plausible that a popu- lation crisis could have turned a group of hunter-gatherers into farmers almost overnight. But one can think of other equally plausible reasons for taking that drastic step — locationally generated pseudo-pressure, conflicts between positionally fixed resources, the social benefits of sedentism, perhaps even at times the increased ease and diminished risk of farming as against gathering. It is true that these other reasons may have had a demographic component, but then demographic causes them- selves must always have been much diluted by other factors. The model of straight population pressure is inadequate as an explanation even of situations where a marked demographic increase can be shown to precede staple agriculture, for of necessity the increased population must be a purely local phenomenon which cannot exist without factors — called here, "locational constraints" — that keep the excess people from wandering off into the surrounding emptiness. And so which is the in- dependent variable, the population increase or the constraint?

To my mind, such questions are both unanswerable and unnecessary. What we are dealing with is a complex, multifaceted adaptive system, and in human adaptive systems (as in real natural and human systems of any kind), single all-efficient "causes" cannot exist. True, it may be advan-

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 45

tageous occasionally to construct models of such systems in which a single factor is given paramount status. But in the case of this particular system, the heuristic value of a simple model is most doubtful, perhaps especially when the paramount factor is to be demography. Population pressure is not the only possible explanation of farming. Nor does it invariably lead to farming. As pointed out earlier, high local densities must have occurred very early in man's history and with great frequency; only in a small percentage of post-Pleistocene cases can these have led to the adoption of large-scale food production. Thus, increase in population is neither necessary nor sufficient as an explanation. It is also among the most difficult of all data to recover archaeologically, depending as it does on excavations on a tremendous scale and on datings of an improbable accuracy. Even if it were true that in a given case a rapid increase in population had immediately preceded and thus presumably caused the appearance of true farming, that fact would be most difficult to demon- strate through any conceivable excavation. And, as I say, the farming may have many other explanations. The population-centered model of subsistence evolution may be pedagogically useful but it is of doubtful value as a research guide.

Much more satisfactory is the rather subliminal model that seems actually to guide much of the research on post-Pleistocene adaptations, whatever the explicit theoretical orientation of the individual researcher may be. The leading characteristics of this model are complexity, factor feedback, and instability. A great many agencies — sedentariness, epidemiology, genetics, environmental structure, technologies of sub- sistence and non-subsistence, political evolution, economic development, warfare, the density and distribution in space of populations •— are recognized as potential influences, without seriously contending that any necessarily have priority. The relationship between each pair of these is visualized as one of feedback; the chicken-and-egg quality of interactions between adaptational factors has long been recognized by most specialists. And the rather Augustinian notion that all recent (i.e. post-Pleistocene) adaptive patterns are intrinsically unstable is gaining ground again after a brief setback during the heyday of functionalism. Change, driven by the sheer impossibility of keeping so many interacting factors out of dis- equilibrium, is a normal condition. What requires explaining is stability, not change.

Under the influence of this implict model, a considerable quantity of significant work has been done. Indeed, in spite of the regrettable lack of detailed and overt consensus, the model has probably generated as much useful research as has the average paradigm of one of the Kuhnian

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

46 BENNET BRONSON

"normal" sciences. It is too good a model to be replaced casually. However, beyond question it will be replaced. The appearance of

numerous proposals for new explicit models of socioeconomic evolution, together with a growing feeling that the field is on the edge (or over the edge) of a breakthrough, signals the old model's approaching demise. As yet the few comprehensive models that have been attempted have not been unqualified successes. But a number of partial models, focused on disentangling only a few strands of the web of factor relationships, have done quite well in terms of generating research hypotheses that are at once testable, non-trivial, and interesting. It would seem that studies of ancient demography could be aimed best at producing partial models like these, at clarifying the connections among a small number of pre- cisely defined and quantified variables of which one is size of population. Such an aim may seem dishearteningly modest when compared with the dimensions of the overall problem of why the long Pleistocene stasis did slip over into a disequilibrating mode and produce the world as we now know it. But a sharply limited approach is the only one that is likely to be productive. Testable explanations for grand patterns are not necessary for research, nor are they practicable in the present state of the art.

REFERENCES

ADAMS, R. M. 1966 The evolution of urban society. Chicago: Aldine.

ALLAN, w . 1965 The African husbandman. Edinburgh: Oliver and Boyd.

BOSERUP, E. 1965 The conditions of agricultural growth. Chicago: Aldine.

BRONSON, B. 1972 "Farm labor and the evolution of food production," in Population

growth: anthropological implications. Edited by Brian Spooner, 190- 218. Cambridge, Massachusetts: M.I.T. Press.

CHAYANOV, Α. V. 1966 The theory of peasant economy. Homewood, Illinois: American Eco-

nomic Association. CHISOLM, N.

1967 Rural settlement and land use. New York: John Wiley. CLARK, C., M. HASWELL

1967 The economics of subsistence agriculture (third edition). New York: St. Martin's Press.

FOUND, w . c . 1971 A theoretical approach to rural land-use patterns. New York: St.

Martin's Press.

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

The Earliest Farming: Demography as Cause and Consequence 47

GORMAN, C. F. 1973 Excavations at Spirit Cave, north Thailand: some interim inter-

pretations. Asian Perspectives 13:79-107. HOMANS, G. c .

1970 English villagers of the thirteenth century. New York: Harper and Row. LEE, R. B.

1972a "Population growth and the beginnings of sedentary life among the !Kung Bushmen," in Population growth: anthropological implications. Edited by Brian Spooner, 329-342. Cambridge, Massachusetts: M.I.T. Press.

1972b "Work effort, group structure, and land-use in contemporary hunter- gatherers," in Man, settlement and urbanism. Edited by P. J. Ucko, R. Tringham, and G. W. Dimbleby, 177-186. Cambridge, Massachu- setts: Schenkman.

MAC ARTHUR, R. Η., E. O. WILSON 1967 The theory of island biogeography. Monographs in Population Biology

1. Princeton, New Jersey: Princeton University Press. MAC NEISH, R. S.

1972 "The evolution of community patterns in the Tehuacan Valley of Mexico and speculations about the cultural processes," in Man, settlement and urbanism. Edited by P. J. Ucko, R. Tringham, and G. W. Dimbleby, 67-93. Cambridge, Massachusetts: Schenkman.

MANGELSDORF, P. C., R. S. MAC NEISH, G. R. WILLEY 1964 "Origins of agriculture in Middle America,," in Handbook of Middle

American Indians, volume one. Edited by R. C. West, 427-445. Austin: University of Texas Press.

MORGAN, W. B. 1955 The change from shifting agriculture to fixed settlement in southern

Nigeria. Department of Geography Research Notes 7. Ibadan, Nigeria: University College of Ibadan.

NETTING, R. M. 1968 Hill farmers of Nigeria. Seattle: University of Washington Press. 1969 Ecosystems in process: a comparative study of change in two West

African societies. National Museum of Canada Bulletin 230:102-112. PEAKE, H.

1928 The origins of agriculture. London: E. Benn. SAHLINS, M.

1972 Stone age economics. Chicago: Aldine. SAUER, c . o .

1952 Agricultural origins and dispersals. New York: The American Geo- graphical Society.

SLICHER VAN BATH, B. H. 1963 The agrarian history of western Europe A.D. 500-1950. London:

Edward Arnold. SMITH, c . E.

1967 "Plant remains," in The prehistory of the Tehuacan Valley, volume one. Edited by Douglas S. Byers, 220-255. Austin: University of Texas Press.

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .

48 BENNET BRONSON

SMITH, P. E. L. 1972 "Land use, settlement patterns, and subsistence agriculture: a demo-

graphic perspective," in Man, settlement and urbanism. Edited by P. J. Ucko, R. Tringham, and G. W. Dimbleby, 409-426. Cambridge, Massachusetts: Schenkman.

SMITH, P. E. L., T. C. YOUNG

1972 "The evolution of early agriculture and culture in Greater Mesopota- mia : a trial model," in Population growth: anthropological implications. Edited by Brian Spooner, 1-59. Cambridge, Massachusetts: M.I.T. Press.

SMITH, τ . c .

1968 The agrarian origins of modern Japan. Stanford: Stanford University Press.

TERRA, G. J. A. 1954 Mixed garden horticulture in Java. Journal of Tropical Geography 3:

34-43. WOODBURN, J. C.

1968 "An introduction to Hadza ecology," in Man the hunter. Edited by R. B. Lee and I. DeVore, 49-55. Chicago: Aldine.

1972 "Ecology, nomadic movement, and composition of the local group among hunters and gatherers," in Man, settelement and urbanism. Edited by P. J. Ucko, R. Tringham, and G. W. Dimbleby, 193-206. Cambridge, Massachusetts: Schenkman.

C op

yr ig

ht ©

1 97

8. D

e G

ru yt

er , I

nc ..

A ll

rig ht

s re

se rv

ed .