The Brenner Debate
Sixteenth- and seventeenth-century England witnessed an agricultural revolution which involved massive changes in land tenure, the organization of production on farms, the techniques employed in farming, and the productivity of agriculture. Thus the sixteenth century represented a sharp change in English rural life: the emergence of the capitalist farm in place of small-scale peasant cultivation, the intensification of market relations, increase in population, and eventual breakthrough to capitalist development in town and country. The social consequences of this revolution were massive as well: smallholding peasant farming gave way to larger capitalist farms; hundreds of thousands of displaced peasants were rapidly plunged into conditions of day labor, first in farming and then in manufacture in towns and cities; higher farm productivity permitted more rapid urbanization and the growth of an urban, commercialized economy; and higher real incomes provided higher levels of demand for finished goods which stimulated industrial development. Thus the agricultural revolution was the necessary prelude to the industrial revolution in England.  It was the growth of agricultural productivity, rooted in the transformation of agrarian class or property relations, which allowed the English economy to embark upon a path of development foreclosed to its Continental neighbours. This path was distinguished by continuing industrialization and overall economic growth through the period when `general crisis' gripped the other European economies (Brenner 1982:110).
It was indeed, in the last analysis, an agricultural revolution, based on the emergence of capitalist class relations in the countryside which made it possible for England to become the first nation to experience industrialization [through higher levels of grain productivity and higher income to stimulate demand for industrial goods]. (Brenner 1976:68)
This process poses at least two problems for historical explanation. First is an historical question: why did breakthrough occur in England in the sixteenth century and not the fifteenth or the nineteenth? And the second is geographic: why did this process of agricultural development occur in England but not on the Continent? In particular, why did agrarian life in the French countryside remain relatively unchanged throughout this period? And why did eastern Europe slide into a second feudalism? 
A variety of explanations have been advanced for these developments. Some economic historians (e.g., M. M. Postan and Emmanuel Le Roy Ladurie) have maintained that the cause of this process of change was an autonomous increase in either population or commerce or both. Robert Brenner argues, however, that these explanations are inadequate, since these large-scale factors affected the whole of Western Europe, while capitalist breakthrough occurred only in Britain. Brenner holds that the determining factor is the particular character of social-property relations in different regions of Europe (particularly the conditions of land-tenure and associated forms of surplus extraction), the interests and incentives which these relations impose on the various actors, and the relative power of the classes defined by those relations in particular regions.
Brenner's explanation of these developments is thus based on micro-class analysis of the agrarian relations of particular regions of Europe. The processes of agricultural modernization unavoidably favored some class interests and harmed others. Capitalist agriculture required larger units of production (farms); the application of larger quantities of capital goods to agriculture; higher levels of education and scientific knowledge; etc. All of this required expropriation of small holders and destruction of traditional communal forms of agrarian relations. Whose interests would be served by these changes? Higher agricultural productivity would result; but the new agrarian relations would be ones which would pump the greater product out of the control of the producer and into elite classes and larger urban concentrations. Consequently, these changes did not favor peasant community interests, in the medium run at least. It is Brenner's view that in those regions of Europe where peasant societies were best able to defend traditional arrangements--favorable rent levels, communal control of land, and patterns of small holding--those arrangements persisted for centuries. In areas where peasants had been substantially deprived of tradition, organization, and power of resistance, capitalist agriculture was able (through an enlightened gentry and budding bourgeoisie) to restructure agrarian relations in the direction of profitable, scientific, rational (capitalist) agriculture.
Brenner's case is primarily constructed in opposition to the prevailing theories of economic breakthrough based on demographic and commercial factors.  According to the demographic theory, medieval European populations went through cyclical phases of growth and decline, and a significant upswing in the late 16th-century caused breakthrough in agriculture by stimulating demand for food.  North and Thomas encapsulate this view in The Rise of the Western World: The growth of population and the resultant expansion of organized markets and a money economy changed the basic conditions which had given rise to feudal society (1973:64).
A central defender of the price and population view is M. M. Postan. He writes,
Behind most economic trends in the middle ages, above all behind the advancing and retreating land settlement, it is possible to discern the inexorable effects of rising and declining population. (Postan 1972:30)
Postan's theory is not exclusively demographic; instead, it accounts for largescale economic and social changes in Britain as a result of the interaction between population change and prices.
Throughout this study, prices and wages have been repeatedly invoked to account for changes in economic activity and in the well-being of the individuals. These references to prices would appear to have at least two important economic implications: first, that prices fluctuated in response to supply and demand, and secondly, that their fluctuations affected economic and social conditions. . . . Skeptical questions about `markets' and prices have in fact been raised by a number of sociologists and historians dealing with underdeveloped societies and economies. (Postan 1972:253)
Emmanuel Le Roy Ladurie explicitly accepts the Malthusian label; part one of The Peasants of Languedoc is called Malthusian Renaissance. Le Roy Ladurie writes of his purposes in that book,
My book's protagonist is a great agrarian cycle, lasting from the end of the fifteenth century to the beginning of the eighteenth, studied in its entirety. I have been able to delineate and to characterize it thanks, naturally, to the price curves, but more particularly thanks to demographic studies (of taxpayers and of total population), to indices of production and business activity, and to the series of charts reflecting land, wealth, and income distribution. (Le Roy Ladurie 1974:289)
And he explicitly identifies population changes as the key to economic developments in Languedoc:
The tragic demographic situation of the fifteenth century--the scarcity of people--was the overriding fact that lend land settlement, economic life, and social relationships their peculiar coloration on the eve of the great advance of the modern period. (11)
Brenner rejects this family of explanations, however, on the ground that these factors--population increase, price and wage changes--are factors which were common throughout Europe during the relevant period, and yet were accompanied by breakthrough to capitalist agriculture in some areas, tightening of feudal bonds in others, and entrenched peasant freeholding in yet others. Therefore population increase cannot be the decisive causal factor leading to breakthrough.
Instead, Brenner holds that it was the particular features of local class and property relations that determined whether increasing commerce and population would lead to breakthrough to capitalist agriculture or not. Changes in relative factor scarcities consequent upon demographic changes exerted an effect on the distribution of income in medieval Europe only as they were, so to speak, refracted through the prism of changing social-property relations and fluctuating balances of class forces (Brenner 1982:21).
I began from the idea that social-property systems, once established, tend to set strict limits and impose certain overall patterns upon the course of economic evolution. They do so because they tend to restrict the economic actors to certain limited options, indeed quite specific strategies, in order best to reproduce themselves--that is, to maintain themselves in their established socio-economic positions. (Brenner 1982:16)
Under different property structures and different balances of power, similar demographic or commercial trends, with their associated patterns of factor prices, presented very different opportunities and dangers and thus evoked disparate responses, with diverse consequences for the economy as a whole. Indeed, . . . under different property structures and balances of class forces . . . precisely the same demographic and commercial trends yielded widely divergent results. (Brenner 1982:16-17)
It follows therefore that long-term economic changes, and most crucially economic growth, cannot be analysed adequately in terms of the emergence of any particular constellation of `relatively scarce factors' unless the class relationships have first been specified; indeed, the opposite outcomes may accompany the impact of apparently similar economic conditions. In sum, fully to comprehend long-term economic development, growth, and/or retrogression in the late medieval and early modern period, it is critical to analyse the relatively autonomous processes by which particular class structures, especially property or surplus-extraction relations, are established and in particular the class conflicts to which they do (or do not) give rise. (Brenner 1976:32)
Brenner's thesis is thus that the central causal factor in the pattern of agrarian development in Europe--the factor which covaries the most closely with patterns of development, regression, and stagnation--is the particular character of class relations in a region and the set of incentives and opportunities which the local property relations impose on the various actors.
My argument thus started with the assertion that the feudal social-property system established certain distinctive mechanisms for distributing income and, in particular, certain limited methods for developing production, which led to economic stagnation and involution. It did so, most crudely, because it imposed upon the members of the major social classes--feudal lords and possessing peasants-- strategies for reproducing themselves which, when applied on an economy-wide basis, were incompatible with the requirements of growth. (Brenner 1982:17)
Agricultural revolution fundamentally involved two processes: the creation of larger units of farming land and the application of more rational and efficient techniques to the farming process (new crop rotation systems, new technologies and tools, etc.). These processes could not proceed except by abolishing older forms of subsistence agriculture: peasant farming on small plots using traditional techniques and tools. These processes thus involved a radical change in the economic environment in a region, favoring the economic interests of some groups and fatally harming the interests of other groups. Peasant farmers, in particular, had little to gain from this process and much to lose, since the concentration of larger farming units unavoidably involved the destruction of peasant proprietorship.
In order for these changes to take place in a given area, therefore, it was necessary that there be a group (a microclass) which had both the economic interest to further these changes and the political power to do so. On Brenner's account, the decisive factors determining whether this process of change would take place or not were the particular set of property arrangements which existed in a given region and the distribution of the means of power which were available to contending groups. Where landlords could effectively force peasants to accept changing property relations, the transition to capitalist agriculture could proceed. But where peasant communities had the power to resist changes in property relations--in particular, abolition of communal lands, expropriation of small peasant property, etc., they were able to block the emergence of the property relations within which capitalist agriculture, wage relations in the countryside, and more efficient production could emerge.
Simply stated, it will be our contention that the breakthrough from `traditional economy' to relatively self-sustaining economic development was predicated upon the emergence of a specific set of class relations in the countryside, that is capitalist class relations. This outcome depended, in turn, upon the previous success of a two-sided process of class development and class conflict: on the one hand the destruction of serfdom; on the other hand, the short-circuiting of the emerging predominance of small peasant property. (Brenner 1976:47)
This analysis depends upon the economic interests and political capacities of the various actors in a given region, especially the small cultivator and the landlord. Consider first the cultivator (peasant). His interests may be defined fairly simply: to achieve the most extensive possible property rights over the land which he farms; to keep rents at the lowest level possible; to reduce to a minimum the forms of labor services and other non-rent obligations he owes the lord; and to minimize tax obligations.
The cultivator's political capacities as an individual were quite limited; collectively, however, peasants in different regions of Europe developed significant powers of resistance to lordly exactions. Brenner identifies several different sorts of political resources available to peasant communities in various regions of Europe: a history of effective collective action (including peasant rebellion) in support of peasant interests and rights; complex and inefficient feudal relations in some regions involving conflicting lordly rights to a given community (thus permitting peasants to play off one lord against another); and peasant alliances with the monarch against landed property through which peasants were sometimes able to entrench and defend property settlements favorable to their interests.
The key to French developments in agrarian relations, for example, may be found in the extensive levels of power and solidarity which French peasant communities were able to achieve with respect to their lords:
By the early fourteenth century, the peasants of northern France had achieved effectively full property rights to the customary land (fixed, minimal dues and right to inherit). (Brenner 1982:50)
These property rights were ultimately grounded in peasant traditions of collective political action:
In Catalonia . . . one also finds increased legislation by the Corts . . . to limit peasant movement and decrease personal freedom. . . . But, correlatively, it provoked in response a high level of peasant organization and, in particular, the assembling of mass peasant armies. . . . Only a series of violent and bloody confrontations ultimately assured peasant victory. Armed warfare ended finally in 1486 with the Sentence of Guadalupe by which the peasantry was granted in full its personal freedom, full right in perpetuity to its property . . . and, perhaps equally important, full right to those vacant holdings which they had annexed in the period following the demographic catastrophes. (Brenner 1976:51-52)
It was rather the manifestation of peasant conquests, achieved through the resistance of highly cohesive French peasant communities. (Brenner 1982:56)
Thus a history of peasant collective action, political organization, and rebellion significantly enhanced the capacity of peasants to defend and extend their property rights.
The property rights which peasants successfully established and defended in France led to technical stagnation in agriculture, however. This was true in part because peasant production was inherently low-yield; it took place on a small scale, using traditional and relatively inefficient techniques of cultivation. But peasants were also less able to make capital investments in their farms because of their vulnerability to surplus- extracting agencies above them. It was possible for other agencies-- landlords and the state, particularly--to extract what small surplus that existed from peasant cultivators.
the emergence of a structure of ownership of land . . . provided the peasantry in most of France (in contrast to England and elsewhere) with relatively powerful property rights over comparatively large areas of the land. This presented a powerful barrier to those who wished to concentrate land. For whatever the market situation . . . the peasantry would not in general easily relinquish their holdings, the bases of their existence and that of their heirs. . . . It was thus, I will argue, the predominance of petty proprietorship in France in the early modern period which ensured long-term agricultural backwardness. This was not only because of the technical barriers to improvement built into the structure of small holdings, especially within the common fields. It was . . . because peasant proprietorship in France came to be historically bound up with the development of an overall property or surplus- extraction structure which tended to discourage agricultural investment and development; in particular, the heavy taxation by the monarchical state; the `squeezing' of peasant tenants (leaseholders) by the landlords; and, finally, the subdivision of holdings by the peasants themselves. (Brenner 1976:46)
At the same time, because of lack of funds--due to landlords' extraction of rent and the extreme maldistribution of both land and capital, especially livestock--the peasantry was by and large unable to use the land they held in a free and rational manner. (Brenner 1976:49)
Brenner also applies his analysis to differences in the course of feudalism in western and eastern Europe. 
The development of peasant solidarity and strength in Western Europe--especially as this was manifested in the peasants' organization at the level of the village--appears to have been far greater in Western than in Eastern Europe; and this superior institutionalization of the peasants' class power in the West may have been central to its superior ability to resist seigneurial reaction. (Brenner 19:56)
The relative absence of village solidarity in the east . . . appears to have been bound up with the entire evolution of the region as a colonial society--its relatively `late' formation, the `rational' and `artificial' character of its settlement, and especially the leadership of the landlords in the colonizing process. (Brenner 1976:57)
At the same time, the planned, landlord-led organization of the settlement in the east tended to place major barriers in the way of the emergence of peasant power and peasant self- government. East German villages were generally smaller and less dense than their western counterparts; they tended, moreover, to have but a single lord. As a result they were less difficult for the lords to control than were villages of the west, where the thick population and, in particular, the tendency of the villages to be divided between two or more lordships, gave the peasants more room to manoeuvre, making gemeinbildung that much easier. (Brenner 1976:58)
The lords of north-east Europe--eastern Germany, as well as Poland--had led and controlled from the start a belated process of agrarian development, imposing `artificial', rationally laid out forms of peasant settlement. In contrast, their counterparts in western Europe had to impose their power `from the outside', against peasant communities which were longer established and better organized--with established traditions of (often successful) struggle for their rights. (Brenner 1982:71)
Let us turn now to the circumstances of the local ruling classes in various regions of Europe. Brenner points out that lords do not always have an economic incentive to increase the efficiency of agriculture. Under some economic circumstances the lord's interests may be best served by intensifying old forms of surplus extraction, rather than by reorganizing the production process. Given his unfree peasants, the lord's most obvious mode of increasing output from his lands was not through capital investment and the introduction of new techniques, but through `squeezing' the peasants, through raising either money-rents or labour-services (Brenner 1976:48). These incentives prevented lords from putting into place technical innovations which were readily available.
There were, in fact, known and available agricultural improvements--including the ultimately revolutionary `convertible husbandry'--which could have brought significant improvements in demesne output. . . . Of course, the methods used on the manor of Marley by Battle Abbey were almost totally ignored by English landlords. They generally did not have to improve--to raise labour-productivity, efficiency and output--in order to increase income. This was because they had an alternative, `exploitative' mode available to them: the use of their position of power over the peasants to increase their share of the product. (Brenner 1976:49)
Where the lord's interests were best served by enhancing the efficiency of agriculture, he could do so only if he could marshall the power needed to transform existing property arrangements accordingly. And Brenner holds that English lords were particularly well-situated to accomplish this transformation.
What may have been responsible for the superiority of the English lords as extractors of a surplus from their peasants was their superior self-organization--their superiority vis-a-vis the French lords as feudal centralizers and feudal accumulators. (Brenner 1982:51)
What were the results of these factors in the English experience?
They [capitalist tenants] were therefore set free to bring in those key technological innovations, most especially convertible husbandry systems and the `floating of the water meadows', as well as to make sizable investments in farm facilities, which were generally far less practicable on small unenclosed farms operated by peasants. (Brenner 1976:64)
The contrast with England is clear. There, over the course of the early modern period, one witnesses an agricultural revolution. Given the technology available to the mixed agricultural production of medieval and early modern Europe, qualitative improvement which would make for significant cheapening in basic food production required that animal and arable husbandry be more tightly bound together . . . ; in particular, animal production had to increase. (Brenner 1982:97)
Thus the capitalist farmer was given a clear field: he had the incentive, the opportunity, and the power to effectively organize the farming process so as to produce a profit; and this gave him the incentive to pursue the path of technical reorganization of farming. Peasant agriculture was not wholly eclipsed: This is not to say, of course, that peasant production was incapable of improvement. Thus small scale farming could be especially effective with certain industrial crops (for example flax) as well as in viticulture, dairying and horticulture. But this sort of agriculture generally brought about increased yields through the intensification of labour rather than through the greater efficiency of a given unit of labour input (Brenner 1976:64). But small-scale peasant cultivation was severely limited in scope, and played a smaller and smaller role in English agriculture. 
It is important to Brenner's case to establish that the agricultural revolution was not the result of new technical advances in agriculture. Rather, alternative techniques and forms of organization of production which were more efficient were available for an extended period of time in England, but were not adopted until it was in the interest of a powerful class to incorporate them.  Thus--against other economic historians--it was not newly discovered technological options which drove the agricultural revolution, but rather a new set of incentives and opportunities imposed on a powerful class.
Postan and Hatcher attribute the lack of technological innovation to the `insufficient supply of technological possibilities'. But, if it were true, as they say, that capital-using technologies capable of increasing agricultural productivity were unavailable, then their complementary contention that agricultural investment was insufficient would not make sense. (Brenner 1982:33)
There is a paradoxical side to Brenner's thesis; for the breakthrough to capitalist agriculture led to a dramatic increase in agricultural productivity, which in turn led to higher per capita incomes. Collectively, then, those social systems which embodied stronger institutional bases of power for peasants were the most disadvantaged in terms of longterm development.
In sum, it is not difficult to comprehend the dismal pattern of economic development imposed by this class structure in France. Not only was there a long-term failure of agricultural productivity, but a corresponding inability to develop the home market. Thus, ironically, the most complete freedom and property rights for the rural population meant poverty and a self-perpetuating cycle of backwardness. In England, it was precisely the absence of such rights that facilitated the onset of real economic development. (Brenner 1976:75)
These constitute the main elements of Brenner's theory of breakthrough. Let us now consider the logic of Brenner's case: the explanatory model upon which it depends and the varieties of evidence which are relevant to assessing its correctness. The empirical problem which Brenner is attempting to solve has these main elements. Europe in the fifteenth and sixteenth centuries consisted of a collection of relatively discrete societies from east to west and north to south. All these societies were primarily agricultural, with varying levels of commercial activity. Various data about these societies are available, and much of this data shows significant patterns of geographic distribution.
Some of this data is aggregative and quantitative for some regional level--local, national, or continental. Examples of aggregative data would include per capita income levels, grain production and land productivity data, per capita tax burdens, and population density. Other information is structural: the political regime and institutions which dominate an area, the property defining land tenure, the dominant form of agricultural technique, and the form and history of peasant political activity.
These data provide the context for a particularly important historical fact. One region (England) underwent a dramatic and self-sustaining transition to higher-yield agriculture, whereas other regions witnessed either agricultural stagnation or regression. Brenner's goal is to provide an adequate historical explanation of this fact: an account whose main factual premises are true, and which explains both the emergence of agricultural innovation and the distribution of patterns of agricultural development over the landscape of Europe. Putting this problem simply, Brenner wants to identify the cause of the pattern of agricultural development found in late medieval Europe. And by this he means the factor whose presence or absence determines the presence or absence of agricultural breakthrough.
Brenner's method is one which has an inductive and a deductive side. The inductive requirement amounts to a requirement of correlation: the factor identified as cause should covary appropriately with its effects. This is generally an extremely weak requirement,  though in this particular debate it is sufficient to distinguish between alternative hypotheses.
The deductive requirement is stronger, however; it requires that there should be a plausible and empirically grounded theory which describes the mechanism through which the cause produces its effect. That is, once we have the theory we should be able to see how the presence of the causal factor, in the background circumstances found in different regions of Europe, would give rise to the phenomenon to be explained (agricultural breakthrough).
The bulk of Brenner's arguments are aimed at one of two goals. First, Brenner articulates his own theory of breakthrough and produces the evidence which leads him to favor this theory. And second, he considers and rebuts a small number of alternative hypotheses--those hypotheses which have been leading candidates within the historical literature. Significantly, his presentation of his own model is largely deductive, while his criticisms of alternative theories are largely inductive. He tries to show, not that the neo-malthusian model is theoretically flawed, but rather that it does not fit the facts. It does not identify a causal factor which shows the right pattern of distribution over the map of Europe during this period.
Brenner's deductive argument--his theoretical model--is one based on individual rationality and an analysis of the particular circumstances (e.g., ecological, institutional, and political) within which the various historical actors (lords, peasants, political officials, etc.) make decisions. He assumes that each agent acts in his own material interest, and that different institutional arrangements dictate significantly different strategies to the various actors. Further, he assumes that institutional arrangements affect individuals as classes: peasants share a set of interests defined by the property arrangements which are substantially different from the interests shared by gentry landowners.
The logical structure of this analysis may be summarized in the following terms.
1 Property and rent relations defining the use of land and labor and the distribution of the surplus product varied across late medieval Europe.
2 The particular property relations in a given region defined a distinctive set of material interests for lords and peasants.
3 The political capacities of lords and peasants varied across late medieval Europe as well: central states had varying levels of power in the countryside, lords were better or worse situated to impose their will on peasants, and peasant communities were endowed with varying levels of solidarity and other political resources.
4 All participants pursued their material class interests (as defined by (2)) relatively rationally, given the political resources available to them (as defined by (3)).
Next we need a principle describing lordly decision making:
5 In any circumstances where lords
(i) find it in their interest to rationalize agriculture (concentrate land, reorganize the process of cultivation, and apply more efficient crops and implements to the process),
(ii) have sufficient power to overcome peasant resistance to these changes,
it may be inferred that they will act accordingly.
Finally, we need a principle of aggregation which permits us to infer from the behavior of individual lords to the pattern which may be expected in regions in which lords behave thus:
6 Regions satisfying the conditions described in (5) will manifest a pattern of agricultural development.
Conditions in the local English economy satisfied the conditions described in (5), whereas conditions in France and eastern Europe did not. In France peasant communities had too much political power to permit landlords to successfully engage in agricultural development; while in eastern Europe lords lacked the incentive to engage in development, since they had the easier option of intensifying surplus extraction through harsher rent arrangements. Therefore Brenner's theory predicts that agricultural development will occur in England and will not occur in France or eastern Europe.
Abstractly, then, the argument begins at the level of a model of a local economy. The model describes the primary units of production found in the local area--small farms and a few large manors, let us say, and the property arrangements which regulate these units. It considers the interests and powers of the various agents involved in production--peasant, lord, smallholder, official, and the economic environment within which production takes place; and it attempts to work out the optimal strategy for each agent. Next it attempts to gauge the relative strengths and capacities of the various agents in order to predict the outcome of conflicting strategies. Finally, the analysis attempts to aggregate the consequences of analysis at the local level to a prediction about the pattern of development at the regional level.
Brenner's theory thus has much in common with the rational choice models discussed in earlier chapters. What distinguishes Brenner's account from that of Popkin, for example, is the set of assumptions which gives his theory a Marxist orientation: in particular, his attention to the grounds of political solidarity among members of a microclass (a regional peasantry, for example).
Let us turn now to Brenner's inductive case. An important feature of Brenner's argument is his use of causal reasoning. He holds that comparative analysis of different regions permits us to go a long way towards identifying the causal conditions of various economic developments. His criticisms of the neo-malthusian explanation of breakthrough, for example, depend on the point that the causal factor the neo-malthusians identify--population increase--does not covary with patterns of breakthrough. My concrete method of critique is exceedingly simple and obvious: it is to observe the prevalence of similar demographic trends throughout Europe over the six- or seven-hundred-year period between the twelfth and the eighteenth centuries and to show the very different outcomes in terms of agrarian structure, in particular the patterns of distribution of income and economic development, with which they were associated (Brenner 1976:37). This approach depends upon a method of causal analysis rather similar to Mill's methods of difference (discussed earlier in chapter 1).
Both Postan and Le Roy Ladurie emphasize the importance of demographic factors in constraining economic growth. It is significant, then, that Brenner offers sharp criticisms of the neo-malthusian model in application to Europe. He holds that comparative analysis of different regions permits us to decisively refute some putative causal explanations of economic developments. In particular, his criticisms of the neo- malthusian explanation of breakthrough depend on the point that the causal factor the neo-malthusians identify--population increase--does not covary with patterns of breakthrough.
My concrete method of critique is exceedingly simple and obvious: it is to observe the prevalence of similar demographic trends throughout Europe over the six- or seven- hundred-year period between the twelfth and the eighteenth centuries and to show the very different outcomes in terms of agrarian structure, in particular the patterns of distribution of income and economic development, with which they were associated. (Brenner 1976:37)
This approach depends upon a method of causal analysis rather similar to Mill's methods of difference: the proposition that an event type A causes event type B can be refuted by finding a case of an event a in the absence of an event b.
The obvious difficulty with this whole massive structure [the demographic interpretation] is that it simply breaks down in the face of comparative analysis. Different outcomes proceeded from similar demographic trends at different times and in different areas of Europe. Thus we may ask if demographic change can be legitimately treated as a cause, let alone the key variable. (Brenner 1976:39)
I would simply reassert that this line of argument can be refuted by demonstrating, . . . that the same demographic trends in roughly the same period were accompanied by the opposite trends in income distribution in different European regions. (Brenner 1982:22)
These passages have the effect of postulating a formal requirement on causal reasoning: that causal judgments are refuted by instances in which the putative causal factor is present but not the effect. Since rising population was sometimes associated with self-sustaining economic development and sometimes with malthusian crisis, rising population cannot be the cause of self-sustaining development.
To the extent that the demographic thesis asserts that population change is the determining causal factor, these criticisms are reasonable.  However, it is difficult to imagine that any sophisticated historian could suppose that this is the case. Normally, when a historian asserts that X causes Y, he does not intend to assert that X is the sole cause of Y, but rather, X is an important factor in the causal history of Y. As it stands, therefore, Brenner's implicit standard of causal adequacy is unreasonably restrictive. It is true that sparks cause fires; but there are many circumstances in which sparks are present without fire. Here the problem concerns the distinction between necessary causal conditions and sufficient causal conditions. In a given set of circumstances a spark is the necessary condition for the occurrence of fire. But the spark is never a sufficient condition for the occurrence of fire; oxygen and combustible material are needed as well. At best the spark is part of a set of jointly sufficient conditions for the occurrence of fire. Given the complexity of any large historical process, the aim of identifying a sufficient condition for an event, or pattern of events, is unpromising; rather, the historian may aim more modestly at identifying a set of relevant causal factors which can be combined into an adequate causal analysis of the pattern. Thus when Brenner writes the following, polemics aside, the answer ought to be yes:
Do Postan and Hatcher really wish to argue that an historical explanation can be counted adequate when the factor imputed to be cause (demographic increase/ decline) can be shown to produce the opposite effects (in terms of income distribution) in very similar conditions? (Brenner 1982:23)
The thesis that The American civil war was caused by P is not refuted by showing that there were other circumstances in history where P occurred and civil war did not. It is rather evaluated by considering the full analysis in the context of which the statement is advanced and assessing the empirical and theoretical justification of the larger analysis. On this approach, then, the emphasis shifts from facts about correlation among factors to considerations of theoretical adequacy.
Brenner's implicit inductive requirement can be extended to a requirement of adequacy of causal analysis as an extended treatment of a phenomenon, however:
* In circumstances in which we are presented with a variegated pattern of distribution of a given variable, an adequate causal explanation of the pattern is one which identifies a set of factors and laws which jointly predict the occurrence of the variable in the observed pattern of distribution.
If a putative causal analysis can be shown to predict a significantly different pattern of distribution of the variable, then the analysis must be rejected. In this form, the requirement appears to be reasonable.
Brenner believes that the factor which he has identified--patterns of property arrangements and political power--satisfies this requirement because it generates the right geographical pattern of development and stagnation:
The different systems of property which were established were responsible, in my view, for structuring widely divergent patterns of economic evolution in the different regions--the imposition of different forms of agricultural involution and ultimately general crisis on most of the Continent, the critical breakthrough to self-sustaining growth in England. (Brenner 1982:69)
But notice that his account also takes as data the demographic and price information which is at the center of the competing theory. In Brenner's hands these factors become background conditions or standing conditions in the environment of which property relations exercise their causal influence. Property relations are the differentia specificia; but demographic and price factors are causal factors on his account as well.
On this treatment, the inductive and deductive arguments must be fitted together. In order to conclude that we have a causal understanding of a complex historical process, we need to have:
* a deductive analysis specifying the factors which are causally relevant and the laws and processes which govern them;
* a demonstration that the deductive theory entails a prediction about the distribution of the factor to be explained which approximately matches the observed distribution.
Brenner writes as though his account is an alternative to the more traditional people and prices models; in fact, however, it is more reasonable to conclude that Brenner's chief contribution is to specify an important causal factor which the demographic models leave out: the role which class and property play in determining the local effect of largescale economic forces. The demographic models specify objective economic forces (Brenner 1976:30) which bathe the landscape of Europe on a large scale. These models also tacitly presuppose a simple, uniform, and transparent local mechanism through which this influence is translated into economic effects: rational individuals pursuing their material interests and responding rationally to changing economic circumstances (chiefly prices). Brenner's argument shows, however, that there is significant variation in the institutions which define the local environment of choice in different regions of Europe, and that the outcome of largescale economic forces is highly sensitive to those differences. This suggests, not that the arguments which identify objective economic factors are false, but rather that they are incomplete and must be supplemented with local and regional analysis of the political and legal environment within which these objective economic factors exercise their influence.
Consider an analogy: Sunlight heats exposed surfaces. This statement represents a general causal claim; it identifies an objective physical factor as the cause of a distributed effect (the warming of exposed surfaces). Now suppose, however, that it is observed there are patterns of varying warmth in an area of uniform sunlight. It is then discovered that the color of the surface makes a difference; dark surfaces heat more quickly, to a higher temperature, than light surfaces. Does this show that the initial causal judgment was false? No; rather it shows that the objective physical factor (sunlight) exercises its effect in conjunction with mechanisms which vary from place to place. The initial causal hypothesis is now seen to be incomplete, but not false.
Likewise, it would seem natural to say that the demographic theory describes a widespread causal factor which exercises its influence across regions with relevantly different local properties; as a result, the effects of the objective economic factors differ across the landscape. Brenner's case (if empirically well-founded) adds to our knowledge of the local mechanisms; but it does not invalidate the causal role of the objective economic factors. This indeed seems like the best way to interpret Brenner's remark that changes in relative factor scarcities consequent upon demographic changes exerted an effect on the distribution of income in medieval Europe only as they were, so to speak, refracted through the prism of changing social-property relations and fluctuating balances of class forces (1982:21). For in this passage Brenner is conceding the point that price and population changes affected the economic circumstances of Europe, and is pointing out that they did not do so autonomously, but rather through various sets of local economic arrangements.
In fact, there is much in common between Brenner's analysis and the demographic-economic models: each depends upon the assumption of individual economic and material rationality and each gives central place to the role of economic parameters (rents, prices, wages) in the explanation of social change. For example, there are significant points of similarity between Brenner's case and that of North and Thomas (1973). These two approaches share the assumption that it is the incentives and opportunities defining the circumstances of action of the individual decision maker (cultivator, inventor, manufacturer, merchant) that determine largescale patterns of development. And they share the belief that the motive structure of individuals is largely determined by their material interests (income, property rights, security). The chief difference between the two is Brenner's emphasis on the politics of class, collective action, and rebellion. Brenner gives an extensive role to peasant political activity, and he places more emphasis on the variability of the institutional context of decision making.  But the emphasis on individual material interest and the causal efficacy of alternative institutional contexts is in common between the two approaches.
Brenner's analysis is grounded in some of the central concepts of Marxist analysis and historical materialism: class, property relations, productive forces, and systems of surplus extraction.
Class structure . . . has two analytically distinct, but historically unified aspects. First, the relations of the direct producers to one another, to their tools and to the land in the immediate process of production--what has been called the labour process or the social forces of production. Secondly, the inherently conflictive relations of property--always guaranteed directly or indirectly, in the last analysis, by force--and by which an unpaid-for part of the product is extracted from the direct producers by a class of non-producers--which might be called the property relationship or the surplus extraction relationship. It is around the property or surplus extraction relationship that one defines the fundamental classes in a society--the class(es) of direct producers on the one hand and the surplus-extracting, or ruling, class(es) on the other. (Brenner 1976:31)
Brenner assumes the basic framework of surplus extraction--counterpart to Marx's aphorism that all history is a history of class conflict--as the fundamental structure of economic organization. Social systems involve a division between producers and non-producers; all social wealth is generated in the production process by producers; and the property relations effectively separate the surplus product from the producer to the control and enjoyment of the non-producer.
Classes are constituted by the structure of the property system; individuals fall into classes according to their positions within the property system. This is a structural criterion of class-membership; a serf is a serf whether he regards himself as such or not, since his economic position is defined by the limited property rights which he has in land and the obligations to which he is subject in relation to his lord. But the theory of class may be elaborated in at least two further directions. Classes may become class-conscious; that is, the members of a class may come to identify themselves as a class. And classes may come to constitute themselves as political agencies: they may develop institutions of political organization and traditions of political consciousness which permit them to engage in collective class activity.
We might ask briefly what the status is of Brenner's assumptions about class and surplus-extraction. Are these empirical observations, research hypotheses, theories, ideological interpretations, or some other alternative? Is it a fact that the rent relation in northern France in the 15th century is a scheme of surplus extraction through which lords exploit peasants? Or is the characteristic of exploitation applied here one which is added by the interpretation, capable of reinterpretation along more class-neutral lines? In another context I have argued that the main ideas of historical materialism function as a set of research hypotheses for social scientists (Little 1986:chapter 2); and this appears to be the standing of Brenner's fundamental categories as well. It would be possible to adopt a different stance on the fundamental social constituents of the phenomena which Brenner is dealing with here; but the adequacy of the various alternative schemes will be determined by the explanatory success of the various research efforts.
Thus Brenner's research falls within the research program of classical Marxism. At the same time, Brenner's is not an orthodox application of classical Marxism to economic history, in that he disputes the primacy of the forces of production (technology) in the historical process, and he denies the existence of largescale laws of motion of a given mode of production.  In particular, he denies that feudalism had one central set of tendencies of development which were entailed by the structure of the feudal mode of production. Instead, feudal property and class relations showed variation across the face of medieval Europe, and these differences gave rise to significant differences in economic development. Guy Bois raises this criticism against Brenner, and Brenner responds in the following terms.
Bois, writing from an explicitly Marxian viewpoint, makes a somewhat analogous charge. Mine is a `political' and a `voluntarist' Marxism: a preoccupation with the vagaries of the class struggle prevents me from discerning the economic `law of motion' of feudal society--in his view, `the falling rate of feudal levy'. (Brenner 1982:28)
Where Brenner's account diverges most sharply from orthodox historical materialism is in the primacy that he accords to the historical development of different classes and their respective political capacities. Orthodox Marxism implies that the political power of a class is a more or less direct function of its economic position within the set of property relations.  Against this view, Brenner's analysis gives the process of class formation and class power a substantial degree of autonomy from economic institutions. The relatively autonomous processes by which class structures were established, developed and transformed have to be placed at the centre of any interpretation of the long-term evolution of the pre-industrial European economy (Brenner 1982:17). The fact that northern French peasant communities shared a history of collective political action which extended over several centuries cannot be inferred from the material and structural circumstances of this group; it is rather an independent factor which has important implications for the career of this class in medieval history.
It is worth noting that there is a surprising degree of convergence between Brenner's neo-marxist analysis and analysis of economic development grounded in neo-classical political economy. A clear example of the latter approach may be found in North and Thomas (1973). These two approaches share the assumption that it is the incentives and opportunities defining the circumstances of action of the individual decision maker (cultivator, inventor, manufacturer, merchant) that determine largescale patterns of development. And they share the belief that the motive structure of individuals is largely determined by their material interests (income, property rights, security). Thus North and Thomas write, Most people prefer more goods to fewer goods and act accordingly. . . . We therefore fall back on the explanation that if a society does not grow it is because no incentives are provided for economic initiative (North and Thomas 1973:1-2). They give extensive treatment of free rider problems, the need for secure property rights to protect the fruits of innovation for the innovator (including patent and copyright laws; North and Thomas 1973:3-4), the centrality of bargaining powers of various groups within a given social system (12), and so forth. 
This parallel is particularly striking when we compare these types of explanations with those associated with the substantivist school. Karl Polanyi's Great Transformation is a paradigm for this school. Polanyi argues militantly against the validity of applying the concepts of economic rationality, profit maximization, exchange relations, and the like, to premarket societies.
The outstanding discovery of recent historical and anthropological research is that man's economy, as a rule, is submerged in his social relationships. He does not act so as to safeguard his individual interest in the possession of material goods; he acts so as to safeguard his social standing, his social claims, his social assets. He values material goods only in so far as they serve this end. Neither the process of production nor that of distribution is linked to specific economic interests attached to the possession of goods; but every single step in that process is geared to a number of social interests which eventually ensure that the required step be taken. . . . The economic system will be run on noneconomic motives. (Polanyi 1957:46)
No society could, naturally, live for any length of time unless it possessed an economy of some sort; but previously to our time no economy has ever existed that, even in principle, was controlled by markets. . . . Gain and profit made on exchange never before played an important part in human economy. (Polanyi 1957:43)
Instead, Polanyi's account requires that the analysis pay primary attention to patterns of reciprocity and redistribution, shared values, traditions, and the determining role of community and politics.
For, if one conclusion stands out more clearly than another from the recent study of early societies it is the changelessness of man as a social being. His natural endowments reappear with a remarkable constancy in societies of all times and places; and the necessary preconditions of the survival of human society appear to be immutably the same. (Polanyi 1957:46)
In place of economic rationality and the market mechanism providing the basis for organization of the premarket economy, Polanyi argues that communitarian patterns of organization are to be found:
The premium set on generosity is so great when measured in terms of social prestige as to make any other behavior than that of utter self-forgetfulness simply not pay. . . . The performance of all acts of exchange as free gifts that are expected to be reciprocated though not necessarily by the same individuals--a procedure minutely articulated and perfectly safeguarded by elaborate methods of publicity, by magic rites, and by the establishment of `dualities' in which groups are linked in mutual obligations--should in itself explain the absence of the notion of gain or even of wealth other than that consisting of objects traditionally enhancing social prestige. . . . But how, then, is order in production and distribution ensured? . . . The answer is provided in the main by two principles of behavior not primarily associated with economics: reciprocity and redistribution. (Polanyi 1957:46-47)
Finally, Polanyi identifies the same element of materialist rationality in common among neoclassical political economists and Marx. Polanyi concurs with the observation above that Marxism analyzes the historical process in terms of individual self-interest, conceived largely in terms of material well-being.
There is the equally mistaken doctrine of the essentially economic nature of class interests. Though human society is naturally conditioned by economic factors, the motives of human individuals are only exceptionally determined by the needs of material want-satisfaction. That nineteenth century society was organized on the assumption that such a motivation could be made universal was a peculiarity of the age. It was therefore appropriate to allow a comparatively wide scope to the play of economic motives when analyzing that society. (Polanyi 1957:153)
All this should warn us against relying too much on the economic interests of given classes in the explanation of history. Such an approach would tacitly imply the givenness of those classes in a sense in which this is possible only in an indestructible society. (Polanyi 1957:155)
The connection between Polanyi's observations here and the terms of the moral economy debate are clear, as is the continuity between his account and the more current arguments of Michael Taylor. 
The main methodological conclusion which may be drawn from Brenner's arguments concerns the need for analysis of the microprocesses of social change, and the social structures and institutions which constitute the local environment of largescale economic and social processes. Brenner shows in one important case that the local structures can fundamentally alter the largescale causal factor (diverting the process from a stimulus to development toward stagnation or involution, for example).
The chief substantive contribution of Brenner's arguments is the centrality of a particular kind of local structure: the relations of class power and interest which are determined by existing property relations, on the one hand, and traditions and practices of joint political activity, on the other.
Anderson, Perry. 1974. Lineages of the Absolutist State. London: New Left Books.
Aston, T. H. and C. H. E. Philpin, eds.. 1985. The Brenner Debate: Agrarian Class Structure and Economic Development in Pre-Industrial Europe. Cambridge: Cambridge University Press.
Berg, Maxime. Origins of Factory Production.
Bloch, Marc. 1967. Land and Work in Medieval Europe: Selected Papers. London: Routledge & Kegan Paul.
Bois, Guy. 1978. Against the Neo-malthusian Orthodoxy. Past and Present 79 (1978):60-69.
Brenner, Robert. 1976. Agrarian Class Structure and Economic Development in Pre-Industrial Europe. Past & Present 70 (1976).
Brenner, Robert. 1982. The Agrarian Roots of European Capitalism. Past & Present 97 (1982).
Cipolla, Carlo M. 1976. Before the Industrial Revolution European Society and Economy, 1000-1700. New York: W. W. Norton.
Cohen, G. A. 1978. Karl Marx's Theory of History: A Defence. Princeton: Princeton University Press.
Le Roy Ladurie, Emmanuel Trans. by ?. 1974. The Peasants of Languedoc. Urbana, IL: University of Illinois.
Little, Daniel. 1986. The Scientific Marx. Minneapolis: University of Minnesota Press.
North, Douglas C. and Robert Paul Thomas. 1973. The Rise of the Western World A New Economic History. Cambridge: Cambridge University Press.
Polanyi, Karl. 1957. The Great Transformation. Boston: Beacon Press.
Popkin, Samuel L. 1979. The Rational Peasant. Berkeley: University of California.
Postan, M. M. 1975. The Medieval Economy and Society. Harmondsworth: Penguin Books Ltd.
Sabel, Charles F. and Jonathan Zeitlin. 1985. Historical Alternatives to Mass Production: Politics, Markets and Technology in 19th Century Industrialization. Past & Present 108 (1985):133-176.
Taylor, Michael. 1982. Community, Anarchy and Liberty. Cambridge: Cambridge University Press.
 Marx's discussion of primitive accumulation in Capital I represents a classic interpretation of this process.
 Perry Anderson, Lineages of the Absolutist State.
 Brenner refers to this set of explanations as a "neo-malthusian" model (1976:33).
 Postan, Le Roy Ladurie; see also Cipolla.
 This is an issue which Perry Anderson confronts in Lineages of the Absolutist State.
 There has been an interesting debate over a parallel issue in recent years: the degree to which capitalist factory production was economically superior to artisan production. Some authors have held (e.g., Zeitlin, Berg, etc.) that artisan production was more flexible and potentially equally efficient; but that factory production served class needs (in particular, the need to control labor costs and labor discipline) better than artisan production.
 "It has been convincingly shown by Eleanor Searle and others that technologies capable of significantly raising agricultural productivity by means of relatively large-scale investments were indeed available in medieval Europe--and they included some of the central components of what was later to constitute the agricultural revolution of the early modern period" (Brenner 1982:33). This point parallels other periods in the history of technology, for example, the water wheel. The technology of the water wheel was known to the Romans; but because Roman agriculture and manufacturing was dominated by slave labor, there was little incentive towards developing labor-saving power technologies. See Marc Bloch's essay on the development of the watermill in the middle ages in Europe (Bloch 1967:136-68).
 We might note that there is other data available as well which might be relevant: patterns of distribution of various religious denominations, patterns of previous Roman habitation, distribution of centers of higher learning, and the like. So selecting the information we have selected implicitly represents a judgment about the likeliest causes of the developments we are concerned with.
 This is so because there is a relatively small number of relevant factors and a small number of regions, so we may expect that many of the relevant factors will covary appropriately.
 Brenner attributes this claim of unique determination to Postan: "the demographic interpreters have been more or less compelled to make surplus extraction or class relations a dependent variable in their population-centred models" (1982:21). If this were the case, then the demographic interpreters would be asserting that "P causes Q and R which in turn cause Z."
 In The Scientific Marx I argue at length that this convergence between classical political economy and Marxism is to be found within Marx's own system; chapter 5.
 Guy Bois's concept of the "falling rate of feudal levy" represents this sort of Marxian interpretation of the laws of motion of feudalism. See also the Dobb-Sweezy debate and Rodney Hilton, ed., The Transition Controversy.
 See Cohen (1978) for a forceful exposition of this view.
 It sometimes appears that North and Thomas commit the sort of error which critics of formalism (Dalton, Polanyi, etc.) criticize in economic anthropology and history quite generally: the careless importation of the concepts of capitalist market institutions into medieval social life and economy.
 "There is the equally mistaken doctrine of the essentially economic nature of class interests. Though human society is naturally conditioned by economic factors, the motives of human individuals are only exceptionally determined by the needs of material want-satisfaction. That nineteenth century society was organized on the assumption that such a motivation could be made universal was a peculiarity of the age. It was therefore appropriate to allow a comparatively wide scope to the play of economic motives when analyzing that society" (Polanyi 1957:153).