Froemming, Steven John; 1999; Rational Choice and Collective Action in an Andean Community. Ph.D. Thesis, University of Washington, Seattle. Excerpt from Chapter 13 (pages 755-770).

Communalism as a Value, the Value of Communalism

Urton (1992) raises the question of whether or not the values and practices informing the cosmology of people in his research site of Pacariqtambo, and by extension, in other communities of the Andes, also represent organizing principles of social organization, and if so, to what degree. He proposes that communalism and differentiation are central values and organizing principles of social life, and provides a rich and provocative discussion of how these values get played out in phayna activity. As he puts it, "it is the dialectical relationship between communalism and differentiation which . . . motivates work by everyone in the community on certain tasks, but which at the same time insists that such tasks should be performed by people working in their different ayllu groupings" (Urton 1992:231). Accordingly, he concludes that "These two values are in constant dialectical relationship with each other in the life of the village."

The question raised by Urton is of broad interest, given that all too often in the Andean literature communalism as intent (norm, value, worldview) is confused or conflated with communalism as institution (ayllu, comunidad campesina) or communalism as action (people working together in ayni, etc.). The first presupposes that cooperation is an ethic, a preferred way of doing things. The second includes rule-making and rule-enforcing mechanisms that go beyond moral incentives to ensure cooperation. The third is measurable cooperative behavior, however it arises and is maintained.

In a broad sense, Urton is most certainly right that in the southern Andes of Peru "relations among individuals and groups are premised on both communalistic and individualistic values and motivations, and that the practices through which socioeconomic and political relations are conducted daily and reproduced over time are products of the dialectic . . . between communalism and differentiation" (Urton 1992:233). The challenge is to make sense of what this might mean in terms of individual thought processes, and how they become aggregated among the whole.

Care needs to be taken in how we define communalism as a value. If we make it context specific, as Urton sometimes seems to do, then the issue of why certain pieces of property are owned collectively and certain tasks are participated in jointly become part of the definition -- rather than part of what needs to be explained (why the value is held in these but not other instances). The same can be said of differentiation. Indeed, it is unclear in what sense differentiation is a "value" in addition to being a condition or a process. The more obvious meaning of this concept -- that people value certain differences and wish to preserve them (they seek to maintain a separate identity) -- may make some sense in the case of ayllu affiliations, but is less intelligible in the case of class. As the term is normally used, the hallmark of differentiation is not that people value difference (as in a multiculturalist or postmodernist sense), but that they value different things.⁽¹⁾

The two principal forms of differentiation identified by Urton in Pacariqtambo are ayllu-based and class-based. The former are traditional, moiety-linked, sociopolitical landholding groups within which members have inheritance and usufruct rights, while the latter is associated with divergent economic interests, the product of the relations of production in an agropastoral village economy linked to regional, national, and global economic orders.⁽²⁾ Used in this way, differentiation becomes a catch-all category for several values rather than an identifiable value in itself. It seems to include, but not be limited to, "individualistic values."⁽³⁾ The fact that the comuneros can be differentiated by class or ayllu status is an indication that they have separate interests, but saying that the comuneros value differentiation does little to explain why individuals cooperate or not to promote these interests in ayllu and class groupings. In any case, a focus on values is inherently incomplete. Caught up in the "dialectic," we can easily miss the role played by opportunities and constraints in peasant decision making.⁽⁴⁾

While I think that an analysis of campesino behavior in terms of values potentially provides a better descriptive basis for understanding the mechanisms involved in actual choice decisions than the "as if" assumptions typical of most "rational actor" analysis, it seems to me that we are far from the point of understanding how these values are weighed and combined in actual decision making. In particular, much, if not most, of what we call communalism in the Andes requires little reference at all to communal values for its explanation. Communal social organization can just as readily be an outgrowth of enlightened and constrained self-interest as it can be of communal values. As Gonzáles de Olarte puts it,

What's true for common property and public works appears to be true for reciprocal labor as well:

The decision rule that Andean peasants appear to follow in labor recruitment goes family labor > contractual cash labor > exchange labor (where ">" means "is preferred to). Guillet concludes that, "the major variable explaining the choice of wage or contractual labor over reciprocal labor is cash."⁽⁵⁾ So when we talk about ayni or mink'a, here too cooperation is more of a practical solution to the problem of making a living than it is an ideological or moral attitude.

As commonly used, the term "values" includes two quite different motivators. We may value an action because of its effects. Or we may value an action in itself, without regard to its effects. The former preferences fall within the domain of rational action, which is always concerned with outcomes. I've followed Elster throughout this thesis in referring to the latter preferences as norms. Given what we know about the constraints under which different forms of property rights and labor recruitment are chosen, it appears that collectivism, communalism, and cooperativism in the Andes are primarily values of the first sort. If we are to understand when these values are applied and when not, it is wiser to follow Gonzáles de Olarte (1984) in focusing on the "efecto comunidad" ('community effect'), than to get caught up in positing Andean "isms."

What are the potential effects of extra-household cooperation? Gonzáles de Olarte identifies three: a) more or better production, accompanied by the reduction of average costs, b) higher incomes than those that can be reached independently, and c) the greater well-being of households in the community and the community as a whole (as measured in health, education, and other locally-determined indicators). Ultimately, the election between individual action and collective action depends upon the continual evaluation and comparison, through experience, of one or another of these forms in terms of these more specific goals.⁽⁶⁾

Despite the impressive range of cooperative action domains in the Andes, the household rather than the community is at the core of most land use and production processes, so we must be careful in speaking of the "centrality" of communalism as a value to not confuse its cross-cultural salience with its relative internal cultural weight within Andean belief systems. Rather than contrasting communalism and differentiation, which leaves us open to conflating preferences and processes, I find it more straightforward to contrast the processes of collectivization and individualization. While this way of framing things perhaps makes it harder to fit intermediate organizational entities such as ayllu into the picture as Urton tries to do,⁽⁷⁾ it allows us to address changing opportunities and constraints, without detracting from the search for the role played by collectivist and individualist values in these processes.

Once opportunities and constraints are brought into the picture, we are in a position to note that the choice between individual and collective control of property and organization of tasks -- and consequently, the tendency toward individualization or collectivization -- is closely related to the relative and absolute quantity and quality of resources, and the size of the community (Gonzáles de Olarte 1984).

The first can be readily seen by contrasting the degree of collective decision-making pertaining to different production zones in Ccachín (recall Chapter 2). The pattern, as Cotlear (1989:47) notes, is that privatization proceeds progressively by zones in a community. The best land is the first and most extensively privatized. The observation that there is a "consistent coincidence that reciprocity, family or kin organization, and even the community, in general, have greater force among the poorest and underdeveloped campesinos" (Sánchez 1987:197) is commonplace in the Andes. There appears to be a tendency to move away from collective solutions as the value of resources increases. The fact that reciprocal labor in the Andes has persisted despite trends toward commercialization, monetization, and proletarianization has been related by some investigators to a shortage of cash (Guillet 1980).

Gonzáles de Olarte (1984:238) observes that where the sale of temporary labor outside of the community is the principal source of income, comuneros abandon participation in common property regimes as their net income exceeds that reachable with the "efecto comunidad." I've already noted that given the coincidence of temporary labor migration and phayna activity in the annual calendar, comuneros in Ccachín often find it advantageous to choose the former over the latter and meet their obligations to the community by paying off the resulting fines. When labor for communal projects itself becomes monetized by development agencies that include provisions for the payment of daily wages to workmen hired from within the community, an incentive is created for the comuneros to withhold their free phayna labor to pressure the agency to come up with the funds to pay wages for work on other projects as well (Urton 1990).⁽⁸⁾

With regard to the second factor noted above -- the size of the community -- Gonzáles de Olarte (1984:231) notes that "in communities composed of around 100 families, that is, some 500 to 600 people, the efficiency of the collective functions much better than in communities composed of 800 to 1000 families." Unfortunately, the effects of community size on communal labor organization in the Andes have not been carefully studied by anyone as far as I am aware, so this conclusion appears to be based more on theoretical expectations or on sweeping impressions than on comparative data. There is nothing empirically or theoretically exceptional about these numbers, and in fact, they tell us little about communities that consist of between 100 and 800 families, a very wide range, and one in which the villages studied respectively by Urton and me fall.

Gonzáles de Olarte suggests that the inefficiency of large community sizes for communal management and organization in good part explains the dynamic toward dismemberment of large communities and the desire for recognition of a separate corporate status by comuneros living in the anexos and parcialidades, but to my mind, this can just as readily be explained by the tendency for the personal benefits of many communal projects to decline the larger the group is and the further that one is located from the village center, and by the opportunity to compete for government resources as equals once official status is recognized. The theoretical expectation that larger groups will fail to further their own interests is largely based on two factors a) the individual incentive to free-ride increases with group size, and b) the larger the number of people to be coordinated, the higher the costs of organizing them to an effective level (Olson 1965). Gonzáles de Olarte seems to primarily have the second in mind, but the expectation that smaller groups will do better than large in collective provisioning likewise follows from the first. The affects of increasing group size, however, are sufficiently complex that we cannot assume that increasing the size of a given group increases the probability of its failure. This depends on other factors as well.

Hardin (1982) emphasizes that as size increases, average cost, total cost, level of supply, or individual valuations also change, so that all else is rarely equal, as assumed by such analyses. Many projects may be too costly to yield net benefits for a small group, but are within reach of large. In some cases, participation may even follow an inverted U-shaped curve. In a study of union participation, Lipset, Trow, and Coleman (1956), for example, found that the highest level of average participation occurred when shop size was around 200 members. Above and below that size, participation fell off. While the explanation offered for this -- that there may need to be enough people to provide the social context for interlocking friendships to lead to solidarity ties -- would hardly seem to apply to Andean peasant communities, where kinship and compadrazgo create interlocking networks in groups of all sizes, my point that the overall effects of group size on participation in community activities is context specific still holds. Accordingly, I do not want to stray too far from the empirical record. Ccachín provides a small natural laboratory for studying this, as it offers the potential to compare participation rates in community assemblies, phayna, and fund-raising assessments between the mother community ( about 225 households) and its two anexos (about forty and thirty households respectively). Communities such as Pacariqtambo, with several ayllu of varying sizes likewise allow for such comparisons. The research, however, remains to be done.⁽⁹⁾

It's time to bring this discussion back to the question of values. Talking about the dialectic between the values of communalism and differentiation seems to me to be a shorthand way of talking about a process in which values are being traded off against each other. By asking people, or by observing their behavior, we can discover what their preferences are. Rarely are we or our subjects in a position to methodically discuss why they attended a phayna or not, or why they chose one form of labor recruitment over another, in a way that would yield candid information. So we mostly rely on proxy indicators. As Urton puts it, "from the point of view of an outsider (such as an anthropologist) -- the authentication of the beliefs and practices of a particular society must be constructed by comparison between present-day remembrances, as recounted in oral tradition and enacted in social practices, and accounts contained in historical documents" (1990:230). This takes us some distance away from what goes on in actual people's heads, so that "what is considered to be persistent in a particular sociocultural tradition is an ideological and analytical construct" (ibid.). If public discourse and remembrance tend to be distorted in any systematic way, say toward the expression of aspirational norms, appeals to the common good, or the invocation of social norms to rationalize self-interests, our problems are compounded, but in most cases, it's the best we've got. Our recognition of the centrality of "collectivism" or "communalism" in Andean belief comes about in part through evidence of this sort.

The other way our understanding of communal values comes about is by inferring them from actual behavior. This is a bit more complicated than deducing the values campesinos place on consumer goods by the quantities of coca leaves, coffee, and wool exchanged in truki ('barter'). It's complicated because we're not so much interested in the value people place on goods or ends, but on the means they choose to their ends. Rather than determining the value they place on roads, running water, and schools (which can't be readily determined either from behavior because of the possibility of free-riding on the goods provided by others), what we are trying to get at is how much people value communal association or cooperation in and of itself, and then, how much this accounts for their propensity to cooperate. Were we to apply the same question to exchange (a more limited kind of cooperation), we would be trying to determine not how much people value coca, coffee, and wool but how much they value exchange itself.

While this would seem to put us in the realm of expressive rather than outcome-oriented rationality, this is not entirely the case, because as Guillet indicates, the rationality of labor recruitment decisions in the Andes involves the evaluation of social as well as material costs:

Urton is even more precise in describing the costs and mechanisms involved:

In this way, social costs tie back to material ones, and communal values are sustained by informal inducements and sanctions.

There are several lessons in this. First, one must look beyond immediate incentives to explain why members behave as they do (a point made for social institutions generally by Hayek 1973). Second, where others' future choices are contingent on one's own immediate choice, and noncooperators face the threat of subsequent defection by others in return, cooperation may be the most rational and strategic choice (Hardin 1982). Finally, the overlapping nature of activities and groups in Andean communities allows cooperative conventions to arise, and provides opportunities for low-cost sanctioning of noncooperators.

This last is probably the most interesting point, because while the two citations above focus on the costs to the sanctioned, the more fundamental issue for explaining cooperation is the cost to the sanctioners. It's an issue I took up in Chapter 6 and which bears repeating here: punishment itself is an investment in the production of a public good -- the amount of cooperation in the group -- and thus raises a second-order collective action problem. It's clear that if enough members of a group choose to punish, rational individuals will be better off if they cooperate. If there are costs to participating in the punishing, the rational, self-interested individual will let others do the punishing. If all face the same choice, no one may punish, and the motivation for defectors to cooperate vanishes. The overlapping nature of activities and groups in peasant communities, however, reduces the costs of sanctioning in at least two ways. First, the comunidad campesina creates and manages, though the cooperation of its members, not only collective goods from which exclusion is impossible (roads, improvements to the church, etc.), but also goods for which exclusion from the benefits is possible (redistribution of land in the agricultural sectors, etc.). Second, there can be a spill-over effect as noncooperators in the public sphere are excluded when comuneros' choose partners for ayni, mink'a, and truki exchange. The costs of such exclusion need not be high. As Hardin puts it,

The net costs of sanctioning, of course, will vary depending on the relationship distance and the power of the sanctioners relative to the sanctioned, but for the average community member, need not be appreciable. More than once I was told not to attend someone's fiesta or go to their work party because they were millay runa or mala gente ('a bad person'), and I assume that the comuneros that advised me in this way often followed their own advice.

Hopefully, the foregoing has revealed some of the problems in determining the coincidence between the ideological or cosmological aspects of cooperation in the Andes and the institutional or behavioral aspects. The former can't be automatically read from the latter. If the contrast between communalist and individualist values is to have any explanatory clout, the framework must be able to account for the synchronic and diachronic patterns of collective action that one sees in Andean communities, and clearly distinguish the motivations and mechanisms. Unfortunately, the ideational content of communalism remains remarkably under-defined. If the arguments that communalism is a "central value," that it "motivates work by everyone in a community on certain tasks," and that there is a "tension or contradiction between communalism and the self-interested actions that promote socioeconomic differentiation" are to have any theoretical value, then it seems to me that communalism must be defined as a commitment to a set of positive moral codes or norms upholding the value of community. This formulation has the advantage of pointing to a set of mechanisms that are distinct from rational choice, and that go to the essence of communality, "the feeling or spirit of cooperation and belonging arising from common interests and goals" (Random House 1996).⁽¹⁰⁾

In principle, whether or not communalism in this sense undergirds cooperation in specific contexts or is an important aspect of social life in general in Andean communities is an empirical issue. In practice, it is difficult to sort out the various mechanisms leading people to cooperate. Self-interest, altruism, and norms of cooperation can coexist and reinforce each other at both the individual and population levels, so we need to find ways of teasing them apart for analysis. The most interesting cases for pursuing the question are those where people cooperate despite material or social incentives not to. Phayna participation is a better domain to investigate this than ayni or mink'a, since typically there is an incentive to free-ride.⁽¹¹⁾ Exactly, because of this possibility, however, the comuneros use a host of methods to detect noncooperators, reduce the opportunity costs of participation, increase the costs of nonparticipation, and provide private benefits for participants, methods that I have described at length in this and previous chapters: attendance lists, pawns, fines, scheduling to avoid time conflicts, token payments of food or seed to attenders, provision of alcohol, chastisement with loathsome tasks, membership probation, and the conditioning of land redistributions and other benefits controlled by the community on participation. Given such methods, it would appear that Urton is not far from the mark when he says that:

But there is more to it than this. Clearly, if fines are set high enough, cooperators always will do better than noncooperators. But if fines are set too low or inadequately enforced, there are situations where both universal cooperation and universal noncooperation are equilibria, and whether cooperation is achieved or not depends on the extent that people can count on each other's cooperation (Elster 1989:153-154 illustrates the payoff structure for such a situation). Fines in Ccachín, recall, are set by the pool of potential phayna participants themselves, and rarely exceed the prevailing day-wage rate at the moment that they are set (as I indicated in Chapter 9, the actual rate faced by a phayna participant is usually much less because no adjustments are made over time for inflation). The other point to emphasize here is that sanctioning itself is a second-order public good and equally subject to underprovisioning, but unsupported by the mechanisms listed above. Elster goes so far as to argue that if norm-guided behavior is only sustained by external sanctions, it is not an equilibrium, and therefore, some sanctions must be performed for motives other than the fear of being sanctioned (Elster 1989:120).

So it pays to take another look at communalism, to identify the specific patterns of thought that might be associated with it, and the specific patterns of behavior that follow from these thoughts. There are at least two kinds of norms that follow from a sense of community or a sense of belonging: those that motivate a person to cooperate, and those that motivate a person to reproach others who do not. To the extent that establishing a system of effective sanctions itself requires cooperation, the second is a variety of the first, but the kinds of sentiments they arouse may be different. They are also logically distinct, because what one expects of others may be stricter or more lax than what one expects of oneself.

There's little evidence that the norm "cooperate whenever the community will benefit (as determined by the assembly)" translates into personal action in Ccachín, although it is the statutory convention (the agreed upon aspirational norm), and legitimatizes the assessment of fines. The relatively low average turnout at assemblies and communal work parties belies the effectiveness of this norm, as does the developed system of sanctions intended to assure compliance.⁽¹²⁾ In this sense we can say that most people in the village are not Kantians, if Kantian-like beliefs motivate people to do X, when X is an activity that would benefit all if all engaged in it. This is not to say that no one follows categorical imperatives in Ccachín -- something that I don't have the individuating evidence about personal beliefs and behavior to determine -- but only that such motivations are not sufficiently strong to compel universal cooperation in any collective endeavors in the village. It's important to note, however, that many who do not have stellar attendance records at phayna do appear to pay off all of their fines for absences. It may be that paying off the fines has come to be seen as a fair alternative to phayna participation by many comuneros rather than as stigmatized. If full phayna participation can be seen as an aspirational norm, the pragmatic norm -- which the comuneros expect each other to obey, and which elicits disapproval if violated -- may be to participate a majority of the time, and to always pay one's fines when requested.

To the extent that the comuneros are motivated to cooperate by normative considerations, their behavior suggests that they must operate by norms that are conditional rather than without exception, norms that are sensitive to circumstances. Given the various interdependencies of social life, the most important of these circumstances is what others are doing, and any behavior guided by such norms will tend to be frequency-dependent. There are two distinct kinds of non-selfish norms to consider here, those that are sensitive to outcomes, and those that are not. Knowing that not everyone will participate, people of the first kind feel a sense of duty to contribute more than their "fair" share. Consequently, their behavior will be characterized by the pattern of cooperating more when others give less, and cooperating less when others give more (assuming that some amount of the good can be provided by their participation, and that there is increasing marginal utility to participation after a certain point is reached). Norms of fairness likewise take into consideration what others are doing, but they lead to the opposite pattern of outcomes: people cooperate more when others can be counted on to participate as well, people cooperate less when others participate less.

One can see that norms of fairness are insensitive to outcomes if one considers the example of someone who has the possibility and the material incentive to provide some of the collective good alone but refrains to do because others fail to participate. To the extent that they operate by a norm of fairness, comuneros would tend to adjust their participation rates over time to the statistical norm. As Margolis (1990:825) puts it, "If social motivation exists at all, we can expect some behavioral propensity which ordinarily functions to limit the extent to which members of the community are likely to act socially much beyond what is typical of the rest of the community." In this way, norms of fairness involve a balancing act to maintain a personal position of being "neither selfish nor exploited" within a the social field (Margolis 1982, 1990).⁽¹³⁾

If a community is characterized by individuals who have the gamut of non-selfish motivations described above, there can be a dynamic aspect to cooperation: participation by those that want to do what's best for all if all would do it (those insensitive to outcomes and circumstances) could trigger the cooperation of those with the propensity to cooperate as long as each new act of cooperation increases the average benefit (those sensitive to outcomes and circumstances) which in turn could catalyze those who are motivated by the norm of fairness (those sensitive to circumstances but not outcomes). As Elster (1989:133) notes, "For each of the latter, there is some number of other cooperators who will trigger off his cooperation. Some are easily shamed into cooperating, whereas others come around only when almost everyone has joined. . . . Depending on the constellation of motivations, the chain reaction may go all the way to universal cooperation or stop short of it." Unfortunately, however, as Elster also points out, "Next to nothing is known about the distribution of these motivations in the population and the way in which they interact to produce decentralized cooperation" (1989:134).

The issue is complicated not only by the methodological challenges of documenting micromotivations, but by the fact that the same outcomes can occur whether individuals always follow the same rule with variance between individuals, or all follow each of these rules a certain percentage of time. The closest I can some to identifying such a phenomena in Ccachín is with the dances, parades, serenades, and other public spectacles that take place in the central plaza. Most people are in a position to see how many others have assembled in the plaza from their homes. There is a cost -- wasted time (and perhaps lost prestige) -- to arriving too early, but events will not start until a sufficient number have arrived. The timing and sequence of arrival, as well as the total number of attenders appears to be determined in good part by what others do. The situation is tailor-made for observing the chain reaction, because an observer can readily see the comuneros monitoring who and how many are in the plaza while they continue with their other activities. One can hypothesize that a combination of word of mouth, visual monitoring, and adjustment of one's personal behavior to the attendance of others at previous phayna leads to a similar phenomena occurring at assemblies at communal work parties, but if true, it's decentralized nature and time breadth makes it much harder to establish.

Much of the above rests on plausibility, but it points out the importance of investigating micromotivations if we are going to explore Urton's intuition that communalism is a central value in Andean life in a systematic way. To give a bit more idea of the difficulty in determining which, if any norms are operative, let me point out that the behavioral consequences of each of the norms discussed above can come about for other reasons. A pattern of some giving more when others give less is often characteristic of chicken games, where each participant has a self-interest in providing some of the public good alone, but would prefer not to contribute when enough others do. Those that are able to precommit themselves to noncooperation effectively force the rest to participate (Taylor and Ward 1982). Arguments about fairness in this case could be used to establish precommitments. In some cases, it may not be self-interest in the good itself, nor a spirit of community that prompts individuals to participate, but a sense of superiority or spite. As well, norms of fairness may be hard to distinguish from conformity rules in general. Just as one may conform to the majority, to the leaders, or to the most successful, norms of fairness may likewise pay attention to the same focal individuals. The comuneros may be more likely to participate if community officials do (this effect is distinct from any contribution the latter might make in facilitating communication, providing direction, etc.). Conversely, they may be unwilling to participate if they see that the wealthiest members of the community (who often benefit the most from the public good) do not participate. Finally, we should not forget that if someday we are able to sift through all of this and determine the distribution of the various non-selfish norms in a population posited here, the explanation of this distribution becomes a paramount task.

Notes

1. As the term is normally used, differentiation refers either to the perception of existing differences, or to a process where entities become unlike or dissimilar (Random House 1996). It should be noted that the first is the polar opposite of an emphasis on likenesses, as this latter is picked up and elaborated upon by Radcliffe in her discussion of ritual attempts to emphasize homogeneity among community members (see Chapter 12). This raises a question in my mind about the difference between communalism and homogeneity on the one hand, and differentiation and heterogeneity on the other (as well as the difference between these features as fact and as ideology). If one were to begin to sort out the distinct and overlapping meanings of these terms, I would add collectivization and individualization to the semantic stew. Some of these values and processes are opposites, some are not. If the terms are to be analytically useful in this context, they need to be clear.

2. It's not clear to me whether Urton considers both self-interested motivations and economic differentiation to be a product of peasant relations of production, or whether he takes the former to be a constant, which under the latter are characterized by conflict. This uncertainty is not critical to the discussion here, but would be so if the two communities were studied diachronically, or were contrasted according to the degree in which they are integrated into capitalist markets. I bring it up here in an effort to be as faithful as possible in summarizing his position.

3. Interestingly, one of the other things it seems to include is the dictionary definition of communalism. The most intriguing for our purposes here is the third definition in Webster's Unabridged, the "strong allegiance to one's own ethnic group rather than to society as a whole", but the first, involving political federation, will do almost as well (Random House 1996). If we take the comunidad campesina as a society in microcosm, then allegiance to one's ayllu is at once a kind of differentiation and communalism.

4. I'm not quite sure how to interpret Urton's use of the term "dialectical" to describe the relationship between communalism and differentiation. In the most general sense, he may simply mean that they are ideas or forces that are interconnected or juxtaposed. In a more provocative but also more questionable sense, he may be trying to point to some contradiction between the two requiring synthesis. I am also unsure whether Urton places the locus of the dialectic in the minds of the comuneros as individuals or in the culture as an abstract whole. There are other ways of talking about essentially the same processes. Margolis (1990), for example, extends the concept of equilibrium analysis of standard economic theory to norms and values to suggest that the social equilibria characteristic of the whole are both contingent upon and influence individual equilibria, the latter which reflect a balance between competing motivations at the personal level. No doubt, the choice of vocabulary and metaphors we use to describe things depends on our audience, and I have no problem with that. My only concern is that we choose a language that is as precise and unambiguous as possible to describe and distinguish psychological and social processes, and for the most part, I find reference to "dialectical relationships" unhelpful in this regard.

5. The explanation for this preference structure is that wage labor is more convenient than reciprocal labor since it reduces the search costs of finding individuals willing to enter into the exchange, the need to pay back labor at a later date, and the problem of adjusting for differences in the size of land holdings (Guillet 1980).

6. In saying this, I am in agreement with Guillet when he writes that "there is no tendency to cooperation, reciprocity, or association, if by this is meant adherence to practices found in past societies" (1980:159). While what worked in the past may continue to serve as a pattern for action under conditions of uncertainty, and the prior investments made in constructing traditional institutions may be used as resources in the present, to the extent that the comuneros are able to anticipate the costs and benefits of their actions, behavioral continuities need to be related to the constraints leading to the recurrent selection of one form of action or another.

7. When contrasted to the comunidad campesina, ayllu memberships may appear to be a kind of differentiation, but it makes little sense to say that they are likewise a kind of individualization (perhaps this is the reason why Urton prefers to use the former term). Taking individuals as our starting point, however, an ayllu can be recognized as a particular kind of collectivity, with a recognizable size and status intermediate between the household and the community. In this sense, the ayllu is as much an indication of communalism as the comunidad campesina is, and the preference to organize the latter in terms of the former may be best explained in terms of nested units rather than dialectics.

8. Urton cites the case of a drainage system being constructed in Pacariqtambo in the mid-1980s. PAIT, a program of employment generation and income transfers instituted while Alan Garcia was in power, paid wages to community members for constructing an addition to the high school and for building a retaining wall around the church, but not for the drainage project. While there may have been other reasons for the lack of community support for the drainage project, Urton suggests that the expectation that wages should be paid for this project too played an important part.

9. Hardin (1982) establishes that the most useful interpretation of the phrase "large groups are more likely to fail than small groups in achieving collective action" is that "If A and B are groups each with a similar common interest, and A is larger than B, then A is more likely to fail than B." When I say that Ccachín (and other Andean communities) provide natural laboratories for studying this, it is not only because the anexos and ayllu are distinct collectivities within a single communal territory, but also because, for comparative purposes, numerous contexts can be found where they meet the condition of having similar common interests.

10. Note that while common interests and goals may enter into the definition of and reason for attachment to the relevant group, once the motivation to cooperate becomes internalized as "spirit" or "sense of belonging," consequences need not be part of the proximate mechanisms of behavior motivated by the expression of this sense or spirit.

11. Given the delayed aspects of reciprocity in ayni, there is the possibility to defection at the ultimate point of exchange, so there is the potential for getting a free ride here too, but the problem is not as pervasive as it is in communal work parties.

12. With regard to this second, however, it is important to note that sanctions are not superfluous even where an ideal is internalized. As Elster notes, "Human nature being what it is, external sanctions are a useful counterweight to weakness of will" (1989:119).

13. The "expectation" arises, for Margolis, from the necessity of grounding the evolution of social motivations in a Darwinian account. Interestingly, the two rules of thumb posited by Margolis to account for the desire to be "neither selfish nor exploited,"

do not say anything about fairness, despite the fact that his concern is with how individuals see their commitment of resources relative to others in a comparable situation. Strictly understood, the relevant circumstance here is what an individual has done in the past, rather than what others have done or are doing. The balancing mechanism between the two, however, lacks an aspiration level, something that reference to the behavior of others provides.

References Cited

Cotlear, Daniel
1989 Desarrollo Campesino en los Andes: Cambio Tecnológico y Transformación Social en las Comunidades de la Sierra del Perú. Lima: Instituto de Estudios Peruanos.

Elster, Jon
1989 Nuts and Bolts for the Social Sciences. Cambridge: Cambridge University Press.

Gonzales de Olarte, Efraín
1984 Economía de la Comunidad Campesina. Lima: Instituto de Estudios Peruanos.

Guillet, David W.
1980 Reciprocal Labor and Peripheral Capitalism in the Central Andes. Ethnology 19: 151-167.

Hardin, Russell
1982 Collective Action. Baltimore: Resources for the Future, by the John Hopkins University Press.

Hayek, Friedrich A.
1973 Rules and Order. Law, Legislation, and Liberty, Volume I. Chicago: University of Chicago Press.

Lipset, S. M., Martin Trow, and James S. Coleman
1956 Union Democracy. New York: Free Press.

Margolis, Howard
1982 Selfishness, Altruism, and Rationality: A Theory of Social Choice. Cambridge: Cambridge University Press.

1990 Equilibrium Norms. Ethics 100: 821-837.

Olson, Mancur
1965 The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA: Harvard University Press.

Random House
1996 Random House Webster's Unabridged Electronic Dictionary. New York: Random House.

Sánchez, Rodrigo
1987 Organización Andina, Drama y Posibilidad. Huancayo: Instituto Regional de Ecología Andina.

Taylor, Michael, and Hugh Ward
1982 Chickens, Whales, and Lumpy Goods: Alternative Models of Public Goods Provision. Political Studies 30(3): 350-370.

Urton, Gary
1990 The History of a Myth: Pacariqtambo and the Origin of the Inkas. Austin: University of Texas Press.

1992 Communalism and Differentiation in an Andean Community. In Andean Cosmologies Through Time: Persistence and Emergence. Robert V. H. Dover, Katherine E. Seibold, and John H. McDowell, eds. Pp. 231-266. Bloomington: Indiana University Press.