14
The Reproduction of the Social:
A Developmental System Approach
Iddo Tavory, Simona Ginsburg, and Eva Jablonka
Biology-inspired ways of thinking about cultural history today are based either on analogies with biological evolution or on the assumption that preexisting genetic predispositions explain the most important aspects of human culture. Far less attention is given to analogies between developmental and cultural processes although it is recognized that both developmental systems and cultural milieus assimilate changes and adjust to them, and both manifest a great deal of stability—what biologists call robustness and what sociologists once called “the problem of order.” Moreover, the recent emphasis that developmental biologists put on open-ended plasticity on the one hand and on processes of canalization on the other suggests that a developmental perspective can be a source of useful insights, that it may be “good to think with,” as Levi-Strauss would have put it.
In this chapter we adopt a developmental system theory (DST) approach, which emphasizes the dynamic, self-sustaining feedback interactions among multiple resources, schemas, and activities that lead to the persistence of a system over time (Oyama 1985; Oyama et al. 2001). DST takes this dynamics as the unit of analysis, focuses on the network of interactions among its constitutive factors that renders it self-stabilizing, and uses it as the basis for analyzing temporal departures from persistence. Our approach, like that of several other authors in this volume, thus emphasizes those aspects of developmental dynamics that constitute the traits of interest, resulting in their stabilization and temporal persistence.
This approach, we argue, allows us to tackle relatively stable sociocultural patterns that would be hard to explain through classical evolution-inspired approaches. To take a prominent example that has occupied sociologists since the emergence of the discipline, urban poverty is a highly complex phenomenon. Poverty is constantly shifting, influenced by institutional macro-organization of economic opportunity and oppression, by the actions of the poor, by interactions they have with people inhabiting other socioeconomic strata, and by organizational interventions that can both alleviate and reinforce poverty. Moreover, poverty also entails a relatively stable intergenerational outcome—low socioeconomic mobility. Even in the most unequal Western societies today, however, poverty is not deterministic—people do move intergenerationally from one socioeconomic category to another. However, mobility is the exception rather than the rule, and most people born into poverty in the United States never leave it (see, e.g., Herz 2006). In other words, remaining poor is a dynamically stable point of canalization people tend to reach if they are born into a specific economic stratum.
A whole set of social factors contribute to the entrenchment, and thus the reproduction, of poverty across generations: from macro-institutional conditions such as prospects restricted by family circumstances and connections, and job opportunities limited to minimum wages and bad schools, to the expectations of peers, parents, and teachers, as well as parents’ own positions and circumstances. Children go to school ill equipped to succeed and face systematic discrimination and constraints when they leave. Even the success of some in escaping poverty often reinforces the difficulty of upward mobility for others, as those who “make it out” usually leave poverty-stricken neighborhoods (Wilson 1987). Other factors include malnutrition or the lack of local markets offering affordable and healthy food. Poor nutrition and exposure to toxins by parents lead to multiple health problems in children that then exacerbate the challenges of overcoming an already disadvantaged educational system (Dasgupta 2009).
Ethnographic studies demonstrate that the urban poor are highly differentiated in the strategies they employ to adapt and improve upon their circumstances. Newman (1999) shows that many residents in blighted neighborhoods attempt to foster a middle-class ethic of work which they put into the jobs available—insecure and minimum-wage jobs. Other studies (e.g., Bourgois 2003) focus on the informal drug economy, showing how drug dealers attempt to realize their dreams of mobility through their dealings. As Bourgois shows, many urban poor go in and out of this informal economy, frustrated by the low wages and insecurities of the formal jobs they obtain, and constantly pulled by the seemingly extraordinary chances of wealth, upward mobility, and respect. Still others try to cope with poverty by constructing “niches” of respectability within it. Thus, for example, Duneier (1999) shows how homeless men and women in New York’s Greenwich Village sold magazines that they pick up from the trash in affluent neighborhoods.
Taking a bird’s-eye view of such studies, it becomes clear that poverty is lived in many different, sometimes contradictory, ways. Although there are macro-institutional similarities in the structural lack of opportunities poor people face, there is quite a lot of variation in the ways that they make sense, act, and interact within their surroundings. However, these different strategies are powerfully canalized with respect to the enforcement of outcomes. Thus, for example, workers in low-paying and insecure minimum-wage jobs remain poor and lack economic and cultural resources to position their children for better outcomes. Despite an apparently quite different strategy, not only are drug dealers in constant danger of being shot by competitors or arrested by the police, most drug dealers actually end up making little more than minimum wages. Although the lives of drug dealers and formal-economy minimum-wage earners may be very different, both ways of contending with their situation are thus canalized into the same general socioeconomic outcome. This does not necessarily mean that without reorganization “from above” poverty will always end up reproducing itself, but rather (in what might be a far more depressing conclusion) that urban poverty is overdetermined by a host of different mechanisms. Such persistence over time, in spite of the variable developmental life trajectories of the individuals that constitute the system, is a typical outcome of historical sociocultural processes.
Such persistent, self-sustaining social systems are not readily amenable to description and explanation in terms of classical cultural-social evolutionary approaches (reviewed in Richerson and Boyd 2005), even when interactions between different social factors and social agents are acknowledged. The four influential genetic-evolution-based models of cultural change described below do not account very well for cases (such as cycles of poverty) with complex, self-sustaining loops of interactions.
1. Richerson and Boyd (2005) develop a dual inheritance theory of evolution and discuss the complementary questions of stability and change, summarizing the results of their mathematical models that show how selection among individuals with different cultural practices may occur and how general psychological and learning biases may affect the spread of such practices. Their analysis—which is couched in terms of differentially successful diffusions of ideas and practices—is difficult to apply to the case of urban poverty. The differential diffusion of ideas and practices which they highlight is, of course, part of the dynamics of urban poverty, but other factors and interactions—such as the behavior of people who make it out, institutional and macroeconomic factors, the discrimination by those outside the system, the effects of alcohol and drugs—are hard to incorporate within this framework, and the way such factors may affect reproduction across generations is not discussed. The assumption that the differential diffusion of ideas and practices can provide a full description of the evolutionary dynamics of a social-cultural system such as the urban poverty cycle seems too narrow.
2. The memeticists offer a framework that regards cultural evolution in terms of the differential replication of “memes” (ideas or practices). Replication is understood as a copying process, akin to the replication of DNA by DNA polymerase, which is unaffected by the history and the functional significance of the replicated sequence. This view ignores the fact that both the acquisition and transmission of ideas and practices are the consequence of processes of socio-developmental, context-sensitive construction rather than history- and function-blind copying. As the studies of urban poverty we discussed above clearly show, there is no sense in which poverty is “replicated” or “copied”: it is constructed during the developmental cycles of individuals and groups in the community (Jablonka and Lamb 2005). Social construction and context-dependent interactional processes, not replication, are at the core of the reproduction of poverty.
3. Evolutionary psychologists and cognitive epidemiologists, such as Sperber, take a more cognitive, innatist view. Sperber’s argument is that humans’ dispositions to acquire knowledge and act in the social world are structured in innate domain-specific ways that lead to the stabilization of cultural practices at the social level. Sperber suggests that what evolves during cultural evolution are mental representations and that this evolution is shaped and channeled by cognitive biases and social products which act as factors of attraction. However, the case of urban poverty shows that the attractor should be considered in terms of a self-sustaining regulatory architecture connecting different factors rather than inherent cognitive biases or even social products. Inherent cognitive biases may be part of a network of interactions, but only the interactions between the different factors leads to attraction.
4. An evolutionary–ecological view that stresses the bidirectional interactions between humans and their ecological and social environments was suggested by Odling-Smee and his colleagues (Odling-Smee, Laland, and Feldman 2003; Kendal et al. 2011). Although this approach is close to ours in that it highlights the ways in which humans construct the sociocultural niche in which they live and in which they are somehow “selected,” the constructed niche and the individuals constructing it are regarded as different causal factors rather than as causally potent constituents that are facets of a single analytical unit, and the developmental socialization of individual and groups within the community is not explicitly modeled.
The DST approach, on the other hand, regards the integrated developmental social system (e.g., urban poverty) which is constituted by individuals and groups of individuals (as well as institutions, cultural webs of meanings, ecological factors, genetic and epigenetic factors, etc.) as the unit of analysis. The relationships between the dynamics of flexibility and stability are one of its explicit foci, and the developmental social trajectories of agents within the system are regarded as central to the dynamics of persistence and change, with each stage in the process of social construction scaffolding the next. In the case of urban poverty, following how different initial conditions and different developmental trajectories end up in the attractor state is central to the analysis, and the different sociological studies highlight different self-reinforcing and self-sustaining activities and schemas. Within the DST framework, analysis in terms of competition and selection among self-replicating entities is considered as only one possible type of process partaking in the dynamics of persistence.
Moreover, in the case of the cycle of urban poverty, as in many other cases, it is far from clear who benefits: which agent, behavior, idea, or group may be considered “adaptive” and from what point of view, if at all. Is it adaptive or nonadaptive to work in the formal economy when such work pays minimum wages and presents little chance for mobility? Is it adaptive or maladaptive to sell drugs? In what sense are malnutrition or alcohol consumption, with their transgenerational deleterious effects, adaptive? Rather than trying to speak of adaptive cultural behaviors—an attempt that is prone to tautological reasoning and oblique moralizing—it seems that descriptions in terms of differential stabilization are more adequate and more informative. With processes such as urban poverty, it is developmental stabilization, not adaptability, that is the key issue. Furthermore, a developmental view focused on self-maintaining interactions between different developmental inputs highlights the dual role of different resources, schemas, and agents (both their causal and the constitutive roles), without externalizing parts of the system. It is the dynamic and persistent product of their systemic interactions which is the unit of analysis. The factors contributing to this network belong to different domains (and may include social, ecological, and epigenetic elements), and together construct individual preferences as well as correcting, reinforcing, and excluding social mechanisms. This sociocultural network affects the developmental paths of individuals and subgroups within the collective, who affect the network they find themselves in, scaffolding it, receiving scaffolding from others, stabilizing it as they do so, but also potentially changing some of its contours.
Waddington’s Epigenetic View
The visual metaphor we choose in order to clarify our views is inspired by Waddington’s (1957, 1968) schema of developmental–embryonic dynamics, which assumes that any stable, recurring, end state—the entire mature phenotype or an aspect of the phenotype (i.e., a functioning heart)—is the effect of a dynamic network of developmental processes (see figure 14.1). In his visual depiction of this process, known as the epigenetic landscape, a large number of genetic factors coregulate each other in ways that produce a recurring, typical, mature phenotype. The regulatory structure of these interactions raises the chances that a particular end point will recur, so that even if one of the genetic factors changes or disappears altogether, and even when the landscape’s paths change, the same type of end state, the attractor, is reached.
We suggest that a dynamic landscape metaphor captures some important features of social reproduction. Such a landscape must, of course, contend with the fact that humans, unlike genes, are conscious agents who construct their own environment and react to it, even as they are structured by it (see, e.g., Giddens 1984). In other words, they actively construct their ecological, social, and epistemic niches and hence the very environments in which they and their descendants live, act, reproduce, and may be selected (Kendal et al. 2011). The social landscape, therefore, must be thought of both as the product of large-scale resources and mechanisms that construct the general environment of action and also as the result of specific actions and interactions that the individuals and groups within the landscape creatively construct and constitute.
In order to understand our use of Waddington’s position, some of its constituent assumptions and terms must be explained. The term epigenetics was suggested in the 1940s by Waddington, who later defined it as “… the branch of biology which studies the causal interactions between genes and their products which bring the phenotype into being” (Waddington 1968, 12). Epigenetic research traditionally focused on developmental canalization and phenotypic plasticity, two aspects of development that decouple phenotypic and genetic variations. The modern epigenetic view emphasizes the centrality of intracellular and intercellular networks of interactions during the organism’s development (and sometimes between generations of individuals as well) and expands Waddington’s concepts of “plasticity” and “canalization.”
“Plasticity” is defined as the ability of one and the same genotype to produce different types of behavior, physiology, or shape, in response to different conditions in its environment. The individual’s responses to the environmental circumstances may be active or passive, lasting or sporadic, reversible or irrevocable, adaptive or nonadaptive (West-Eberhard 2003). Furthermore, the range of plastic responses to new conditions may be narrow and predictable or wide and open-ended. An example of the former is the effect of the type of food provided to the larva of a bee, which determines whether it will develop as a queen or as a worker, while examples of unpredictable and open-ended plasticity are seen in animals that learn new behaviors through trial and error. The open-ended plasticity in humans who practice novel cognitive behaviors is particularly important for understanding culture.
Open-ended plasticity is usually based on processes of “exploration” and “selective stabilization.” Such processes, which may occur at cellular, physiological, behavioral, and social levels, are all based on a common principle: the generation of a set of local variations and an array of interactions, from which only a subset is eventually stabilized and ultimately endures. The initial conditions, the number of points around which development can be stably organized (i.e., the “attractors”), and the ease or difficulty of deflecting developmental trajectories from their present paths are the factors that determine which specific output is actually realized. “Attractor” is thus shorthand to describe a dynamic state which is the product of a large number of interacting factors that tend to coregulate and to reinforce each other in ways that regularly produce this recurring end result. There could be different permutations and plastic shifts in many of the underlying processes, creating a number of variations in trajectories, but the attractor state remains the same.
The exploratory behavior of unicellular organisms (such as bacteria and paramecia) is an example of selective stabilization. On blocking of its path, a paramecium will start moving at random and will start directional movement only when it encounters an attracting stimulus, such as light or food. Similar initially random exploratory behavior is observed in the growth trajectories of plants. Other examples are the selective stabilization of synaptic connections during development and learning, and, at the behavioral level, trial-and-error learning in animals. Exploration and selective stabilization processes therefore involve extensive plasticity that is manifest in the variability of inputs and the multiplicity of possible trajectories that, nevertheless, are canalized, leading to a given attractor. Canalization can be defined as the adjustment of developmental pathways so as to bring about a uniform result in spite of genetic and environmental variations (Jablonka and Lamb 1995, 290, based on Waddington 1957). Canalization thus leads to robustness, reliability, and stability in a “noisy,” ceaselessly varying world. A single typical phenotype is generated by many different underlying genotypes and under various environmental conditions.
Canalization and plasticity are related, so that for virtually every instance of canalized development one may detect plasticity at other levels of organization, underlying or overlying (Jablonka 2006). For example, animals pant, sweat, and seek shade and shelter from the sun in hot weather, a mode of behavior that keeps their internal temperature within narrow bounds. Through physiological and behavioral plastic changes, a stable (canalized) internal temperature is secured. Trying to capture this dynamic stability of a given morphological or behavioral pattern during embryonic development, Waddington described the process in his famous “epigenetic landscape” metaphorical model of development (Waddington 1957; figure 14.1a, b).
Waddington’s epigenetic landscape describes ontogenetic development as a ball rolling down a tilted landscape, with many hills and branching valleys descending from a high plateau, which represents the initial state of the fertilized egg (see figure 14.1a). The attractor toward which the ball rolls is a morphological, physiological, or behavioral state which is relatively stable: the whole mature phenotype or part of it such as a functioning heart, a functioning eye, or the locomotor ability in a human child (Thelen 1995). The downward slope represents temporal progression, with the bottom representing the steady state (the mature phenotype) and the top representing the initial state. The steepness of a slope leading to a valley may be thought of as the strength of biases and propensities, constraints, and specific affordances, and it indicates how much the system can vary—the steeper the valley, the more constrained and buffered the developmental path and the more difficult it is to deflect it.
The landscape is underlain by a network of interacting elements (see figure 14.1b). It is shaped by the tension of guy ropes that are tied to pegs in the ground. In Waddington’s schema, the pegs represent genes, and the guy ropes represent their biochemical and regulatory effects. The topological relations of the interacting pegs and guy ropes determine the architecture of the visible undulating landscape. These interactions could be described today by network-theory models to which a time dimension has been added. The different pegs interact in complex ways, and some pegs are more “central” than others because they have connections with many other pegs. It is not only or mainly the number of different “pegs” (different interacting units) that is important, but the way they interact (Brandes and Erlebach 2005). Within a certain range, the attractor is reached whatever the route and trajectory taken and whatever the precise initial conditions. Although in certain circumstances the existing biases and propensities may be overcome and a different attractor may be reached, this requires that a new challenge (e.g., strong stress, Lyon, this volume) alters the multiple interactions among the elements and generates a new network of interactions that divert the ball from its normal course. The steepness of the valleys, and hence the outcome of development, can be altered through natural selection; it can evolve.
Translating the Metaphor
We take Waddington’s visual landscape model as a point of departure for describing and making sense of the dynamics of stability of social outcomes. Our social-developmental landscape is a description of a pattern of social life in a theoretically circumscribed population, which is the outcome of the dynamics of multiple social resources, schemas, and activities that affect the development of individuals and groups of individuals in the community; the trajectories in the social landscape are the developmental paths of individual subjects as they go about their lives.
Although inspired by Waddington and visually similar to it, the social landscape departs from that envisaged by Waddington in many ways. First, it is far more open-ended than the original Waddingtonian landscape. Second, unlike Waddington’s landscape, which describes the dynamics of a single individual development, the social landscape describes a group-level process with multiple trajectories individuals typically follow/construct, with individuals developing along a given trajectory following a time axis until they reach the attractor. Third, and crucially, the pegs and guy ropes in our schema are both causal and constitutive: they are not separate from the landscape itself, they make up the “hills” and “valleys” and are not external causal factors. The social landscapes in figure 14.2 are therefore sculpted up from the network (the interacting factors are not underlying it or overlying it), like a fisherman’s net which is crunched up and modeled into a 3-D landscape shape. Fourth, the resources/causal factors that interact are not all made up of the same “fabric”: they belong to different domains, and, depending on the case analyzed, may include genetic, epigenetic, ecological, institutional, and symbolic resources. For an individual, these parts of the web fold up to construct its family structure, school, curricula, professional society, religious organization, work organization, interest groups, and the transient and enduring objects and products of activity that make up the physical landscape of individual and social life. In addition to these, there are those physical and symbolic aspects of social life that are more indirectly connected to the individual and that include laws and other normative demands, ways of communicating, public resources, and so on.
Like the ball in Waddington’s model, individuals roll down the landscape. People, however, are not passively driven down the slopes (as the ball seems to be in Waddington’s original picture): people affect their own trajectory to some extent, as they construct their social cultural and epistemic niche—their passage is an active process which always changes it somewhat. Although individuals go through partially preexisting trajectories which they typically tend to stabilize and “deepen,” their activities may also alter the local features of the landscape with which they interact. To use sociologist Anthony Giddens’s terms, there is always a “double hermeneutic” in the case of people: people are both structured by, and actively structure, social life (Giddens 1984).
A subject’s steady state is a point (or a path) in the landscape where that individual “settles.” This region is identified by the researcher and is thus a theoretical construct defined as a relatively stable state in the population members’ trajectories, which we want to account for. The individual’s trajectory is produced by the sociocultural landscape he or she is born into, or otherwise finds himself or herself in, and the individual follows a route already “paved” and “dug” by others who have been following it. However, as we pointed out earlier, individuals are also constantly remaking their trajectories through their actions and interactions with an early stage in their social development that serves as a scaffold for the next stage. The landscape is thus in constant flux, with the trajectories of individuals’ lives being continuously molded, altering some of the contours of the neighboring landscape but only slightly modifying the attractor state. Finally, as shown in figure 14.2, a given social landscape is part of a larger social landscape—there is always a larger context, and there are different social worlds that are related to the landscape we are interested in. Interactions within a specific landscape, however, are not only more numerous than interactions between landscapes, they are also mutually more reinforcing, with more auto- and heterocatalytic effects and with more coscaffolding effects. Some parts of the focal landscape are shared with a neighboring landscape, and hence entering a particular trajectory may lead the individual to end up in a different attractor. A “social earthquake” in one part of the landscape can ripple to others.
The case of the reproduction of urban poverty, with which the previous section began, can be described in terms of this social landscape model: the various social activities, resources, and schemas canalize the social state in spite of the variability of initial conditions and the developmental plasticity manifested in the various developmental trajectories of different individuals. The “social landscape” is not only a good metaphor for interpreting existing studies but can be used as a point of departure when conducting research. The following research note, drawn from the ethnographic work of one of us, on Jewish Orthodox neighborhood life in Los Angeles (Tavory 2010b), makes this point.
The Case of Orthodox Jewish Life in L.A.: Persistence and Thriving in the Midst of Profanity
The empirical puzzle that Tavory worked through was the following: the largest and most successful Jewish Orthodox community on the West Coast of the United States today—both residentially and institutionally—is situated in Los Angeles’s Melrose–La Brea neighborhood, an area known for its “transgressive” youth culture rather than for strict adherence to religious edicts. This site thus seemed a strategic place in which to ask how urban spaces sustain ways of life that the modern city was supposed to erode and extinguish. Relying on almost five years of ethnographic participant observation, interviews, and archival data, Tavory examined the history of the neighborhood, Orthodox neighborhood institutions, and the everyday lives of members. For the purposes of this chapter, we will focus on three mechanisms that contribute to the relative stability of the landscape of religiosity—the educational system, edicts regarding proscribed and prescribed activities on the Sabbath, and the way in which Orthodox Jews are identified and reconstituted as Jews in everyday life. These mechanisms are far from being the only processes through which religiosity is sustained (see Tavory 2010b for further elaboration). Rather, they exemplify different kinds of mechanisms—institutional, cultural, and interactional—operating through different time scales and different actors.
First, educational institutions, beginning in the early school years and continuing even after members are married, sustain religiosity both intergenerationally and on the level of people’s everyday life. Orthodox schools are private and cost between $11,000 and $25,000 a year. Although most members in the neighborhood earn a comfortable middle-class income, they also have a large number of children, with five children being the norm rather than an outlier. In such a context, school fees become one of the most pressing issues in the lives of members. In fact, some members reported that they had paid more than $500,000 for children’s schooling throughout the years. In a neighborhood where the average income per household is about $71,000 per year, households with three school-age children would need to pay about $33,000 for schools, more than half of their yearly income after taxes.
In fact, however, relatively few people actually pay full school fees. Families with five to six school-aged children could not be expected to pay full tuition for all of them; children of teachers usually secured large discounts; children from less affluent families almost universally received financial aid. Indeed, in conversations with the principals of two of the largest Orthodox Jewish schools in Los Angeles, both administrators noted that they were receiving less than full tuition for about two thirds of the students. However, parents who did receive discounts were often expected to contribute to the school in other ways. Some parents organized “pizza lunches,” volunteered to drive children to school, or actually tried to fund raise for the school if they had the skills to do so. Along these lines, parents reported that they were drawn into greater involvement in the school’s affairs, leading them to interact more with other parents with whom they collaborated.
Thus, on the one hand, the educational system cemented families’ involvement in Orthodox life by making demands on families’ resources and time. From the children’s perspective, on the other hand, the educational system constructed and reproduced Orthodoxy in multiple ways. The most obvious has to do with children’s friendships and personal ties. All students being Orthodox, the schools ensured that children did not have friends who were nonreligious, or—much “worse”—non-Jewish. This, in the view of most Orthodox, would translate into a smaller chance of “marrying out.” Posted in Hebrew on the walls of one neighborhood synagogue, a poster advertisement for an Orthodox school in the neighborhood made the message clear: “Mother, I think congratulations are due,” and then, immediately below, “The heart of a mother cries: the bride is not Jewish.” The subtext was clear: only by sending the child to the “right” school can such a scene be avoided.
This is an obvious boundary marker, linking discursive practices to more “interactionally mute” (Tavory 2010a) practices of boundary formation and involvement, so that even without having to actively construct boundaries between themselves and the non-Orthodox world in their everyday lives, the boundaries are strictly kept in place. However, schools coconstitute not only levels of institutional involvement but religious involvement more generally. Like most other aspects of the community, schools are differentiated along an axis of religiosity. The more strictly Orthodox schools have entry requirements for the children’s parents. To be admitted into the most prestigious schools, parents sign a document stating that they do not have a television and that married mothers have their hair covered. Other aspects of family life are also monitored, though less strictly. Thus, for example, one family owned a dog, something usually frowned upon in Orthodox Jewish circles. The school two of their boys attended notified the family a few times that if they did not get rid of the dog, the school would expel their children. They did not ultimately expel them, probably because the family was one of the few that paid full tuition, but the fact that such a threat was even possible is telling. Thus, schools monitored both the friendship networks of children and the behaviors of their parents—both important if Orthodox life is to be reproduced.
Also crucial for the intergenerational reproduction of Orthodoxy, specific patterns of mobility were created and fostered, especially at the stricter Orthodox schools. The most prestigious schools thus offered a large number of religious classes and very little in the way of secular subjects such as English, mathematics, history, and science (see also Heilman 2006). In post–primary school education, the amount of nonreligious studies was further curtailed. Although schools did the bare minimum required in order to receive some public support, children—and especially boys—finished their schooling without “advanced placement” classes that would ease their way to college. As a result, a secular educational trajectory was made much more difficult to pursue. The few teenagers in the study who came out of these schools but decided that they wanted to go to college instead of advancing their careers in the religious field had to go through a community college first.
Thus, the educational system helps in the reproduction of the Orthodoxy in the neighborhood in a few different ways: by capturing the time or resources of families, by providing work for others, by controlling some of the practices members subscribe to even beyond the domain of the school, and by placing Orthodox children on a specific educational trajectory, creating a situation in which joining the secular society (i.e., potentially moving toward a different attractor) becomes harder.
Second, religious edicts themselves solidify participation in different ways. To take the simplest of the ways religious edicts and the reproduction of Orthodoxy are tied, the edicts proscribing different work activities on the Sabbath constrain members’ choice of residential space, while strengthening relationships between members. As driving a car on the Sabbath is considered to be a form of proscribed work, members have to live within walking distance from an Orthodox synagogue. Doing so, in turn, means that they have to live in proximity to others who go to the same synagogue. In an era in which residential neighborhoods are often decoupled from the spaces of sociability people attend, religious edicts sustain a layering of neighborhood space and spaces of sociability. In such a situation, it thus becomes much easier to sustain friendships with cocongregationists, to have children go to each other’s homes after prayers, to organize impromptu study sessions, and so forth.
These edicts also had some other unexpected consequences. The Sabbath begins at sunset on Friday, ending an hour after sunset on Saturday. This meant that, in the winter months, the Sabbath often started relatively early, sometimes as early as 4 p.m. Although this was not a problem for most people, some Orthodox Jews who worked in Los Angeles’ Downtown area, relatively close to the Melrose–La Brea neighborhood, lived farther away, in the San Fernando Valley. These individuals were often pressed to arrive to their homes before the Sabbath started. They were, in effect, caught between the secular time in which they worked and the religious reckoning of time that could not allow them to reach their destination. To solve this problem, some of these people were often invited to spend the Sabbath, along with their families, at the houses of acquaintances in the Melrose–La Brea neighborhood, further cementing friendships and tying Orthodox members of different neighborhoods in the Los Angeles metropolitan area together.
The last mechanism operates on a different level of analysis, focusing on mundane interactions Orthodox people in the neighborhood have as they are recognized as “religious Jews” by outsiders, owing to their distinctive attire (see also Tavory 2010a). Being an Orthodox Jew, members are either born into, or later enter, a world of signs: the black Hasidic garments, the beard, the yarmulke. Like other uniforms that signify a particular membership categorization—such as the priest’s collar, the police officer’s uniform, or the Amish man’s clothes—the strict Orthodox attire provides clear signs of belonging. For members there are many gradations of dress and attire, each signifying varying degrees of observance and membership in a specific Orthodox subaffiliation—a round hat means something different than a flat one, and a shaggy beard is different from a well-trimmed one.
However, even simply wearing the yarmulke, the most minimal sign of Orthodox Jewish men’s belonging, allows members to be recognized and categorized by both members and nonmembers as Jewish and somewhat observant. Though it has specific meanings known to the wearers, it is usually worn without thinking of its symbolic meanings. In fact, Hasidic Orthodox men sleep with their yarmulkes on, taking them off only when they go into the shower. When asked if they thought about their dress and yarmulke or attended to their signification on an everyday basis, most members said that, although they were acutely aware of them at certain events and on certain occasions, they were usually worn nonreflexively, as a matter of habit.
Yet a specific interactional effect emerges when members wear a yarmulke. Without initiating interaction, the yarmulke provides a space for interaction in its automatic denotation of Jewishness for a general audience. People at bus stops approach Orthodox Jews every now and then with questions such as “Is it true that Jews believe that…?,” and—not often, but not as rarely as might be hoped—someone will shout an anti-Semitic slur at them as they are walking down the street.
Thus, though in most cases members themselves pay no mind to their clothing in everyday life, wearing the yarmulke—and of course the full Hasidic garb—is a constant resource for interaction initiated on the part of others, an interactional hook marking the wearer. With the yarmulke, members become Jews actually “doing being Jewish” while engaging in other activities in which Jewishness is far from the relevant facet of self. In terms of the stabilization of Orthodox religious life, these mundane interactions are important in two ways. First, categories of membership are often constructed not only by the people we study but often by others. Second, over time, people learn to expect their environment to treat them in a specific way, further stabilizing the way they perceive themselves.
These mechanisms were chosen because they specify the different orders of social mechanisms that are at work in constructing and reproducing a specific way of life. From institutional structures and cultural edicts to the microinteractions of everyday life, religious life is powerfully canalized—although people can leave Orthodoxy behind them, it is quite rare. Indeed, as the examples above suggest, a network of complex interactions among different processes reconstitutes the community and the social and religious identity of its individuals. Like urban poverty, religious life is powerfully canalized, held and reproduced by multiple social mechanisms operating at different levels.
In terms of the Waddingtonian landscape, the mechanisms described above can be seen to entrench and stabilize the transmission of Orthodox Judaism not only by pulling people into a dense shared social world but also by curtailing the possibility of people’s moving to different attractors. In these terms, the three mechanisms described have somewhat similar effects. Participation in Orthodox schools ensures that childhood friendship circles are limited to Jewish Orthodoxy and that people sustain Orthodoxy (at least minimally) in the home. The edicts of the Sabbath sustain a different temporality upon which members live their lives: as a shared temporality is one of the sine qua nons of sociality, it pushes members away from others who are located on different temporalities (say, the secular calendar, in which the weekend is precisely the time to meet friends, go on trips, etc.). By the same token, the visibility of Jews as Orthodox Jews practically excludes the non-Orthodox passerby: by interacting with others as Orthodox Jews, they effectively curtail the possibility of Orthodox Jews taking on other identities in interaction.
These different mechanisms are also interrelated. It is easier to make sense of having only Jewish friends when non-Jews look so different and act within a different weekly temporal structure; it is easier for children to make sense of the different temporal rhythms of life when all their friends share the same temporality. In other words, the different mechanisms do not only coproduce a relatively stable Waddingtonian sociocultural landscape but also scaffold each other, so that each becomes more plausible and easier to make sense of and to practice.
General Discussion
A DST approach to cultural evolution suggests that the most useful unit of analysis is the dynamically reproducing sociocultural system under study, rather than any particular causal factor that contributes to the reproduction dynamics of such a state. We suggested that the dynamic regulatory architecture of the network that builds up the social dynamics should be under scrutiny, and that the focus of the analysis would be those interactions that lead to the reproduction—albeit always with some modifications—of the social attractor state. Understanding the dynamics of social reproduction is, we suggested, also a good point of departure for understanding how departures from a particular reproduction dynamics occur.
As we noted in the introduction, the idea that cultural evolution should be understood in terms of attractors has been proposed and developed by Dan Sperber. Sperber argues that attraction can lead to cultural stability even in the face of variations and a low fidelity of copying (Sperber 1996, 2005, 2011; Claidière and Sperber 2007) and suggests that cognitive biases and cultural products are the factors of attraction that bias and channel the evolution of cultural representations. There are similarities between Sperber’s approach and ours, in that like him we believe that viewing cultural change in terms of replicators (memes) or differentially reproducing model individuals’ behaviors is insufficient and often misleading (Jablonka and Lamb 2005; Tavory et al. 2012). However, our view of attractors and units of sociocultural evolution differs from his: while Sperber concentrates on the evolution of cultural representations with the factors of attraction being cognitive biases and cultural products, we see the social attractor state as the unit of evolution and regard the mental representation and the social products as some of the schemas and resources that interact through human interactions to form the self-sustaining dynamics of the attractor state. We therefore focus on the producers of social and cultural products, and on the regulatory structures of social dynamics that are constructed by producers’ activities, especially when they reach their most stable and socially constructive ontogenetic stage. Thus, we focus on developmental histories and the regulatory nature of the (essentially stabilizing) social interactions that occur at each generation.
On a broader theoretical scale, an approach based on processes of exploration and selective stabilization has a much more general application than a view based either on the selection of replicating entities or on attraction alone. Natural selection of genes, memes, organisms, or traits can be seen as a special case of exploration and selective stabilization (a special case that involves the differential multiplication of entities). Selective stabilization accommodates both “replicator” and “attractor” dynamics and therefore seems particularly applicable for the study of cultural evolution, which, as we see it, includes both. Attraction can sometimes overwhelm selection, and therefore, as the case of urban poverty shows, stability and recurrence in the face of perturbation may not be the consequences of adaptive superiority. Since in complex social systems adaptedness has a very vague meaning, an analysis in terms of selective stabilization is far more appropriate.
Throughout this chapter we suggested that using a “landscape” metaphor, based on Waddington’s epigenetic visual metaphor, may help in capturing the dynamics of stability and change in complex sociocultural environments. Like Waddington, we assume that there is very rarely, if ever, a situation in which a recurrent state of affairs is underlain by a causal process with a single dominant factor. Our Waddington-inspired schema assumes that any social attractor state is the effect of interactions and feedbacks between many different factors that lead to the persistence of social life. The regulatory structure of these interactions raises the chances of a particular outcome: if the social landscape changes, if one of the causal–constitutive factors is modified or disappears altogether so that some of the ridges and valleys are altered, and even if many actors leave the community, the attractor is nevertheless likely to be reached.
As we pointed out, the constituents of the social landscape belong to different domains, and what they are in a given case depends on the social pattern we study and the question we ask. In the case of both urban poverty and religious life in Los Angeles we included among our resources and schemas different social and meaning-making processes that constitute a complex social landscape and a complex attractor which is a part of it. And in both cases the landscape is relatively stable in spite of great variability in individual developmental trajectories and in individual starting conditions.
Sociologists make a similar point by using different vocabularies, talking, for example, about the “overdetermination of causes.” There are, however, two important correctives that our approach offers sociologists. First, thinking through this metaphor would force researchers to explicitly study the complementary dynamics of change and stability. As many sociologists today stress, any social outcome is an end point in a long series of both powerfully patterned and highly contingent interactions. Structure, then, is not something that exists out there, independent of these contingent interactions. Indeed, if structure “exists,” it exists as a process, a “verb” (Bourdieu 1977; Giddens 1984). However, students of culture are often seeking the “golden practice”—the piece of the puzzle which would explain the difference between conditions. Raised on Mill’s (2002/1843) “method of difference,” they (like other scientists) often look for the one practice that would account for the difference between conditions. Thinking through Waddington’s metaphor would make this project highly dubious. If indeed sets of practices are canalized toward a specific outcome, then the comparative study of meaning should not focus on specific practices but rather on the system of canalized relationships and the combinations that may subvert a well-canalized attractor. Although one well-placed change can cause a different course to be more common, this would be rare. In most cases, it is only a combination of changing relationships and mechanisms that produces change. The task shifts from a specification of a single cause to thinking about networks of relationships which produce a similar effect. Even when sociologists do acknowledge that many factors affect a given social outcome, they tend to focus on “hubs,” junctions where many factors come together, rather than, as we would prefer, on the architecture of “hubs,” that is, on those dynamic interactions that lead to autocatalytic, self-sustaining outcomes. Identifying those parts in the landscape where different factors create autocatalytic, self-sustaining interactions leading to the reproduction of the social structure might lead to better targeted attempts to understand, and perhaps even intervene, in the landscape.
Second, thinking about social phenomena in terms of plasticity and canalization would be useful in providing a spectrum of different conditions—a scale that goes from the well-canalized process (such as the reproduction of poverty or the reconstruction of Jewish orthodox life in L.A.) to more precariously canalized phenomena that are thus reproduced with less fidelity or that can readily change. Thinking along this continuum would allow sociologists to create a scale of effect canalization, and thus a comparative tool. In this regard, the metaphor’s pragmatic value is in allowing sociologists to ask different questions than they have been asking and opening up a comparative dimension that is important for understanding of both stability and change.
In summary, the landscape metaphor provides a general approach to the study of cultural development/evolution which could help both cultural evolutionists and sociologists tackle the kinds of problems they routinely engage with. Due to its complexity, it is not a quantitative or formal model easily amenable to quantification and simulations. What it provides, however, is a framework for articulating and testing specific hypotheses about the dynamics of social reproduction in different cases. Such a framework, in turn, would not only help researchers to account for the relative stability of sociocultural systems. As we focus on the developmental trajectories of actors that forge those interactions that maintain the general dynamic structure of the system, we can also account for its changes. It is only on the basis of reproduction dynamics that we can identify the feedbacks that drive the system beyond its attractor, or those regulatory interactions that diminish or even eliminate certain outcomes.
Acknowledgments
We are very grateful to the members of the “evolution of human cognition” group in All Souls College, Oxford, in 2011 for their critical and thoughtful discussions of the ideas presented in this paper and to the editors of this volume for their incisive, detailed, and constructive comments. We are especially indebted to Anna Zeligowski, who drew the figures, and to James Griesemer for helping us with figure 14.2.
References
