Throughout chapter 5 we discussed the different roles played by friendship and trust in an organizing campaign. Many of the conclusions we reached, but not all, are familiar to organizers themselves, who have often had an inchoate sense that it mattered less that one was a friend of a potential member than if that potential member was able to trust one to have his or her interests at heart. While friendship often invokes such trust, it carries with it a range of other expectations that are difficult for an organizer to meet—perhaps the most obvious of which are demands for time spent, dining, hanging out, and so on. Organizers don’t have this kind of time available, as a general rule. And yet, at the same time, they are trying to bring about and induce a communal experience defined by a shared focus and orientation to the project of social change. To achieve that shared experience, people have to come to see themselves in a new light, as active members of a self-conscious community of similarly situated people with similar aims. They have to see themselves as sharing a project. New selves emerge out of new interactions with others, out of new social structures.
Generations of sociologists have argued that there must be some set of processes by which social structure finds its way into our ways of seeing and acting. Social structures are not just things outside us that we must navigate; they become us, shaping us and shaping the ways we see and relate to others. The internalization of structures is expressed in lots of different ways. By Pierre Bourdieu it is described as “habitus,” which gives rise to the kinds of embodiments that characterize his description of peasant men as men who cannot dance in The Bachelors’ Ball; by Richard Sennett and Jonathan Cobb, in The Hidden Injuries of Class, it is the set of beliefs and feelings that underwrite a class society by leading to actions that voluntarily reproduce it—like the self-blame and resignation that accompany one’s inability to advance in a nominally meritocratic job market. Michael Burawoy, in Manufacturing Consent, and Nancy Chodorow, in The Reproduction of Mothering, observe how microinteractions—on the shop floor and in the home—reproduce class and gender relations, respectively, by way of the embodied understandings that such interactions produce.1 In all of these cases, larger social structures—divisions between rural and urban, rich and poor, men and women—are seen to be reinforced and reproduced through embodied dispositions, beliefs, and feelings that emerge from our membership in such structures.
The premise that structures get into us is one that many sociologists take for granted. It is clearly right. But at the same time it is mysterious. Perhaps the particular causal mechanism doesn’t matter; the outcome—the reproduction of social structure—is what we care about. And yet reproduction is not the only outcome. Sometimes people have understandings of themselves and others that generate social change. Organizing is about inducing that kind of realignment of self and other. And this means that the mechanisms by which structure gets into us might also contain the foundations for how change takes place. Therefore, it may matter how social structure gets inside us.
There are some hints to the puzzle of how it happens in the microstructures that arise—seemingly naturally—when people get together. For example, we know that humans create and navigate social structures, and we know that we do so in somewhat predictable ways. Status hierarchies emerge among people across all human groups, even if people start off with the best of intentions (like commune adherents, or social activist–oriented millennials).2 We also know that there are basic relational building blocks out of which social structures tend to emerge: like the fact that we tend to like people who like us (dyadic affective reciprocity), or the fact that we tend to dislike the enemies of our friends and like the friends of our friends (balance).3 Our idea is that these processes are the foundations for both social change and social reproduction.
Despite much theorizing about the instantiation of social structure in our perceptions, we know very little about how it happens. Perception is something that happens in our bodies, and mostly in our brains, which makes it somewhat foreign for sociologists. Nevertheless, if we care about how social structures reproduce themselves through individuals’ perceptions and the conditions under which such structures might be unsettled, then it makes sense to try to see what is happening in our brains when we see and respond to social structure. That is what we tried to do when we embarked on using fMRI scans to understand the formation of social structure among the OUR Walmart volunteer students. In this appendix we introduce the results of those studies; both because they motivate the conclusions in chapter 5 (as already indicated) and because they are intrinsically interesting.
SOCIALIZING SOCIAL COGNITIVE NEUROSCIENCE
Back in late May 2014, while the students were still getting to know one another, most had agreed to climb into an fMRI scanner so we could look at their brains while they looked at one another and looked at themselves.4 In early August, they went back into the scanner so we could see if and how their reactions to themselves and one another had changed. This time, physically and emotionally exhausted from the field, many of them fell asleep in the machine—this in spite of the deafening racket and constant stimuli.
How the brain works is mysterious but not a complete mystery. We know from people who have suffered traumatic brain injuries—where some part of their brain is removed or damaged—that different areas of the brain are responsible for our different capacities. For example, from one famous patient, H.M, we know that if we remove the hippocampus, the amygdala, and the entorhinal cortex, it is completely impossible for a person to store memories.5 We know that people who lose their left superior temporal gyrus can’t understand any language. And that those who lose their left inferior temporal gyrus can’t say anything, even if they understand what is being said to them. So, we know what parts of the brain are required for memories, language reception, and language expression.6
Neuroscientists have also begun to think about the neurological underpinnings of social life, although here there is much more work to be done. We can now identify, at least generally speaking, the parts of the brain that are associated with self-recognition and self-reflection, as well as those parts of the brain that link us to others. We are beginning to understand, at the neurological level, the ways in which we rely on conceptions of self to understand those close to us, and—reciprocally—how our understandings of ourselves are influenced by what we think others think about us. And we can see some evidence that there are the neural pathways by which inequalities in the social world get internalized—for example, the amygdala seems to be a neural correlate of implicit negative bias; and the dorsal anterior cingulate cortex is activated during the experience of social exclusion or loss.7
Current social cognitive neuroscience is, however, starkly limited with respect to understanding the ways that the complex social worlds in which we live are shaped by and shape our neural responses to others, mainly because its standard research paradigm is completely asocial. The usual fMRI study oriented toward understanding how people respond to others puts the research subjects in a scanner and then subjects them to a “treatment,” typically a highly stylized stimulus on a single dimension that is interesting to the researcher—for example, gender, or race, or age.8 So, for example, one might wonder: Do our brains respond differently to older people than younger people? Neuroscientists “test” this idea by showing pictures of an old person and then a young person to someone while he or she is in the scanner. It turns out that our brains do respond differently to these kinds of abstract stimuli, at least sometimes. And yet, these kinds of abstract stimuli may not have anything to do with actual social processes. The motivation behind these kinds of research designs is the idea that by putting research subjects into purified environments one can more precisely estimate the effect of the treatment. If people lived in vacuum tubes, this kind of research project would make sense. But we don’t.
We are interested in the same problem that the social cognitive neuroscientists are interested in—Can we identify the neural signatures of relationships? Are there neural processes that lead inexorably to inequality? Are there neural bases to social solidarity, to mutual attraction or distaste for other? But rather than abstract away from people’s lived experiences of specific social relationships to answer these questions, we wanted to make sense of this specificity. Could the fMRI data that we collected at the beginning of the program help us understand how a group of actual people with actual relationships evolves over time? Could it help us understand how people see themselves, and those with whom they are interacting, and how this changes as relationships deepen? Would it help us understand resilience in the face of opposition and hardship?
We believed that we could learn the answers, or at least preliminary answers, to these questions, although to do so we would have to model face-to-face social networks as they evolved over time in the minds of our students. That required building a platform to see how our students saw themselves, and everyone else in their group.9 To see the activity in our students’ brains when they saw images of themselves or other group members, we needed to get the students into scanners. That turned out to be a good idea, independent of the science. It built community, and it indicated to the students just how unusual the project they were engaged in was. After all, no other internships are designed to capture the experiences of the interns as they unfold. The fMRI scanning helped convince everyone that whatever was going to happen would be interesting.
The students were both the target stimuli presented during the scan and the perceivers that viewed other stimuli. In other words, each participant in the scanner looked at repeated images of the other participants in the project, images of themselves, images of us, and images of a morphed or “ghost face” (as a control). While inside the scanner, they performed a simple cover task to maintain alertness (which worked during the first wave better than the second wave, when several fell asleep)—they were instructed to press a button with their pointer fingers each time a group member’s face was presented and a different button with their ring fingers each time a “ghost face” was presented. We complemented this neurological data with more standard social network techniques. We asked students—at the beginning of the project, in the middle of the summer, and a few months after the project ended—how they felt about one another, and how they thought others felt about them, on a number of different dimensions. This meant that we were able to see how patterns of neural activation were related to patterns of reported affiliation.
INSIDE OUR BRAINS
Now we need to take a little detour into neuroscience to understand what gets measured in the brain and why we bother to measure it. What we see in an MRI study is the trace of brain activity. When the neurons in our brain are activated, when they are firing, blood flow increases to the regions where they are located. The MRI measures this increase in blood flow and by association allows us to infer that nearby neurons are active. In an fMRI, everything about the mechanics of an MRI stays the same. The “f” stands for functional, and the idea behind the fMRI is that by giving the brain things to respond to—stimuli of different kinds—we can detect where the neurons dealing with that stimuli are activated. We can look at activation, captured by blood flow, at the voxel level. A voxel sounds like it describes an animal like a hamster or ferret, but it is actually a place in a three-dimensional grid, like a pixel on a bitmap, or a location in a house. Say you are reading this while sitting on your couch, in your living room, in your house or apartment. If your apartment is your brain, we capture you on the couch (well, actually, just elevated a bit off the floor) in the region of your apartment that is the living room. There are over 100 billion (100,000,000,000) neurons in our brains on average. We have resolution at 160,000 voxels, so from one perspective, our data are not very granular: each of our locations “holds” around 600,000 neurons.
Because voxels are in space, they naturally aggregate up to regions with names, and there is a lot of neuroscientific data about what kinds of mental processes these regions represent. When it comes to the question of reading social structure, previous research points us to two different regions of greatest interest. The first is the ventromedial prefrontal cortex, the amygdala, and the ventral striatum, henceforth discussed collectively as the “valuation system.” These are parts of the brain that are consistently implicated in processing the affective value and motivational significance of various stimuli, including other people. The second comprises the dorsomedial prefrontal cortex, temporoparietal junction, and precuneus. These interconnected regions are consistently activated in neuroimaging studies involving judgments about others’ psychological characteristics, mental states, and intentions or the passive viewing of social stimuli—such as familiar faces—for which we might spontaneously make such attributions. There are lots of good reasons to suspect that the set of activities involved in theory of mind are taking place in and at the intersection of these regions.
For all of our work, we study these regions specifically, and for the work that we will discuss here, we focus most on the valuation region. A lot of others have worked on this region as well, and we know from their work that this is an area that is activated (lights up, colloquially) when people are presented with opportunities to earn money, and when, for example, primates have a chance to earn juice or bananas.10 We think prizes, money, juice, and bananas are important, but our guess is that this part of the human brain evolved to solve important real-world problems that confront members of social groups: Who likes me? Who is important in this group? Who do other people/primates treat as important? Who should I look up to? And if solving these kinds of social navigation problems are important, we would solve them with the same system that tells us that juice is good—that is, our neural response would help us orient attention and focus on people whose actions are potentially rewarding so that we could then call up to consciousness what they might be up to, what they might be thinking about us, and so on.
AT FIRST, POPULARITY MATTERED, BUT ONLY AT FIRST
We have done this kind of experiment with a variety of groups, and in other settings we have been able to trace the neural signature of social status. Simply put, when one looks at someone who is of high status in a group, the valuation system is activated, even after netting out how much one likes that high-status person. In these other contexts, we have shown that this valuation activity seems also to lead to increased activity in regions of the brain related to social cognition.11 People see high-status others as having motivational significance for them (valuation), and so they work harder to try to understand those high-status others’ emotional states (social cognition). At the beginning of the summer we saw a similar pattern among the OUR Walmart students—those students who were more popular among their peers, whether they liked them or not, elicited more reward region activation than did the less popular.
Popularity mattered for neural activation before the project really got underway, in those first few days when students were meeting one another, making speeches, describing their aspirations, and introducing their motivations for joining the project. What is fascinating is that by the end of the summer, popularity was no longer related to activation of the reward system. Something happened over the summer, and whatever that was, it changed the social structure. The neural mechanism undergirding inequality in the group—what seemed to us to be a basic human characteristic—gave way to a much more nuanced response to group members, and in its wake a more supple form of social solidarity.
RECIPROCITY IS A SUPPLE FORM OF SOCIAL SOLIDARITY
Relationships among the Summer for Respect students evolved over time. One thing we discovered, based on the social network surveys alone, was that sociometric status—or popularity—explained a declining percentage of people’s individual liking over time. In other words, the amount of variation in people’s liking that was explained by targets’ overall popularity declined sharply over the course of the summer. Relationships became more idiosyncratic, more individuated, determined less by group status and more by the specifics of people’s relationships with one another.
We also discovered something else. The valuation activity in a perceiver’s brain, what we might call implicit liking, when he or she looked at a target at the beginning of the summer was statistically unrelated to how much that perceiver reported liking—what we call explicit liking—that target at the time. But implicit liking at the beginning of the project was significantly associated with how much the perceiver explicitly reported liking the target months after the end of the project, even when controlling for how much the perceiver explicitly liked the target earlier. In other words, we discovered that we can predict explicit liking in the future based on implicit neural activation in the valuation region months earlier. One’s brain knows who one is going to like and who is going to like oneself in the future.
The simplest explanation for affective reciprocity at the end of the program is that people who like each other at the start of the program like each other at the end, and vice versa. But that is not what is going on. All of our findings control for the effect of initial liking on later liking. Equally interesting, the students really did change their opinions about their peers the more they got to know them. In some cases, they liked each other a lot more, and in some cases, they really came to dislike each other. Continuity in explicit liking is not the mechanism for the increase in reciprocity that we observe.
What does the emergence of dyadic affective reciprocity have to do with the broader issues with which we are concerned in this book? What we observed among the students over the course of the summer is a movement from a version of mechanical solidarity, a solidarity based on common beliefs, to one of organic solidarity, a solidarity based on complementary differences. At the beginning of the summer, our students’ perceptions of one another were driven largely by others’ perceptions—a person’s valuation region was activated when he or she looked at those who were popular among others, regardless of how much that student liked those popular people. By the end of the summer, the relationships among the students had deepened and differentiated; people’s orientations toward one another were driven less by group-level status and more by the complementarities of their experiences with one another.
Reciprocity is one of the most basic building blocks of human cooperation.12 Dyadic reciprocity generates group-level cooperation and cohesion in experimental and nonexperimental settings, in some cases more so than explicit group identity.13 And this tendency toward reciprocity, or symmetry, in affective exchange is a mechanism that often works against the emergence (or reproduction) of hierarchy in those exchanges. This is true on an intuitive level—reciprocity, doing unto others as they do unto you—seems a plausible basis for interpersonal equality. But it may also be true at a broader, more systemic level as well. Affective reciprocity—at least how it is defined here, as mutual liking (or disliking)—is in some sense the opposite of domination; and as a human inclination it seems to run just as deep.
At the beginning of the summer, the students’ brains were attuned to one another in ways that seemed likely to reproduce inequality among them—they were rewarded most when they looked at those most sociometrically popular. But what predicted the students’ ultimate feelings about one another were the unique (or idiosyncratic) signals they were giving and receiving, which allowed the emergence of a community built around reciprocal dyads.
Interestingly, we recall thinking during those last days in August that maybe what we were seeing in the sweltering basement offices of INCITE was not just emotional exhaustion and release (although there was plenty of that), but the emergence of a stronger collective identity among the students. It would take a few more months for us to collect the data that supported that idea, but it turns out to be the case.
TRUSTING AND LIKING AND OTHER RELATIONSHIPS
“Liking” isn’t the only relationship we have with others. And what it means to like someone is not as fixed as one might think. We used “liking” as a measure of how people felt about one another because other scholars have used this measure too and because most people think they know what it means. They do, but they don’t necessarily share the same beliefs about what kinds of characteristics undergird it. That is because liking is context specific.
Imagine asking undergraduates how they felt about one another along a number of different dimensions and then asking the same questions to students at an elite professional school, for example, a business school at an elite northeastern university. For social justice–oriented undergraduates, “liking” was highly correlated with thinking that the person they were asked about liked them. Liking for these students was most strongly negatively associated with considering someone “shallow” and “carefree.” In contrast, within the professional school groups, liking was most strongly associated with thinking someone was competent; and it was most strongly negatively associated with considering someone needy. Liking can mean strikingly different things in different social contexts.
Here is another surprising thing we learned from scanning the brains of the Summer for Respect participants: the brain itself seems to organize our responses to others around this distinction. The brain regions that are activated when we look at someone that we trust are the same regions that light up when we look at someone we help or we think we influence. In turn, the regions that are activated when we look at someone with whom we gossip are the same that are engaged when we look at someone with whom we get along or hang out. Our instrumental and affective orientations to others seem to have distinct neural correlates. Our brains organize relations in such a way as to keep largely distinct those relations in which we engage to achieve an external (shared or unshared) goal, and those which are ends in themselves.16 This discovery provided the useful clue for organizing, which we discuss in chapter 5—namely, that it is far less important for the organizer to be a friend of those he or she organizes than to be trusted by them.
The new idea here is that there is a neural foundation to the distinction between affective and instrumental relations. Organizers have long thought this, albeit in much less academic language. Why does it matter? Well, for the reasons we embarked on this project: to figure out how people might change the social circumstances in which they live. The distinction we’ve discovered at the neural level supports the organizer’s intuition that friendship and cooperation are different kinds of relations. And it is cooperation, rather than friendship, that, under the right conditions, can lead to social change.