, ,

In 2007 Harvard researcher Robert Putnam published the long-awaited results of his research which showed that increased diversity lead to a reduction in social capital. Putnam’s massive study concluded that:

…inhabitants of diverse communities tend to withdraw from collective life, to distrust their neighbors, regardless of the colour of their skin, to withdraw even from close friends, to expect the worst from their community and its leaders, to volunteer less, to give less to charity and work on community projects less often, to register to vote less, to agitate for social reform more, but have less faith that they can actually make a difference, and to huddle unhappily in front of the television (2007: 150).


In the presence of diversity, we “hunker down”, he argued, “we pull in like a turtle” (2007: 149).

These results have been echoed in other studies as well. For example, Dronkers in his study of immigrant neighborhoods in the Netherlands finds that:

1) neighborhoods’ ethnic diversity reduces individual trust in neighborhoods; 2) those with neighbors of a different ethnicity have less trust in neighborhoods and neighbors 3) a substantial part of the effect of neighborhoods’ ethnic diversity on individual trust can be explained by the higher propensity of having neighbors of a different ethnicity. We conclude that ethnic diversity can have a negative effect on individual trust. (Dronkers 2008)


And Hero concurs that “When we considered the interaction of diversity and social capital, a powerful dampening effect of the former on the latter was shown” (Hero 2007: 157).

The explanations for the findings of Putnam’s and other similar studies have usually been couched in moral terms: liberals see it as confirmation of persistent prejudice; conservatives see it as a confirmation that multiculturalism is destructive to society. Both sides see social capital as an important good in society, but they differ in that whereas liberals believe that social capital and diversity are compatible, and that the promotion of social capital is frustrated by prejudice, conservatives believe that social capital and diversity are incompatible and that the good of social capital can only be achieved by reducing diversity through integration. My hope here is to resolve this dispute by offering a non-moral explanation of the mechanism by which social capital is created and maintained, and thereby come to understand how this mechanism is in turn undermined, so that ultimately we can come to judge the compatibility of social capital and diversity.

In his groundbreaking account of social capital, Coleman wrote that “Social capital is defined by its function” (Coleman 1988: 96). If so, it should prove fruitful to consult the literature that has been dedicated to the understanding of the notion and nature of natural function, and see how this may be applied to social capital. This might provide insight into the distinctive function of social capital: what is its function, how it performs its function, and how it fails to perform it. Specifically, I will apply the teleofunctional framework developed by philosopher Ruth Millikan to these ends. There are other accounts of function in the literature, but Millikan’s is specifically geared towards understanding the nature of interpersonal cooperation and communication, and as such is equipped with a set of conceptual tools applicable to the study to social capital.


  1. Teleosemantics

Millikan’s great insight in her landmark Language, Thought, and Other Biological Categories was to look language, thought, and behavior from an evolutionary frame to apply this to the philosophical problems of mind and semantics. Starting from a naturalistic account of the phenomena of teleological function, Millikan is able to move up into the spheres of intentionality, meaning, representation, and interpersonal communication. For Millikan, functional items like hearts and kidneys get their functions not by what they currently do, or have a disposition to do, but by possessing a certain history. Specifically, there must be a history of both reproduction and selection. “Reproduction” is to be understood counterfactually as had the ancestor differed in some respect, the descendent would likewise differ in that respect. Picture the way the words on the paper that comes out of a copying machine are copies of the words on the original, or the way that children’s genes are copies of their parents’ genes, or the performance of social customs such as shaking hands or bowing when greeting are copies of previous performances of those gestures, or the way manufactured items on an assembly line are copies of a blueprint or prototype (1984: 23).[1][1]

The second requirement is that the item must be selected for the possession of a feature. “Selection” here is to be understood not as conscious choice, but in the Darwinian sense that the reason an item or feature exists is that this feature has correlated sufficiently often with some effect, and this effect helped account for its reproduction as opposed to things lacking this effect (1993: 35). The effect in question is the proper function of the item. In the case of biological items such as organs or inherited behaviors such as mating displays, the proper function is that effect an item’s ancestors had that accounted for the proliferation of the genes responsible for its production (1993: 14). Put more intuitively, a specific animal mating display, for example, has proliferated because this behavior corresponds more positively with the ability to attract mates than some other behavior. This behavior was selected by natural selection over some other behavior at least in part because of that correlation; those who performed this display were more likely to attract mates, and thus to pass these behavior-producing genes into their progeny, than those who didn’t. Hearts, to take another example, have pumping blood as their proper function not because they currently pump blood, or have the capacity to pump blood, but because the genes that produce hearts are copies of an ancestor’s genes, and the reason these genes have proliferated is because they correlate with the production of things that pump blood (1993: 35). Thus diseased, malformed, and damaged hearts that lack the ability to pump blood still have the pumping of blood as their proper function because the genes that produced them are copies of genes that have proliferated because they sufficiently often produce things the pump blood, whether or not the current item in question has this ability. Many mating displays fail to perform their mate-attracting function, maybe they even fail more frequently than they succeed, but they succeed often enough to make it worthwhile to pass on the genes.

It is not only the case of the copying and selection of genes that thereby have proper functions in this sense. The account of selection merely requires that a feature be reproduced because it correlates positively with some effect more positively than if it were lacking this feature. Thus things may have proper functions that exist for reasons other than natural selection working on genes. Manufactured artifacts are also copied and selected because they produce some effect (1984: 28). Hammers have driving nails as their proper function because it was the ability to drive nails that has lead to the copying of these artifacts by humans in manufacture. In addition to innate behaviors, a learned behavior can have a proper function if it is reproduced because it leads to a reward; it being the function of the behavior to bring about this result. Behaviors learned through trial and error or through imitation, for example, fall into this category (1984: 28). When a child first learns to imitate (reproduce) light-switch-flipping behavior because they have observed the correlation between light-switch-flipping and room illumination, the proper function of this behavior is to illuminate the room. The proper function of taking the bus is to arrive at one’s destination; the proper function of opening the refrigerator is to get food, and so on.   Likewise, saying/reproducing language items such as words, sentence syntax, phonetic accents, and so on have proper functions; the child learns to iterate tokens of “ba-ba” because this correlates sufficiently often with the reception of a bottle and thus food.

In some cases two items have a common purpose that has been achieved in the past by each party reproducing their share in a pattern of behavior. This serves to coordinate each one to the other in order to cooperate in the achievement of this common purpose. For example, drivers have a common interest in avoiding collisions, so, in the United States and many other countries, a convention was instituted whereby drivers drive on the right. Drivers thus coordinate with one another in order to cooperate in the achievement of their common goal. Other countries of course may have instituted the similar convention of driving on the left. Producing and maintaining this mutual adaptation whereby each party contributes to the shared goal is the “stabilizing” function of the item or behavior; it is what keeps both parties to the coordination responding in standard ways (1984: 31, 2005: 54).

One such case are the reproduced patterns involved in certain animal mating dances where both the producer of the dance and its audience (called the “consumer” by Millikan) have a common purpose and have come up with a convention in order to coordinate with one another in order to reach it. The intended audience “expects” a dance of a certain form to be performed, its conventions followed, and the producer expects the audience to react in a standard way to its performance. Because consumers often enough respond to the dances in a way that benefits them, producers are encouraged to keep producing. And because responding to the dances aids consumers, they are likely to keep responding in the standard way. Despite the fact that often the dance may fail in its purpose, it is more likely to succeed than some random motions, and that makes it worthwhile to keep it in use.

Language devices have their stabilizing functions as well. For instance, if the listeners of an utterance reacted randomly to one’s speech it would be pointless to continue to make those utterances, and the speaker would soon stop. Likewise, if speakers’ utterances never communicated useful information, or if the hearer was too often manipulated against their best interest, hearers would soon stop believing what is said (1984: 31). Speakers and listeners thus arrive at a convention to which each is attuned in order for each of them to succeed in performing the stabilizing function. Speakers must produce forms that the listener is prepared to accept, and the intended audience, the consumers of the language form, must react in ways that reinforce the speaker in so speaking. For instance, it is a convention amongst English speakers that we utter “dog” when trying to discuss dogs. Likewise, it is a convention amongst English writers that we write the characters d, o, and g in order when we are trying to discuss or refer to dogs. One’s hearer must be coordinated with this convention and so be prepared to respond to the utterance of “dog” such as to know that it refers to dogs. It will do no good to utter “dog” amongst those who are not coordinated with the speaker by the possession of a common linguistic history so as to respond to this utterance in a predictable way.

Speakers within a language community are, simply, adapted to an environment in which hearers are responding, sufficiently often, to the forms speakers produce in ways that reinforce these speaker productions. Correlatively, hearers in the community are, simply, adapted to conditions under which speakers, sufficiently often, produce these language forms in circumstances such that making conventional responses to them aids those hearers (2005: 57).

Thus, people continue to utter “dog” in order to draw attention to dogs only insofar as listeners often enough continue to respond to this utterance appropriately. It is in both the speaker’s and hearer’s interest to continue to respond in this way since the hearer is liable to gain useful information and the speaker is liable to meet his goal of spreading such information. “Dog” has been copied from person to person for generations because it is successful in doing so in relation to dogs. The stabilizing function a linguistic form performs is one of the aspects of the term’s meaning (2005: 58).

There are ways of communicating besides through the use of spoken human language. There is the common phenomenon of non-spoken communicating signs that have been designed to coordinate between producer, consumer, and environment. In these cases items with stabilizing functions may attain intentionality. To use a now classic example, after finding a source of nectar, a honey bee returns to the hive and does a dance. Other bees watch the dance and so learn the location of the nectar relative to the sun and the hive. They then fly off in the direction of the nectar in order to retrieve it and bring it back to the hive. Millikan calls items such as the dance of the honey bee “intentional icons” because they are about the location of nectar (1984: ch. 6). In order to be an intentional icon the sign in question must, firstly, be able to vary in accordance with variations in the environment. Secondly, it must be a function of the producer of the sign to produce it for a consumer in accordance with certain mapping rules by which the sign maps its intended environmental feature. Finally, it must be a function of the intended consumers of the icon to use it in the way the mapping relation indicates (1993: 106).

For example, the form of the dance of the honey bee varies depending on the location of the nectar relative to the sun and the hive, it is produced by the dancing bee in order to indicate this relation, and the watching bees then use the mapping relation indicated to direct their direction of flight. In another example, soldiers wear their ranks on their uniforms where they can be clearly perceived by their intended audience. The insignia varies according to the soldiers rank and both tells consumers what the rank is and so prescribes appropriate behavior. Finally, beavers slap their tails on the water when danger is near in order to tell listening beavers of its presence. The slaps vary with the time and place of danger, and the listening beavers use this icon to initiate hiding or other avoidance behaviors.

For each item that possesses a function in the sense described here there will be an explanation of how the item has historically managed to perform this function. This explanation will mention how the structure of the item in question has managed to “do its job” historically, what conditions were in effect, what the environment was like that allowed the item to successfully perform it function. Millikan calls such an explanation a “Normal” explanation and the conditions that have historically held in order for the item to succeed in performing its function “Normal” conditions (1984: 33). “Normal” is capitalized to prevent confusion that might occur if one was to think that Normal conditions are average or frequent since “normal” often has that connotation. For example, just think of how few sperm manage to perform their function of fertilizing an egg, or how infrequently the skull needs to perform its function of protecting the brain from impacts. It might be helpful to think of Normal conditions as “activation conditions” or “enabling conditions.” In abNormal conditions an item will fail to perform its function, or at least fail to accomplish it in accordance with a Normal explanation. Diseased hearts are in abNormal conditions, being underwater for extended periods is an abNormal condition for otherwise healthy lungs, and whatever it is that prevents a specific sperm from fertilizing an egg is also an abNormal condition.

We will put all this theory to use in part 2.