Close-Knit Social Networks and Innovation
What kind of social network is most conducive to the spread of new ideas? What kind is most resistant?
Lesley and James Milroy write in “Social Network and Social Class: Toward an Integrated Model” (1992) that “a close-knit, territorially based network functions as a conservative force, resisting pressures for change originating from outside the network. By close-knit we mean relatively dense and multiplex.” “Dense” and “multiplex” have specific meanings here: “In a maximally dense and multiplex network, everyone would know everyone else (density), and the actors would know one another in a range of capacities (multiplexity). Close-knit networks, which vary in the extent to which they approximate to an idealized maximally dense and multiplex network, have the capacity to maintain and even enforce local conventions and norms.”
In other words, if you have a group of friends, and some of those friends are also your co-workers or members of the same bowling team, your ties with those friends are multiplex. If your friends are all friends with each other, the network is dense. If your friends aren’t all friends with each other but are instead friends with other people (who also don’t all know each other), the network is loose. Dense networks with multiplex ties—close-knit networks, that is—tend to self-police more; as the Milroys write, they “maintain and even enforce local conventions and norms.” The Milroys offer this self-policing as an explanation for how low-status dialects and vernaculars have managed to persist for centuries in the face of pressure from a higher-status standard. From the point of view of the minority speaker community, standard-isms are innovations that can stand out in vernacular speech and be disapproved of.
This is all true, and it may be very helpful in explaining the survival of minority dialects and other nonstandard cultural traits. But close-knit networks may also have certain advantages for the spread of innovations, which may at least partially offset their self-policing conservatism. Let’s leave self-policing aside for a moment. When an innovation enters a social network, either through dissemination from outside or through a new idea or mutation in one of the network’s members, it needs certain conditions in order to spread. Many innovations, particularly linguistic ones, have meanings that are not immediately self-evident but require a shared context in order to be correctly interpreted. Social actors can be most assured of finding this shared context in a close-knit network.
For example, a year or two ago I came upon both the term hapax legomenon and the idea that most sentences spoken in English today (and presumably in any other modern language) have never been spoken before. I decided to define “hapax” as “a sentence that has never been spoken before,” reappropriating an uncommon scholarly term to use in daily life. As I used it, the word with this meaning was picked up by friends of mine, and it’s now spreading to people I don’t even know. (Its utility is based on the alienation that occurs when you point out that a sentence that has just been spoken in the course of a conversation is a hapax.) This term, an innovation, has spread more quickly to people with whom I have multiplex ties and who are part of the same dense social network as me. This is quite simply because I’m less likely to use the term when I’m around people I don’t know as well—as is often the case with looser social networks—or with people I spend less time around—as is the case with less multiplex ties—because they are less likely to understand what I’m talking about, and having to explain it is a hassle.
That was an example of a lexical innovation that was helped in its spread by a dense social network. There are also examples of other kinds of linguistic innovation that spread similarly. For example, let’s say some friends of mine use a particular accent to convey an attitude towards what they’re saying, perhaps a skeptical or sarcastic attitude. They might avoid using that accent to convey that attitude around people they don’t know, since it might be misinterpreted; the innovation will therefore only spread through close-knit social networks of which they’re a part.
These are linguistic examples, and I leave it as an exercise to the reader to come up with non-linguistic innovations that are or aren’t helped by the density of a social network and the multiplicity of its ties. I don’t know how exactly this innovation-promoting property of close-knit networks interacts with the self-policing property the Milroys discuss. My guess is that the self-policing property comes into play more when an innovation seems to be imposed by an outside force, whereas innovation stemming from within the group itself triggers the innovation-promoting property more. In reality, of course, I’m sure it’s more complicated than that.
Questions for commenters: What non-linguistic innovations can you think of that are or aren’t helped by a close-knit network? How do you think the innovation-promoting property and the self-policing property of a close-knit network interact? The Milroys emphasize that the self-policing property is territorially based; is the innovation-promoting property also territorially based, and is it more or less so?
This entry was posted on May 26, 2009 at 1:51 am and is filed under Vicente with tags Social networks. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.