I liked Granovetter’s paper on “The Strength of Weak Ties” quite a bit. The concept of information diffusion gave his investigations of social networks the clarity I felt was lacking in the questions I was complaining about earlier. While information diffusion is certainly not the only perspective from which you might want to look at social networks, it is very useful for the kinds of problems Granovetter seems to be concerned with (labor markets, neighborhood coalitions, etc).

While Granovetter argues convincingly for his hypothesis that information diffuses primarily through weak ties in social networks, I found some of his starting assumptions problematic.

Granovetter defines stronger ties as ones that involve larger time commitments. Personal experience tells me that this is not the case. My family ties are stronger than my professional ties, despite the fact that I spend far more time with my co-workers than my family. Among my friends, there seems to be little correlation between time spent together and tie strength: I have very strong friends whom I see once a year or so, and “weaker” friends whom I see quite often.

Granovetter’s theory that network overlap correlates with tie strength also conflicts with the intuition that strong ties are created by shared experiences. For example, two travellers who meet overseas may form a strong bond based on their shared experiences in exotic lands. This bond may last a lifetime despite the fact that the two travellers live in different places and share no mutual friends. These kinds of stories seem to be fairly common among soldiers and expatriates.

In general, however, Granovetter’s theory coincides with my personal experience. I’ve found my most satisfying and enjoyable jobs through weak ties. And I generally find that useful information comes from people outside my immediate social group. For example, I rarely receive interesting information from my fellow SIMS students. This isn’t a reflection on them, but is a consequence of the fact that we all have many common interests, and thus I’ve usually already heard about the things I hear from them.

My class on Analysis of Information Organizations is focusing on the topic of social networks. The instructors have asked us to answer the following questions:

• What are the purposes of your various social networks?
• What are the shapes of your social networks: how large are they? How do they combine face-to-face and online contacts? How old are they? What are their functions?
• Are you going to use your social networks to get a job? Find a partner?
• How, and when, do you use your social networks?
• Propose a typology of the functions, origin and duration, and size and density of social networks, based upon your own experience.

I have some serious problems with these questions. Let’s take them one by one.

What are the purposes of your various social networks? This question assumes that one can unproblematically enumerate one’s “various social networks.” My first inclination is to point out that there is really only one social network in the world (assuming that there aren’t any tribes cut off from all other civilization yet to be found). Which part of the network is “mine?” Of the part which is mine, how would I (and why would I) divide it into sub-parts? And does it make sense to ask what the the “purposes” of these parts would be? What is the “purpose” of a family, or a group of friends?

What are the shapes of your social networks: how large are they? How old are they? Again, we’re brushing a slew of more fundamental questions under the rug here. But let’s assume that I’ve come up with some way to divide the people I know into groups. Some of these groups may be small, some may be large. But measuring size in that way seems shortsighted; if one of these groups has 40 people, each of whom know only a couple of people other than me, is that part of my network larger or smaller than another part consisting of 3 people who each know 100 people other than me? Measuring age is similarly problematic: if I recently met a friend of my parents, is the part of my social network comprised of the four of us old or new?

How do they combine face-to-face and online contacts? I rarely consider anyone whom I haven’t met face-to-face to be in my social network. But having moved locales and jobs frequently in my life, I take advantage of online tools (primarily email) to keep in touch with people I can’t easily see face-to-face. I use online tools to maintain links, and only rarely for making them.

What are their functions? Back to “purposes” of social networks again. Why this focus on teleological explanation?

Are you going to use your social networks to get a job? Find a partner? Well, yes. I’ve used personal contacts to find jobs in the past and will no doubt continue to do so in the future. And I met my wife through personal contacts, not through a dating service. Having built on-line recruiting software and seen the workings of recruiting companies first-hand, I have no desire to commoditize myself through those kinds of formal systems. It’s kind of amusing that many people who would sooner die than sign up for Match.com have no problem marketing themselves on Monster. The two systems are more similar than different.

How, and when, do you use your social networks? This question basically boils down to “How, and when, do you exchange information with the people you know?” “How” is pretty straightforward: in person, via email, on the phone, and on IRC. “When” is more difficult:

• when I want to gather socially
• when I want to kill time
• when I want to give people information they might find useful
• when I want to learn something
• when I need help
• when I want someone to remember me
• when I feel a duty to

Propose a typology of the functions, origin and duration, and size and density of social networks, based upon your own experience. Having read thus far you can probably guess how I feel about this question. I don’t think it makes sense to classify things without a clear statement of why we are classifying them. For example, the way we classify books will depend strongly on why we are doing the classifying. If am a researcher, I am concerned primarily with things like subject and author. If I am a collector, attributes like “signed” and “first edition” become salient. If I am a manufacturer of recycled paper I may have an entirely different classification system. Any typology of social networks will necessarily emphasize some things and gloss over others. I can’t specify concepts to emphasize unless I know why I’m doing it.