Tag Archives: social network analysis

Influence and Butterflies

Seems like “influence” is a key theme in social media, these days. An example among several others:

Influenceur, autorité, passeur de culture ou l’un de ces singes exubérants | Mario tout de go.

In that post, Mario Asselin brings together a number of notions which are at the centre of current discussions about social media. The core notion seems to be that “influence” replaces “authority” as a quality or skill some people have, more than others. Some people are “influencers” and, as such, they have a specific power over others. Such a notion seems to be widely held in social media and numerous services exist which are based on the notion that “influence” can be measured.
I don’t disagree. There’s something important, online, which can be called “influence” and which can be measured. To a large extent, it’s related to a large number of other concepts such as fame and readership, popularity and network centrality. There are significant differences between all of those concepts but they’re still related. They still depict “social power” which isn’t coercive but is the basis of an obvious stratification.
In some contexts, this is what people mean by “social capital.” I originally thought people meant something closer to Bourdieu but a fellow social scientist made me realise that people are probably using Putnam’s concept instead. I recently learnt that George W. Bush himself used “political capital” in a sense which is fairly similar to what most people seem to mean by “social capital.” Even in that context, “capital” is more specific than “influence.” But the core notion is the same.
To put it bluntly:
Some people are more “important” than others.
Social marketers are especially interested in such a notion. Marketing as a whole is about influence. Social marketing, because it allows for social groups to be relatively amorphous, opposes influence to authority. But influence maintains a connection with “top-down” approaches to marketing.
My own point would be that there’s another kind of influence which is difficult to pinpoint but which is highly significant in social networks: the social butterfly effect.
Yep, I’m still at it after more than three years. It’s even more relevant now than it was then. And I’m now able to describe it more clearly and define it more precisely.
The social butterfly effect is a social network analogue to the Edward Lorenz’s well-known “butterfly effect. ” As any analogy, this connection is partial but telling. Like Lorenz’s phrase, “social butterfly effect” is more meaningful than precise. One thing which makes the phrase more important for me is the connection with the notion of a “social butterfly,” which is both a characteristic I have been said to have and a concept I deem important in social science.
I define social butterflies as people who connect to diverse network clusters. Community enthusiast Christine Prefontaine defined social butterflies within (clustered) networks, but I think it’s useful to separate out network clusters. A social butterfly’s network is rather sparse as, on the whole, a small number of people in it have direct connections with one another. But given the topography of most social groups, there likely are clusters within that network. The social butterfly connects these clusters. When the social butterfly is the only node which can connect these clusters directly, her/his “influence” can be as strong as that of a central node in one of these clusters since s/he may be able to bring some new element from one cluster to another.
I like the notion of “repercussion” because it has an auditory sense and it resonates with all sorts of notions I think important without being too buzzwordy. For instance, as expressions like “ripple effect” and “domino effect” are frequently used, they sound like clichés. Obviously, so does “butterfly effect” but I like puns too much to abandon it. From a social perspective, the behaviour of a social butterfly has important “repercussions” in diverse social groups.
Since I define myself as a social butterfly, this all sounds self-serving. And I do pride myself in being a “connector.” Not only in generational terms (I dislike some generational metaphors). But in social terms. I’m rarely, if ever, central to any group. But I’m also especially good at serving as a contact between people from different groups.
Yay, me! 🙂
My thinking about the social butterfly effect isn’t an attempt to put myself on some kind of pedestal. Social butterflies typically don’t have much “power” or “prestige.” Our status is fluid/precarious. I enjoy being a social butterfly but I don’t think we’re better or even more important than anybody else. But I do think that social marketers and other people concerned with “influence” should take us into account.
I say all of this as a social scientist. Some parts of my description are personalized but I’m thinking about a broad stance “from society’s perspective.” In diverse contexts, including this blog, I have been using “sociocentric” in at least three distinct senses: class-based ethnocentrism, a special form of “altrocentrism,” and this “society-centred perspective.” These meanings are distinct enough that they imply homonyms. Social network analysis is typically “egocentric” (“ego-centred”) in that each individual is the centre of her/his own network. This “egocentricity” is both a characteristic of social networks in opposition to other social groups and a methodological issue. It specifically doesn’t imply egotism but it does imply a move away from pre-established social categories. In this sense, social network analysis isn’t “society-centred” and it’s one reason I put so much emphasis on social networks.
In the context of discussions of influence, however, there is a “society-centredness” which needs to be taken into account. The type of “influence” social marketers and others are so interested in relies on defined “spaces.” In some ways, if “so-and-so is influential,” s/he has influence within a specific space, sphere, or context, the boundaries of which may be difficult to define. For marketers, this can bring about the notion of a “market,” including in its regional and demographic senses. This seems to be the main reason for the importance of clusters but it also sounds like a way to recuperate older marketing concepts which seem outdated online.
A related point is the “vertical” dimension of this notion of “influence.” Whether or not it can be measured accurately, it implies some sort of scale. Some people are at the top of the scale, they’re influencers. Those at the bottom are the masses, since we take for granted that pyramids are the main models for social structure. To those of us who favour egalitarianism, there’s something unpalatable about this.
And I would say that online contacts tend toward some form of egalitarianism. To go back to one of my favourite buzzphrases, the notion of attention relates to reciprocity:

It’s an attention economy: you need to pay attention to get attention.

This is one thing journalism tends to “forget.” Relationships between journalists and “people” are asymmetrical. Before writing this post, I read Brian Storm’s commencement speech for the Mizzou J-School. While it does contain some interesting tidbits about the future of journalism, it positions journalists (in this case, recent graduates from an allegedly prestigious school of journalism) away from the masses. To oversimplify, journalists are constructed as those who capture people’s attention by the quality of their work, not by any two-way relationship. Though they rarely discuss this, journalists, especially those in mainstream media, typically perceive themselves as influencers.

Attention often has a temporal dimension which relates to journalism’s obsession with time. Journalists work in time-sensitive contexts, news are timely, audiences spend time with journalistic contents, and journalists fight for this audience time as a scarce resource, especially in connection to radio and television. Much of this likely has to do with the fact that journalism is intimately tied to advertising.

As I write this post, I hear on a radio talk show a short discussion about media coverage of Africa. The topic wakes up the africanist in me. The time devoted to Africa in almost any media outside of Africa is not only very limited but spent on very specific issues having to do with Africa. In mainstream media, Africa only “matters” when major problems occur. Even though most parts of Africa are peaceful and there many fabulously interesting things occuring throughout the continent, Africa is the “forgotten” continent.

A connection I perceive is that, regardless of any other factor, Africans are taken to not be “influential.” What makes this notion especially strange to an africanist is that influence tends to be a very important matter throughout the continent. Most Africans I know or have heard about have displayed a very nuanced and acute sense of “influence” to the extent that “power” often seems less relevant when working in Africa than different elements of influence. I know full well that, to outsiders to African studies, these claims may sound far-fetched. But there’s a lot to be said about the importance of social networks in Africa and this could help refine a number of notions that I have tagged in this post.


Facebook Playing With My Mind

Took a look at the homepage for my Facebook account and I notice something new, below the birthday announcements. Some profile summaries with a mention that I might know these people. Nothing really awkward there, probably just a new feature. Although, Facebook has this strange (and potentially annoying) habit of changing features without warning us.

But still not mindblowing, or even mindplaying.

There’s a “Show All” button in that box and, when I click on it, I get to a Friend Finder page where I see a series of profile summaries with the heading: “People You May Know. Found based on your existing connections. Do you know any of these people? Add people you know as friends to make these results even better for you.”

Next to each profile summary:

You both know: [links to mutual friends]
Add To Friends|(View Friends)|Message

Again, nothing really weird. (Without warning,) Facebook browsed my connections and found some mutual friends. Some applications do things like these.

But, here’s where things get a bit less obvious: the first time I look at this page, I see a list of people I don’t recognize with mentions of some of my contacts (friends and acquaintances). Overall, these contacts are people I had assumed were unconnected. Granted, they all live or have lived in Montreal (my hometown). And some of them are somehow involved in music. But even the musicians among them are working in quite separate music scenes within Montreal’s music landscape.

According to this list, Richard (one of my contacts) has eleven connections in common with twelve of my friends and acquaintances. These twelve friends and acquaintances of mine presumably have little in common with the people that both Richard and I know. None of these twelve contacts of mine are connected directly to Richard on Facebook. They all know some of Richard’s contacts but my connections to them are very diverse: former students, former bandmates, a childhood friend, a fellow brewclub member, etc. I’ve met these people at very different stages of my life and I just couldn’t assume any of them would know one another. Again, all of these people have some connection to Montreal but given Montreal’s population, I find it quite surprising that my network would cluster so much across contact types.

I felt compelled to send a couple of messages about this. To Richard (this acquaintance of mine who seemed to have many mutual acquaintances with people I know). And to two of the people who were listed as possible acquaintances of mine (one of whom I probably did meet, a number of years ago).

Fascinating stuff for a social scientist like me.

But where it gets mindplaying is when, coming back to the Friend Finder page, the list of possible acquaintances is radically different from what it was the first time. This time, most of the people in the list belong to YulBlog, Montreal’s blogging community. That community has a relatively high clustering coefficient so I basically assumed that many of those YulBloggers are friends with some of my blogging friends. I did meet with several of these bloggers at blog meetings but I prefer letting them judge whether or not we should be linked through Facebook. So, this new Friend Finder page looks pretty normal, Which makes the first Friend Finder page seem more unusual. Playing with my mind.

It’s possible that the first Friend Finder page was a glitch. Facebook has been known to have some bugs recently, as they implement (some would say “impose”) changes in the way they handle things like privacy and contact lists. But, looking at Richard’s contact list, it does seem that these people really are all connected, albeit indirectly.

Lest you should mistake my enthusiasm for flabbergastment, I must say that while I find these connections surprising, I still understand that they’re fairly easy to explain. The effect, though, is one of puzzlement at the extent of the Small World Effect. I feel as though my world were much tinier and much more clustered than I had ever assumed. Especially the Montreal portion of my social world. And I thought my friends were diverse… 😉

Yes, I know. I should just draw the network chart and let people reach their own conclusions.

Ah, well…