I liked the example about Touching the Void too, but for a slightly diffferent reason. In my class "Globalization and Religion" our instructor didn't order books for us at the bookstore. This was intentional on her part; she believed we could save money buying online (the books were too new for any real savings, though). Anyway, what happened is that we, the 6 of us, ordered from Amazon (separately) and as a result of that effort, the books we ordered became recommeded pairs. So Amazon started telling the world that people who bought John Keane's Global Civil Society? also bought Phillip Goodchild's Capitalism and Religion. I thought is was amusing, but also telling that such a small group of buyers could affect the front page for any one title.

I was thinking as read the part about Rhapsody. The Power law curve predictably drops off at a certain point, but then re-emerges further out on the tail, particularly when the tail is aggregated. It seems to me that if we can look at the head of the curve as an aggregate of a few megahits and far part of the long tail as an aggregate of the smaller ones, then what we might actually see would be the inverse of a normal distribution curve, with the middle being the hollow (of mediocrity?). This is a shift in perspective - the wide angle lens - but it seemed worth noticing.

I'm struggling with the examples Anderson brings in. The power law, as I've always thought of it, is dynamic. It is very much like hit bands, particularly one-hit wonders. When Touching the Void coat-tails on the success of Into Thin Air it becomes part of the top 20. Eventually, it will fall back into the lower 80 and be sold on discount at bookfairs. Derek has pointed out how the relationship between TTV and ITA constitutes a small-world network, the same way that Nirvana's success brought Pearl Jam, Soundgarden, Mother Love Bone, etc, and Pearl Jam brought Neil Young back into fashion.

I suppose the biggest problem I'm having with these examples is, well, have you ever been to Rhapsody? It would appear that they cater to the latest hit music crowd. I don't know what their business profile is like, but the last I heard (and it's been a year or so) Amazon.com was not profitable, though they had other interests that were funding the sieve of their online retail adventure (in essence, a shell game to make the publicly traded Amazon.com look profitable). And as I recall, sometime in the past year Netflix had to up their prices and do a reshuffle because they were about to go out of business because they were losing money on heavy volume.

While I understand that the long tail serves niche markets, I don't believe (and will gladly see some statistics that prove me wrong) that the average Netflix or Amazon.com or Rhapsody customer is doing anything more than re-inscribing the power law dynamic. At Rhapsody, I would guess that the most profitable downloads are the hit singles, and people largely don't care about the rest of the album, but Rhapsody has to keep prices high so that they can afford to offer the other 97%.

It also occurs to me that this is not a new thing. I walk into Wal Mart and my chances of finding a Poison Idea CD are just about none. I walk into a local chain, like Record Exchange, and my chances go up depending on the chain. But I walk into Smash Records in Georgetown and it's just as likely that Poison Idea will be blaring from the house sound system, and my chances of finding Usher's latest have gone through the floor. Every town (or the next one over) has its own verson of Smash Records (prices, at Smash, are sometimes higher, but because they also carry independent labels like Discord, some CDs are $8).

I'm also reminded as I'm typing this of the case of WHFS 99.1, Baltimore. For something like 40 years, HFS was the leader in alternative rock for the Baltimore/Washington area. In the early 1990s they started a local festival show that got so popular they had to move it to RFK Stadium. They played music by bands people had never heard of, and it rocked. Well, alternative became the mainstream, and about two months ago, HFS became a Spanish radio station. The market was saturated. In the space of 10 years, they went from an industry leader to the long tail, but whose to say that Spanish radio won't be the next big thing?

Customers might be reinscribing the unequal distribution depicted by the power law, but the power law is just a model applicable to any scale-free network involving preferential attachment (or choice), growth by new nodes and fitness (dispersion of energy resulting from a new node). In other words, this stuff applies to commerical examples, but it also applies to social examples. I'm interested in complicating the power law as a smooth curve, replacing it instead with arc texturized by the lesser, locally defined curves populating the tail.

In the book example, one kind of cluster might form around topic: high-risk mountain climbing. As a lesser sphere, those books topically associated would be carried toward the curve because of the fitness of a new node--the new, popular book on climbing. So the acceleration is by small-world association; the cluster--nested in the tail--gets pulled into greater circulation. So for individual customers *not* to reinscribe the power law dynamic, they would have to make their purchasing decisions randomly, yes?

I like this Derek. As I read for this week I keep thinking of Power Laws as descriptive and not prescriptive, so when Anderson advocates changing how we view movies and music popularity to think about the long tail, I begin to wonder if this ideology can really be applied to social theory and collective social action. I tend to think that this larger framework for discussing trends does not really give us any insight into how to make trends begin. In other words, how can the long tail and power laws help us to change the world. (I know not really the intention, but nevertheless . . .) Your comments about social networks and even Vygotsky are helping me think through this quandary a bit. How can power laws take into account word of mouth, social clusters, etc. It seems as if social change can happen through networks it is at this level -- clusters of like minded individuals -- where the change can happen.

My exposure to Vygotsky is pretty light, but I wanted to throw in a comment about the dynamics of the power law and the long tail with respect to social change.

It seems to me that when it comes to organizations or movements engaged in social change, there is a period of "newness" when there is enthusiasm and a certain chain reaction of interest and a push toward the goal. But many organizations that start out with a bang end with a quiet fade. In this scenario, the top of the power law curve would be "big names" - think Greenpeace of Amnesty International or Habitat for Humanity - the organizations that reach smaller groups, like local community centers or groups, make up part of the long tail. Somewhere in that curve are the groups that have a splashy entry, fade into the background, but don't go away. They chug along in some obscurity, but if there is a right moment, an opportunity that builds to a peak, they are there. And that moment can bring them back into the limelight again. Or the collective action of many such groups in the tail can make for changes that no one group alone is responsible for.

I'm not sure that's why Jen or Derek were thinking of, but it made a certain sense to me.

I didn't start out thinking about activism as much as activity and energy in a local or regional sphere. It makes sense to me that all clusters (whether social activism or a fad-accelerated market) enjoy a kind of activation or stimulation with new ideas, products, causes, events and so on. And in terms of activism, excessive activity might be as problematic as stagnation or the "quiet fade" you ref, Chris. I'm probably not explaining it very well in this comment, but I tend to think of power laws as a useful explanatory device (descriptive, as you say, Jen), helpful for mapping activity and relationships in all sorts of ways. Understanding them might not revolutionize social/political activism, but it might helps us recognize organizational variables we wouldn't see without network theory.

Derek's discussion of sub-spheres, desire and contentment leads me to return to Shirky's comment that

[p]ublishing an essay and having 3 random people read it is a recipe for disappointment, but publishing an account of your Saturday night and having your 3 closest friends read it feels like a conversation, especially if they follow up with their own accounts.

In our readings, the rhetorical strain in the discourse about power laws, as the term suggests, tends toward narratives of power. This is reinforced by the contexts of marketing and celebrity, the first regarding successes in the competitive marketplace and the second, to invoke Weinberger's point about interest as a driving force in the Web, on bids to attract attention (links).

I wonder what other narratives, or micro-climates of meaning and value, are possible, especially for life in the long tail, other than its boom or bust market potential. The quote above suggests one alternative axis of meaning (I think also of Madeline's blog). And the discussion here about activism and social change, while still a narrative of power, suggests another. In Jen's comment above, reflecting on how power laws could help us enact social change, I see a possibility for revising Bitzer's understanding of the rhetorical situation (as a context calling for change) in terms of the context of networks.

When he distinguishes the thin attention of mass appeal from the more engaged, robust variety of niche attention (groupies, posses, crews, various fandom clusters--the *always* interested), Anderson begins to get at something like this too, Henry, and I like the move to see network models as a way to complicate Bitzer's model. Network structures also force an intervention on commonplaces about audience, asking us instead to accept that audiences are relatively more proximate--in interest, power, space--than others. To differentiate the frame of power, we might prefer a schema of interest/attention/engagement instead.