Which collective? The collective tending towards the mean (Was Re: [iDC] Table Tennis, Schmeb 2, List Dynamics, and Autonomous Uses of Sociable Web Media)

nick knouf nknouf at media.mit.edu
Fri Jun 16 00:55:57 EDT 2006

Dear list members,

Perhaps too late for this discussion, but I will write anyway...

I'm new to the list and have followed the recent posts about sociable  
web media with great interest.  Like Miguel, I too harbor intense  
reservations regarding the ultimate "transformative" power of  
collective workspaces like Flickr, MySpace, Wikipedia, etc.  The  
remarks in this post will certainly come across as criticisms, yet as  
Ryan mentioned in his recent post in regard to Latour, I am still  
searching for a "positive" politics that can use these techniques in  
a constructive way.  I lack such a methodology, but I hope that that  
discussions on this list and elsewhere will lead to such a thing.

I also encourage any corrections to things I write!

== Which collective? ==

In the doubly-removed quote in Trebor's original e-mail below,  
O'Reilly mentions the "collective intelligence" in the context of  
blogs.  But of course this "intelligence", to use his term, could be  
said to exist amongst the images on Flickr, the sounds and networks  
on MySpace, and the amalgam of words on Wikipedia.  Even if  
collectives are by most definitions amorphous, they still posses some  
semblance of structure, some set of constraints that in a certain  
sense limits their output to a set of combinations of their inputs.   
To move beyond this simplistic definition, the results of human  
collectives reflect in some fashion the desires, motivations,  
feelings, and knowledge of the individual people who are (or find  
themselves lumped together as) a collective.  The trope of the whole  
(collective) still requires the individual parts (people) and thus  
the individual still influences (in some fashion) the process or result.

Thus when we talk of the "collective intelligence" of tagging  
communities on Flickr or del.icio.us, or aggregation services like  
digg, we have to ask: which collective?  Which group of people are  
taking on the responsibility of adding annotations to these images,  
selecting the news for us to see?  We put our trust into an  
amorphous, oftentimes anonymous, crowd; they select and organize  
information on our behalf.  That collective process is, in at least  
the cases I've seen so far, a simple "majority rules" measure.  The  
popular rises to the top; the marginal and unpopular never makes it  
to the front page.  (See some of Walter Bender's early work in  
electronic publishing, specifically his ideas of the personalized  
newspaper and the "Digital Me": see the end of this page: http:// 
web.media.mit.edu/~walter/)  In many cases we have no practical way  
to respond if we dislike the selection of tags or links: the standard  
response is "get involved", tag things yourself, digg links yourself,  
and so on.  Which might actually be the appropriate request.   
However, it flies against one of the purposes of these sites, which  
is to enable others to experience the results of the collective work  
of the site's users.

So we have to ask ourselves, what are the motivations of people who  
work for free, finding the most novel links, burrowing through the  
depths of horrible site after horrible site to "digg" the novel one,  
developing and tagging all of their photos and works for others to  
browse?  The usual recourse is to the overtly-economic term "social  
capital", the unmeasurable value that somehow determines the stature  
of someone within a social network.  And while, yes, that or some  
other theoretical formulation might be appropriate, but what I want  
to suggest is that these people have a much more measurable amount of  
capital: _temporal_ capital.  The people who post the most to these  
collaborative Web 2.0 sites are those who have the most time to spend  
on the task.  For them, the garnering of social capital (or whatever  
you want to call it) is the best thing for them to do with their  
limited amount of time.  For they could use this time to write a  
poem, cut up zucchini for dinner, or work for their local political  
party; but they would rather work for free for some broader public.   
Those who have other commitments cannot participate to the same  
extent, and I would hazard a guess that this would include the usual  
groups that are commonly disenfranchised when the luxury of time is  

Not only, then, do the results of these collective activities depend  
on people for whom we do not know their motivations; the results also  
depend on those who have the most temporal freedom to participate.   
Not that we do not have similar concerns with more traditional media  
systems.  While a print journalist is spending her time writing the  
article because she is being paid for it, we do not necessarily know  
her motivation for doing so.  Yet we _do_ know who she is, in some  
objective sense; with the interesting exception of anonymous sources  
and pseudonyms, we know who all of the actors are in the news  
article.  The details of the collective are potentially unpredictable  
or unknowable, but in my experiences with these sites, this has not  
happened interestingly in practice.

== The disturbing trend towards the mean in collectives ==

I recently read an intriguing article and discussion on the Edge  
prompted by an essay by Jaron Lanier entitled "The Hazards of the New  
Online Collectivism":

essay: http://www.edge.org/3rd_culture/lanier06/lanier06_index.html
"discussion": http://www.edge.org/discourse/digital_maoism.html

(I put discussion in quotes because, interestingly, it does not use a  
format of either a wiki or commenting system; I'd describe it rather  
as a set of responses limited to selected participants, in a direct  
and oddly comforting contrast to what I was talking about earlier..)

Lanier makes a number of provocative points, but I want to focus on  
the idea of averaging.  As mentioned above, so many of these sites  
use simple rankings, or majority measures, to determine the featured  
tags or the highlighted articles.  This de facto or overt averaging  
of the selections of the collective has infiltrated the traditional  
space as well, with Lanier describing the recent practice of the New  
York Times in "averaging opinions" by featuring op-ed pieces  
supporting intelligent design.  While this might represent a lack of  
scientific knowledge on the part of the editors (disturbing in its  
own right), I would agree with his interpretation.

A collective approach to things tends to marginalize the radical, the  
novel, the contrarian.  Such ideas become lost in the variance,  
defined by terms such as "outliers".  Discourse clusters around the  
mean, with the most potentially interesting and meaningful thoughts  
deemed to be "insignificant" as they fall plus or minus two sigma  
away from mu.  Where is the space for the unorthodox _grounded_  
ideas?  Where the visitors to these sites can really discover the  
imaginative?  Where we can debate the original thoughts that might be  
able to move people towards new means, to new averages, or to  
eliminate the need for a single "average" altogether?  I guess part  
of the design of lists such as this one is to help explore answers to  
the last question, but I am still wondering about the more popular  

Not that I have a solution to this problem (if it exists) of  
averaging.  Simple solutions (if they exist) might be to consider  
other metrics of voting, as some municipalities and states (in the US  
and other places) have used, such as instant runoff voting.  Yet I am  
certain that the real answer is much more complicated.

== And this is where I ask others on the list... ==

Perhaps the not-so-well-hidden value judgments here might be a bit  
too harsh; I certainly would appreciate thoughts from others about  
places where I've considered things in an over-simplistic fashion.   
That is, if you have the time :-)



On Jun 12, 2006, at 10:34 AM, Trebor Scholz wrote:
> [...]
> Nicholas Carr:
> ³Not long ago, the big-time tech publisher and conference  
> impresario was
> talking up Web 2.0 as a means of achieving a "technology-mediated"
> higher consciousness. But a shadow seems to have fallen across
> O'Reilly's optimism. In a commencement speech last month, he  
> cautioned,
> "If history is any guide, the democratization promised by Web 2.0 will
> eventually be succeeded by new monopolies [which] will have enormous
> power over our lives - and may use it for good or ill." A couple of
> weeks ago, after being mugged in abstentia by a hysterical blog mob,
> O'Reilly said the experience "has shaken my faith in the collective
> intelligence of the blogosphere."
> [...]
> Call it whatever you like, sociable web media will be my term of  
> choice,
> but the question of autonomous uses of these participatory web
> architectures despite their enormous potential to turn into a
> participatory panopticon, still matter a great deal.
> -Trebor

More information about the iDC mailing list