The Connection

"Humberto Maturana's theories of autopoiesis and cognition form the core of the new understanding of living systems that is now emerging at the forefront of science. Maturana was one of the first scientists to identify the relationship between mind and matter as one between process and structure, thus finally overcoming the Cartesian division that dominated Western thought for hundreds of years."
    -- Fritjof Capra (1)
Maturana, a biologist, invites us to understand life on earth as a series of levels or cycles of interaction. Each response of one part of the network to activity in another part changes the quality of the response. This 'autopoietic' transformation -- called becoming or learning -- defines different levels or orders, which in familiar terms are typified by the cell, the organism (a community of cells), and the community of organisms.

But he goes further: the process of making higher-level patterns (which themselves nest within other patterns) is, essentially, intelligence. A worthy question is whether the increasing prevalence of computers contributes to and furthers, or detracts from and stultifies, this 'organic' quality.

On one hand (or mouse button), algorithms can draw out from massive amounts of data connections -- patterns -- that were never suspected. Sick a search engine on the most obscure pair of keywords, and almost invariably it produces a hit; that is, evidence that someone, somewhere, had already beat you to the (pattern-)punch. (Whether the pattern is of any use to you is another question entirely.) But this power all too clearly isolates the human element from the 'chore' of examining the data directly: how can we expect to have new insights if we no longer look at (rather cherish and absorb and massage) the old ones?

On the other, the opportunity to jot down every fleeting thought (whence many a remarkable 'breakthrough' has come) with little need for preparation can lead to the recognition of patterns that might otherwise be missed by the most powerful of algorithms. Here, the issue is not the amount of data, but the very process of conversion of experience to information.

Current consensus has it that algorithmic pattern detection is not to be considered 'intelligence.' In the past, I have been reluctant to agree, since the criterion, the cut-off point, was left undefined. Now -- having been able to recover and develop the fleeting thought which prompted this page -- I am ready to suggest that the second sort of process necessarily and sufficiently discriminates 'mechanical' from 'mental' pattern-making. (We will know 'AI' has arrived when the processor not only connects the dots, but -- on the odd occasion -- jots.)

This observation onfirms my suspicion that its not the quantity of data, but the quality -- and that while it's easy to interpret quantity as a property of the data set, quality is something we have to work into it, rather like one kneads bread to develop the gluten. This cannot be done one datum at a time, because it's a relational process -- comparing and contrasting, adducing reasons and educing implications, making and baking patterns that connect. (2)


If the computer comprises the web of relations by which a single 'cellular' thought fits into and becomes part of an 'organismic' frame of reference, is the Internet a community of 'points' of view? That this question seems to arise naturally in the course of almost any netted discussion suggests that our readiness to look for such autopoietic evidence is not misplaced. However, these discussions never seem to go very far, bogging down in debate whether it is useful to speak of the net, when the various tentacles (mailing lists, MOOs, chat groups, web-rings, etc.) are still very isolated from one another.

Indeed, in view of the (circuit court) rejection of COPA, it seems altogether likely that separate nets will come to be seen as the only practical way to reconcile the demand for 'local' standards and the imperative perception that 'information wants to be free.' But this segregation will hardly correspond with the different 'technologies' referred to; rather, I envision a population of 'service communities' between which traffic is 'gated' according to their individual standards of propriety, or specifically negotiated 'trade agreements.' An individual 'user' would elect to join the community that seems best to fit hyr own (cellular) standard -- or, conversely, a community would invite users to participate.

One obvious difficulty in formulating this vision is a language which, while it has any number of terms for collections and aggregations (and reduces them to singular entities; e.g. train, battery, and magazine), has few tools with which to develop series of aggregates (one may speak of a convoy of trains, perhaps, but then what -- a gaggle of convoys?)

A more serious problem is the lack of a pedagogy; that is, a clear means to convoy cells between levels -- from cellular isolation to communal integration, from communal isolation to gaggular coherence, and so on. Every attempt in history has always stalled out, as if arriving at the next stage is an end in itself; that is, the end. Children become students -- but how many become teachers? Worse, children become parents -- but how many 'turn out' to be 'good' parents? Computers make the task of information processing more efficient -- but if they contribute to a sense of global community it is, at most, a happy accident, and one which is likely to be rectified by the imminent corporatization of internet administration, as the only sequel to a corporation is a bigger corporation, not a better one. Quality, however one chooses to define it under the commercial bale, reduces to quantity.

To this quandary, I can only suggest that, individually and collectively, we make every effort to seek out and sustain autopoietic potentials wherever we find them; to insist that, whatever terms come into currency, they always describe only endpoints of a full spectrum of possibility; to demand that the 'information infrastructure' assures not merely 'access' to 'resources,' but the opportunity to be heard. Highways can be arteries, but they must be veins as well. Otherwise, the body politic becomes the body putrescent, in which only Homo vermix will thrive.

In particular, I think we have a chance to develop a metaphor for cyberspace propagation which avoids the easy polarities which will be offered: of walled cities within an anarchic wilderness, or oases in a desert wasteland. Rather, taking a cue from Dante (whose seven circles of Hell were matched by seven of Paradise), we can imagine, and operate, an Internest. In it, negotiable pathways would be defined not by venial and mortal sins but by degrees of proficiency, the primary criterion of which would be simply, Has the candidate helped guide or mentor another through the prior level? 'Each one promote one' is not only catchy but comprehensive. And only through comprehension does any tool, no matter how 'advanced,' serve to advance its user.


(1) In reference to a [course ] at Schumacher College. See also [Biology of Cognition]
(2) On the E-Lekh, a 'frameset' facilitates this exercise. Choosing from a menu (which is itself split), one can open any two pages in a 'split screen,' and then open a mail form (which is linked at the bottom of the menu) in order to contribute between them. The frames can be dragged to maximize the desired panes.
(3) Can goodness be anything but the quality of parenting? -- to which word I apply the intuitive etymology para <separate> + ent <p. part. nominative marker>: one who 'raises' or 'cultivates' another to perform independently. Despite every effort, our machinery has yet to replicate this ability, due (I suspect) to its inability to distinguish form and function, container and content, etc. See Intro.

logos links future demos past
Comments? Contributions? Write to Serchan.