Wednesday, February 23, 2005

Renormalization

I 'm still grogging through a hyper-invasive head cold.  'Snot easy to read and write through the sinus pressure, drainage and trips to the kettle for another hot tea.  I've gotten on with reading Robert Connors, Hayden White and a couple of chapters from Barabasi's Linked for tomorrow in addition to a project overview for comp history and a trip to the tax prep office to sign the proper filing forms.  This was only the second time I've sought an accountant's assistance with taxes; with the move, bi-state filing, and limited opportunities for sorting through tax forms, we turned over to the pros.  And today when we were at their office nodding our heads to the double-check of name-spellings and socials, the friendly accountant showed us an itemization asking nearly twice what they'd initially quoted.  I'm too ashamed to share the number; I'll just say that I'm ordering the software next year and doing it myself again.  The $70 worksheet pushed me off the edge (of chair, cliff, reason).  In the late age of computer-spun accounting, no single worksheet prep should cost seventy bucks.

So we signed all the forms and came back home. I was too head-coldy to give the guy hell for puffing up the tax prep bill. 

Once home, about hot tea, I learned this: Peppermint will re-steep for a second good cup.  Raspberry zinger will not.

Renormalization is a term credited to physics (perhaps other fields, too).  According to Barabasi, it comes from the work of Cornell professor and 1982 Nobel prize winner Kenneth Wilson who assigned the term to the event following a physical phase transition.  Renormalization (granted, read second hand) accounts for the paradoxical flux and structural stiffening of real networks--the shift from chaos to complexity, from disorder to system:

By giving a rigorous mathematical foundation to scale invariance, [Wilson's] theory spat out power laws each time he approached the critical point, the place where disorder makes room for order.  Wilson's renormalization group not only called for power laws but for the first time could predict the values of two missing critical exponents as well. (77)

And...

We had finally learned that when giving birth to order, complex systems divest themselves of their unique features and display a universal behavior that has similar characteristics in a wide range of systems. (78)

It's quite possible, even likely, that I'm misunderstanding or misapplying this concept.  Even so, I find something catchy--sticky--in renormalization, and I wanted to put it on the table, set it out there for other possible connections (even if only my own, later on).  Barabasi goes on to explain two basic features of scale-free networks that complicate earlier theories that treat such structures as static and random.  To detail the importance of hubs, he names the characteristics of growth and preferential attachment.  And although his notion of growth seems teleologically biased, allowing for decay or detritus only briefly in a discussion of aging, I continue to be struck by the application of concepts from physics to social patterns.  At least for tonight (no telling how much credit to my head cold or tax fiasco), renormalization seems an especially interesting idea.

[Added: Renormalization is the same as order-disorder-order in story structures?]

Posted by dmueller at February 23, 2005 10:21 PM to Networkacy