Date: Fri, 17 Jul 1998 10:09:00 -0700 (PDT)
From: Narda Alcantara
To: Doug White
Subject: fall seminar
Review of Social Networks
i have, so far, detected four main trends (I made it five, breaking the fourth into two parts -- DW) in the present social networks
paradigm. there might be even more but, in my opinion, newcomers should be
introduced to, at least, these four trends that are not necessarily
exclusive but that should be considered as part of the historical
process of any discipline that tends to grow and, eventually, split in
subfields:
1. the manchester people and their egocentred networks, looking for the
non formal, hidden, personal networks, and interested in conflict and
negotiation. Mitchell, Barnes, Bott, et al, and Wellman, Fischer, Danching
Ruan, et al, on the other side of the atlantic.
2. the cognitive/sociometrists based in the Gestalt tradition, via Kohler,
Lewin and his "field theory" (i don't recall just know other names),
looking for "homophily" (Blau's thing), and the reasons why people like
some people but not others, etc.
3. the cognitive/cliquish people looking for group and subgroup formation
(cliques, clans, clubs, and the like), in the Jacob Moreno tradition
(though i dont think Moreno could be considered networks...), like Warner
and Mayo. examples: the Davis & Gardner deep south data, the Sampson data,
etc. lots of networks people are working on this stuff.
4. the kinship network analysis, including Pattison's kinship algebra,
Boyd's kinship stuff (a bit esoteric), and what John Scott calls the
"great breakthrough" in network analysis, Lorraine and Harrison White's
thing on structural equivalence and the like.
5. coming out of the latter, P-graph analysis and large scale network
analysis by D White and his uci group plus the Pajek people
(Batagelj and Mrvar) and interested historical poliscientists (Padgett),
historical sociologists (Bearman and others) and social demographers
(James Lee, etc.).
if i were to plan an introductory course of network analysis, i would
start by telling people about these four or five trends, so that everyone could
choose the one that interests them, but most of all, to make them aware of
the fact that network analysis is becoming more and more complex, and that
different trends respond to different objectives and research traditions.
in other words, we should discard the idea that networks is a whole,
single, homogeneous paradigm, with a single theoretical stance of its own.
of course, the option will be always open to people who want to work on
the four trends, but that is something feasible only for a few
ni modo: "lo que dios no da, Salamanca no lo presta".