computer simulation of kinship

Eric Silverman (ERICS@DEPAUW.EDU)
Sat, 14 Jan 1995 16:40:21 -0500

I have a (long) question concerning computer simulations/representations of
kinship charts. This is not a question concerning computer software
for genealogies, however. What I have in mind is something like this:

I work in a village of about 1000 in the Sepik River, Papua New Guinea.
Like most folks who work in such places, I have a lot of pages of genealogies
and kinship charts. However, like all or most of us, I have been taught to
represent this facet of social life using the conventional two-dimensional
kinship chart that we all know and love so well. In essence, this has
become for many of us a key model whereby we conceptualize and then
represent social life. Of course, it has its problems, one of which is
that we cannot properly represent the entire social matrix of the village
in a glance, because of the limitations of two-dimensions and the general
architecture of the standard diagram. Is anyone aware of a computer
program that will represent kinship charts in three-dimensions? Moreover,
if this program could in some sense rotate the three-dimensional model,
would our understanding of the relationships in the village alter? That
is to say, if we had a different type of kinship diagram, we might see
things--perhaps--that were not apparent in the conventional diagrams
that many of us now use. I sometimes wonder what a three-dimensional
depiction of the totality of a lineage or clan, even the entire village,
would look like. Would such a model more closely approximate the "image,"
loosely speaking, in the mind of someone in the village during a social
event or action? (I am well aware of the theoretical problems that can be
raised by this last sentence)

As I look at my book shelves that have
ethnographies on the Pacific, I doubt that no more than a handful lack
the conventional kinship diagram, in some capcity. But, as Bourdieu
discusses briefly in Outline, there are serious problems with these
static, two-dimensional representations. Although my query will not
resolve many of these problems--nor the thorny issue raised by Schneider--
such a program might nevertheless be useful or insightful. Computer
scientists deal with these sorts of diagrams and different ways of representing
diagrams all the time; some are called "fish eye views" and "graphs." I have
been discussing the matter with a colleague in Computer Science here,
but I would hate even to begin to reinvent the wheel. Is anybody aware of
or working on or interested in any such programs?

(Sorry to be so long-winded)

Eric Silverman
Dept. Soc/Anthro
DePauw University