Stirling's data is extensive. Even the information that just relates to migration is extensive. His attention to detail has provided us with a wealth of data and theory distilled from many years of experience. In `the wiring diagram' he has constructed a visual metonymy to try and convey certain diachronic relationships between conceptual and `real' classes of `objects' but he is still not happy with it. His model will never be complete so how can it tell us anything? He already knows it can, by extracting his causal chains he has illustrated how the diagram can be used as a hypothesis generator.
But are they true? How can we test his hypotheses independently?
Theory (a framework for proceeding)
If we consider that his mind has `abstracted' relationships from his lifetimes experiences. These are already a reduced set of all the possible views and interpretations of the `social reality' he was embedded in (Oxborrow, 1989).
The act of thinking itself therefore is a `reductionist act' yet from this `reduced set' of impressions he perceived even in 1953 that one social event had an influence on another. He was already thinking causally. Many of his predictions based on this thinking have come to pass. In his mind, whether consciously or not, we can say he had a `conceptual model' (Oxborrow, 1989) of what was happening in the world around him. He was mentally dynamically and diachronically simulating this model as he went through life.
The term simulation is like role playing (Chodorow,1991) or acting in the theatre (Johnson, 1981). It is a way of rehearsing the real before it happens38 (Schechner & Appel, 1990)
Fischer(1994) [13] is interested in formalising simulations to help understand specific anthropological problems.
In making the translation from the idea of simulation to the logical domain he says:
When we apply a rule or set of rules to a set of input information, we derive a set of results corresponding to the inputs. The process of deriving this set of instances of applying the rule can be called simulation. The rule is a model which describes relationships, the application of the rule to generate an outcome is a simulation.(Fischer,1994)
and in analogy to role playing he says:
Such simulation, predating computers, has been used in anthropology at least since the nineteenth century (Mulvaney 1970). (Fischer,1994)
He makes the point that:
Simulations are distinguished from other kinds of models more in terms of goal than form. Simulations are typically used for problems which are seen as complex and intractable, where no direct means of evaluation is known or the conventional means of evaluation is extremely difficult to execute, or which requires interactive decisions by the investigator during the course of the model. (Fischer, 1994)
Of particular value are the ideas behind systems theory[14] as developed from the work of, amongst others:
Norbert Wiener[15]; Heinz von Foerster[16] & Jay Forrester[17].
Forrester's most ambitious and contentious simulation is described in "World Dynamics" MIT Press 1973 and criticised by Brian Bloomfield in "Modeling the World", 1986
Relatively recent advances in computer technology have made complex simulations, which would have been hideously expensive or basically impossible, viable.
A useful web based resource for this kind of study is
The PRINCIPIA CYBERNETICA32 WEB[18]
Which is the web based arm of an international project33 based at the Free University of Brussels.
Simulation techniques have been used successfully in a variety of serious anthropological contexts for many years.(Fischer, 1980) (Fischer & Finkelstein, 1991) & (White, Colson & Scudder, 1987) (Lansing, 1991).
In order to test his model - to see if it `works', I propose to use a variety of computer based tools: relational databases; various types of word processor; spreadsheets for calculations and graphing of data; Statistics packages for more complex statistics analysis; The internet for access to on-line publications, general information and the Stirling homepage; a Java development environment for bespoke programming solutions including simulation implementation etc. I will conduct some standard demographic analyses of the census data with a view to using simulation techniques were appropriate.
In simulations we can try to `model' the objects and processes we are researching and thereby generate data based on our ideas (Fischer 1994)[19]. If all of our assumptions are correct about logic being an inherent quality of any `structured' entity i.e. our physical worlds and the constraints that this imposes on us, and we have modeled correctly based on the `evidence', then we can expect this data to conform to our indexes from the real world study. We are looking for correspondences between the sim data and our real data to tell us whether we are on the right track or not! The simulation gives us `consistent' control data. More likely is that the sim data will diverge from the real data showing us that our ideas are not working. Either way we learn something.
Some of the data Stirling has fits the criteria for a simulation experiment (Fischer & Stirling, 1991)[20] because we have and probably will always have incomplete knowledge about the nature of the variables underlying Stirling's point of conceptual agglomeration and the processes working on them. For the historical period in question it seems reasonable to attempt a cyclical agent oriented stochastic simulation of the migration process.This is effectively a Newtonian conformed, statistically driven individual events generating sequencer (a cogwheel engine with worn cogs), however, at this stage it seemed simpler (but not as interesting an approach) to use a cyclical aggregated deterministic simulation (a cogwheel engine with very shiny new cogs). (There seems to be no need to worry about systemic chaos at the level of macro abstraction these processes are probably working on so I will assume Newtonian physics apply).
1. I will first of all gather Stirling's data on migrations and analyse his database in order to construct indexes of change in the village which reflect migration patterns.
2. I will construct an events chronograph to show the sequence of possibly relevant events which could have effected village migration patterns.(Appendix C)
3. I will then formalise his model of migration, in preparation to implementation.
I will need to further deconstruct his conceptual boxes into the substance of the DB Data and use his 1950 census data to set the initial conditions. For the purposes of looking at migration I could try and simulate all the variables but some of them would not make much difference to the outcome. Therefore they must be prioritised. By starting off with just a few variables i.e. birth rates etc. we can test to see which variables make the most difference.
4. I will discuss the implementation of this formal model using the Java object oriented programming language.
5. If time allows I will try to see how this simulated data corresponds with the
real data. From my progress or from these results I will draw my conclusions