Journal
of Memetics -
The Call for Papers to this the first special issue of the JOM was broad in scope. We sought papers positioned within computational memetics, loosely conceived. All we asked of researchers was that their work proposed some sort of formal model and that computation had a substantial role in their study (above and beyond that of simply word-processing the final submission).
From the CFP we were pleased to receive eight submissions that we sent out for review; here we publish the three accepted papers. Our thanks to all of the authors and to the referees.
The project of memetics is generally conceived as a process of observing animal social activities and attempting to infer rules from these behaviors. Of course the particular memetic stance means that the observation and inferencing is informed by a replicator-minded, evolutionary-theoretic, and human-centered position. (Though none of those positions are uniformly agreed to within the community.) These particularities notwithstanding, memetics still owes much to anthropology, ethology, and other research communities involved in observing social behavior and inferring from it certain rules.
In computational memetics, however, the project is put on its head. Here, we encode a set of rules into a model, simulation, or system of analysis and then proceed from this encoding to a description of the results. Instead of description to rules we go from rules to description.
Why bother with this head-spinning inversion? Because this type of research allows us to move beyond historical contingency, to gain new and fundamental insights, and to ground our ideas in solid yet falsifiable foundations. Let me impose a quick taxonomy onto computational memetics - simulation, computational theory, population memetics - and use it to motivate these claims with some examples from the literature.
Examples of this work include Bura (1994), Gabora (1995), and Hales (1998). My co-editor for this special issue, Bruce Edmonds (1998), has compared and contrasted various simulation models within memetics concentrating on some of my work (Best, 1997) and the more abstract models of William Calvin (1997).
Why simulate? The world we find around us, including the memetic world, is informed fundamentally by historical accident and contingency (Gould, 1989). But a computer simulation let's us replay (over and over if we like) the same bit of history - tweaking parameters with each run. From these replays, we can distill from the simulation what properties are universal, what are clearly useful, and what are pure happenstance (Langton, 1995).
Memetic simulations have been used to ask quite fundamental questions, such as: What is the value to an agent in a community of having any memetic (or social) system to begin with? Is social activity mostly an accident? How has the presence of memetic systems altered the course of genetic systems?
For instance, I have explored through simulation the relationship between genetic evolution, individual learning, and memes (Best, 1999). This simulation work suggests that memetic processes can allow genetic evolution to solve problems that genes alone would not solve. By continuously "rewinding the tape" of geme/meme co-evolution these simulations suggest that memes can really matter to genetic evolution.
A recent and related paper has explored the role of the relative speeds between genic and memetic evolution and the prospects for one replicator to generate selective pressures against the other (e.g. memes against genes) (Bull, Holland & Blackmore, 2001). This work demonstrates the value addition of simulations vis-à-vis their ability to "rewind the tape". Blackmore argued earlier (1999) that human language and brain encephalization is due to competition between memetic and genetic evolution. In the real world there is no way to retry history under differing patterns of competition to test this hypothesis. Under simulation, she and her co-authors have replayed history again and again while varying these parameters.
A computational theory is set off from a simulation because, simply put, a simulation need not ever have been run. Indeed, a computational theory can be thought of as that subset of general mathematical models that are algorithmic (step by step) in nature. But like the more general set of mathematical models, formal analysis is often applied to a computational theory in order to secure some results and deepen understanding.
An important example due to Holland (1992) illustrates the value of computational theories to reveal fundamental elements within a science of evolution. Holland developed an algorithmic theory of artificial evolution and employed a set of analytic techniques (relying on the multi-armed bandit statistical argument) to prove that his evolutionary algorithm optimally trades-off the two goals of search within a non-linear or epistatic landscape: the exploration of new potential answers and the exploitation of existing known solutions. This is a stunning result as it suggests that nature (qua memetic or organic evolution) optimally searches its problem space!
While we do not, to my knowledge, have as strong a result specific to memetics, this line of computational memetic inquiry is quite valuable and should pay off handsomely in tomorrow's research outputs.
Population memetics is the study of variation within populations amongst replicating memes. This diversity is the very stuff of the evolutionary process - and an important method to observe the memetic process in situ. I am offering it as a form of computational memetics since, in practice, it often is predicated on substantial computer analysis (in contrast, indeed, to a computational theory which might require little or no computer time).
My own computational study of memetics within text discourses offers an example of this line of research. I have used a formal model of analysis, and a large collection of software tools, to study the population diversity and dynamics of replicating elements of text within corpora over time (e.g. Best, 1998, in press; Pocklington & Best, 1998). I have argued (at times with my co-author Richard Pocklington) that it is possible to distill units of selection within replicating texts, monitor their time-valued dynamics, discover the accumulation of usage amongst variant forms, and even make attempts to reverse-engineer their adaptive value.
While this work offers only a modest first step towards an elaborated population memetics, it suggests way that we can apply computational methods to existing bodies of data and reveal the memetic process.
The papers of this special issue provide additional and important examples of work in this space of inquiry. Baldassarre uses simulation to further investigate the relationship of individual and cultural learning to organic evolution. In particular, he studies the value of transmitting evaluative criteria (rules on whether to select and retain various behaviors) versus the transmission of behaviors themselves and argues that transmission of behavior is more effective than transmission of evaluative criteria.
Derek Gatherer uses simulation to study the effects of cultural taboos on rates of human genetic homosexuality. Gatherer shows here how computational methods can be used to support or refute existing qualitative theories. In this case, Gatherer suggests that his simulation casts doubt on Lynch's (1999) previously published musings on the relationship of memetic taboos and genetic homosexuality.
Finally, Robert Reynolds, Robert Whallon, and Steven Goodhall employ a simulation developed within the Swarm environment to explore the role of consensus decision making and emulative learning in primate social groups. In particular, they find that primate groups are able to advantage social learning when those resources which are pooled collectively amongst the social group are distributed using a method which rewards the individual agent responsible for group success. In the absence of such distribution sensitivity the group performance is equivalent to a simulation where social learning is replaced with a random walk algorithm.
Best, M.L. (1998). An ecology of text: Using text retrieval to study alife on the net. Journal of Artificial Life, 3(4): 261-287.
Best, M.L. (1999). How culture can guide evolution: An inquiry into gene/meme enhancement. Journal of Adaptive Behavior, 7(3/4), 289-306.
Best, M.L. (in press). Adaptive value within natural language discourse. To appear, Evolution of Communication Journal.
Blackmore, S. (1999). The meme machine. Oxford: Oxford University Press.
Bull, L., Holland, W., & Blackmore, S. (2001). On meme-gene coevolution. Artificial Life, 6(3), 227-235.
Bura, S. (1994). MINIMEME: Of life and death in the Noosphere. In D. Cliff, P. Husbands, J-A Meyer & S.W. Wilson (Eds.), From Animals to Animats 3: Proceedings of the Third International Conference on Simulation of Adaptive Behavior (pp 479-486). Cambridge, MA: MIT Press.
Calvin, W.H. (1997). The six essentials? Minimal requirements for the Darwinian bootstrapping of quality. Journal of Memetics [On-line serial] 1(1). Available: http://cfpm.org/jom-emit/1997/vol1/calvin_wh.html.
Campbell, D.T. (1960). Blind variation and selective retention in creative thought as in other knowledge processes. Psychological Review, 67, 380-400.
Edmonds, B. (1998). On modeling in memetics. Journal of Memetics [On-line serial] 2(2). Available: http://cfpm.org/jom-emit/1998/vol2/edmonds_b.html.
Eigen, M. (1992). Steps towards life: A perspective on evolution. Oxford, UK: Oxford University Press.
Gabora, L.M. (1995). Meme and variations: A computational model of cultural evolution. In L. Nadel & D.L. Stein (Eds.), 1993 Lectures in complex systems. Reading, MA: Addison-Wesley.
Gould, S.J. (1989). Wonderful life : the Burgess Shale and the nature of history. New York: W.W. Norton.
Hales, D. (1998). Selfish memes & selfless agents - Altruism in the swap shop. In Proceedings of the 3rd International Conference on Multi-Agent Systems. Los Gatos, CA: IEEE Press.
Holland, J.H. (1992). Adaptation in natural and artificial systems: An introductory analysis with applications to biology, control and artificial intelligence (Rev. ed.). Cambridge, MA: MIT Press.
Langton, C.G. (1995). Editor's introduction. In C.G. Langton (Ed.), Artificial Life: An Introduction (pp. ix - xi). Cambridge, MA: MIT Press.
Lewontin, R.C. (1970). The units of selection. Annual Review of Ecological Systems, 1, 1-18.
Lynch, A. (1999). Thought contagion: How belief spreads through society. A response to Paul Marsden. Journal of Artificial Societies and Social Simulation [Online serial] 2(3). Available: http://jasss.soc.surrey.ac.uk/JASSS/2/3/lynch.html.
Marr, D. (1982). Vision: A computational investigation into human representation and processing of visual information. San Francisco: Freeman.
Pinker, S. (1997). How the mind works. New York: W.W. Norton & Company.
Pocklington, R., & Best, M.L. (1997). Cultural evolution and units of selection in replicating text. Journal of Theoretical Biology, 188, 79-87.
Richards, W. (Ed.) (1988). Natural computation.
Cambridge, MA: MIT Press.