This paper began with an brief description of my experience in the complex group dynamics of a political protest in Washington D.C. I chose to open with that topic because the bulk of the discussion I have presented is fairly academic and abstract. I wanted to provide a real world example of why this sort of quasi-philosophical hair splitting is relevant to everyday life and real problems. I believe that a better understanding of communication and cultural processes could have important real world applications, especially as this is a period in which the modes of media, and the associated power implications, are somewhat in transition. I also feel that it is important to keep the pragmatic elements of a realistic setting salient as a reference point for the more academic discussion: how does one decide what course of action to take in a confused and possibly dangerous setting? My hunch is that we look around to see what others are doing, ask people who may be in the same situation, try to recollect what we have seen or heard of others doing in the past, or try our luck with constructing a response based on our knowledge of the world and how it works. This set of basic “common sense” knowledge is one of the crucial aspects of culture.
Arguments have been made that the ability to construct and pass on systems of interpretation, meaning and behavior is uniquely human. But there at least is some evidence that species other than humans have limited non-genetic means of transmitting behavior. The very limited selection of research presented here on communicative processes and culture in animals suggested that a better understanding of what is meant by “Culture” in humans in necessary before any conclusions can be drawn about its presence or absence in animals. Because many kinds of communicative behaviors have been observed in animals, it seems likely that the presence or absence of language may be a matter of degree. I do not necessarily mean to say that there are other species that use abstract symbol systems to communicate, just that it is difficult to see where a sharp line could be drawn between intentional and “mindless” use of symbolic signaling. Humans certainly appear to have the most extensive and well developed language/symbolic/culture system. Why this might be and how it may have occurred over evolutionary time is an open question. Are cultural processes adaptive, that is, do they confer fitness benefits on the individuals of species that employ them, or are they a side benefit from some other process that was selected for at some point in our evolutionary past?
It is easy to come up with reasonable situations where the presence of cultural transmission could affect genetic fitness, so it must be taken into consideration when testing evolutionary hypotheses. (Consider the possible genetic effects of the emergence of assortitive mating which is based on a cultural parameter.) Because the mechanics of cultural transmission may be quite divergent from genetic transmission, it is difficult to predict how they might interact, or which effects would dominate in a given evolutionary situation. It may be possible for traits to be adaptive in a cultural regime and not in a biological one. Certainly there is the potential for the emergence of structures which are more complex than might be predicted by biological evolution acting alone. The concept of a bio-culturally stable strategy may prove to be a useful tool for dealing with the intimidating proliferation of possibilities. I am also attracted to the idea of describing the evolution of culture-capable organisms as a co-evolution: traits are being selected for in multiple regimes, the species is balanced between them, but minor change in one direction may drive further evolution in another.
The biologically evolved and adapted activities of humans as communicators and sense-makers could potentially create adaptive landscape in which some patterns or messages, cultural or semantic, can increase in frequency at the expense of others, and possibly at the expense of traditional biological fitness. In other words, some of the formal properties of communicative processes in humans seem to suggest the possibility for differential propagation of cultural traits or information “messages.” If this is the case, the concept of diffusion or percolation along network ties may be a useful way of thinking about culture transmission. But what are the “messages,” “cultural items” or “information units” which I have so consistently avoided defining?
Two general frameworks for thinking about information have been brought forward in the course of this paper: Shannon’s concept of information as reduction of uncertainty, and a rough idea about information transfer increasing similarity of perceptual schemas or interpretive frameworks through some process of modification or induction of corresponding semantic elements in the interactants’ meaning spaces. Obviously a great deal more work is needed if these suggestions are to be refined from mumbo-jumbo to a usable definition. The factor analysis techniques of Osgood and his more recent colleagues (Pocklington, et al., 1997, Romney, et. al., 2000) seem like a promising avenue to pursue.
The research for this project has greatly increased my appreciation of the tensions and paradoxes inherent in the process of creating meaning and knowledge for oneself and trying to communicate it to others. Statements which conform entirely to expectations might as well be left unsaid, as they contain no information, yet uses or messages which are too bizarre and diverge too far from the norm will probably not be recognizable enough to be decoded. I now see meaning as a consensus process involving the individuals in communication; the use and connotations of each reference must be negotiated by the participants into something which is reasonably congruent between them, yet still consistent with their own interpretive frameworks. To me this leads to an increased sense of the relativity inherent in systems of symbols, language, and meaning.
I have found Campbell’s use of the concept of bias, and its development by Boyd and Richerson, to be quite widely applicable. Rather than thinking of effects as rigidly derministic functions (never a good idea for anything involving real people!), they can be thought of as an increased tendency to err in a particular direction or oversample in a certain way, with the effects perhaps only detectable as the deviation of a population mean from its expected value. There are, for example, some patterns of “human error” in communication which can be expected whenever humans open their mouths – even if individuals are working to tell the “truth” to the best of their knowledge, they are still more likely to omit what they do not understand, etc. Yet, remarkably, we manage to get along just fine most of the time, using contextual cues, previous knowledge, and multiple channels of information to assemble an appropriate meaning from even the sparsest of signals with surprisingly little error.
It seems likely that part of the reason for the relatively low rates of errors or specific misunderstandings in ordinary conversational communication is that we utilize a tremendously wide bandwidth of contextual cues to fill in for the details we miss. Communication and information transmission can involve material culture, proxemics, direct action, and omission, as well as the more explicit symbolic content of speech or text. It is important to consider the patterns of contact with material culture and its overt and implicit informational content when considering models of knowledge collection. I believe that a social networks perspective can provide an important framework for dealing with social structure, interaction, and environmental effects simultaneously. Some phenomena, like the small world problem, which initially seem counter-intuitive suddenly make sense when considered from this perspective. In a footnote to their paper, Robust Action and the Rise of the Medici, Padgett and Ansell (1993) present what is to me a compelling argument for sociological relations from a network perspective:
Our general position on the interrelation of social attributes and social networks can now be clarified. Obviously, (contra some occasionally overstated polemics by social network aficionados), we do not believe that social attributes are irrelevant: the particular way in which the Medici recombined social attributes through networks is the heart of the story here. What we object to is the arranging of attributes discretely in groups or spatially as grids – a procedure which presumes that attributes are behaviorally meaningful in a network vacuum. Of course actors in the system, as well as researchers, do exactly these clustering procedures mentally when they analyses their own social structure; this is what “boundedly rational” cognitive classifications are all about. But there is a widely underappricated gap between these macro cognitive (or “cultural”) operations and micro behavioral local action taken by concrete individuals in very particular, heterogeneous, and often cross-pressured circumstances. Simplifying social reality into homogenous subsets “with common interests” rips individuals out of their (often contradictory) multiple network contexts and obscures the very heterogeneity and complexity of which organizations like the Medici party are constructed! (Padgett and Ansell 1993, p 1285)
Not that network data always simplify the situation. The problem with working with complex data sets is that the analysis becomes more complicated as well. And, as network analysis is still sort of an infant science (although the concepts have been around for at least 50 years), there are few agreed upon measures and indices for making meaningful comparisons across studies or networks. In many cases the analysis takes a very qualitative mode, using formal criteria to lay out a visualization, and then relying on inspection and intuition for judgments about meaning and further analysis. But I find the models and tools which are being developed from a networks perspective to be both exiting and promising. The notion that contact networks imply transmission networks and so might provide a window into cultural processes is tantalizing, and it is at least not disproved by the data I collected and analyzed.
In this paper I have generally used “communication” to mean an attempt to convey information or the transfer of information between two interacting individuals. But my usage has often been quite sloppy. Part of this is due to the argument I’ve been trying to make: that there may not be any clear-cut distinction between interaction, information exchange, and communication – they are all part of a graduated process and are to some degree implicit in each other. I’ve also tried to suggest that communication and culture ought not to be considered separately either; they both involve flows of information, just on different time scales. Although I certainly have not presented a framework comprehensive enough to deal simultaneously with the scales of cultural evolution and gossip, I believe that I have pointed out some similarities which suggest that such a synthesis might eventually be possible.
I think one of the key images which fascinates me, and which I hope to have conveyed effectively, is the idea that even if humans were viewed as nothing more than a collection of arbitrary sense-makers – sucking up information from their social and physical environment and re-emitting it in some fashion – the simple fact of their immense and unpredictable connectivity could lead to the emergence of complex and interesting structures. The individuals in a society could be viewed as enzyme-like “black boxes”, spinning out endless variation and catalyzing the syntheses of socially transmitted cultural information, extracting from it patterns and regularities while simultaneously permuting it into innovative, surprising, and often beautiful forms.