On the heels of Kemp & Regier’s article “Kinship Categories Across Languages Reflect General Communicative Principles” published in Science, Stephen Levinson calls for a revival of kinship studies. Dan Sperber echoes with a post entitled “Is Kinship Back?” Levinson’s specialty is language and cognition, his piece was published in the Psychology section of Science. It fascinates me how versatile “kinship” is. People with different academic backgrounds end up talking with each other about the same thing – kinship. That’s why it has been possible to build a comprehensive bibliography of kinship studies across a wide range of scientific disciplines – anthropology, linguistics, psychology, economics, history, biology, etc. All these studies turn out to be interrelated either in the kinds of underlying questions being tackled or, more directly, in the overlapping references.
Levinson’s plea is of course very welcome. One thing to be mindful of is the danger of re-inventing the wheel under the pretext of reviving something. Kinship terminologies have always been the mice of the social sciences in a sense that a myriad of types of formal approaches were either designed for or tested on kinship terms. I’m talking about componential analysis, scalogram method, equivalence-rule analysis, graph theory, set theory, optimality theory, etc. Kemp & Regier (2012) add another layer to this plethora of formalizations. But it remains unclear what problems all of these approaches are trying to solve and whether the questions have already been answered or should not be asked in the first place. One of the problems with “traditional” kinship studies was their insularity. The formal approaches tend to reaffirm it because it is impossible to concisely model something that has empirical connections to naming, pronoun use, reported speech and a host of other phenomena. But it is precisely the multi-faceted nature of kinship terms (not their boundedness as a lexico-semantic class) that needs to be accounted for.
On another note, Levinson suggests the application to kinship studies of “computational techniques of biological phylogenetics to extract the historical development of patterning in cultural categories.” “Traditional” kinship studies out of anthropology has been doing this for the past 150 years since Lewis H. Morgan’s “invention of kinship” and my book “The Genius of Kinship” is the 2007 state-of-the-art in this subfield. Biological phylogenetics may be a more distant cousin to these approaches, while the comparative methodology in historical linguistics is a close sibling. And it’s precisely in the extent to which students of kinship can cross-pollinate with historical linguists that the visible progress in the dynamics of language-bound categories can be made. The application of Bayesian method derived from biological phylogenetics to kinship terminological evolution, as exemplified by Fiona Jordan’s 2011 paper “A Phylogenetic Analysis of the Evolution of Austronesian Sibling Terminologies” referenced by Levinson as showing “patterns of irreversible evolution,” can be misguided because the essential units of analysis, which require knowledge of both anthropology and linguistics, are not coded properly. (I communicated the problem to Fiona but apparently it was too late to fix it, hence the paper came out in its original form.) This will correspondingly result in the revival of “conjectural history” decried by Radcliffe-Brown, not of kinship studies.
Levinson is trying to enter kinship studies but, from the very onset, he does it with a wrong set of assumptions about kinship studies. These questionable assumptions can be dispelled by a thorough historiographic dive or they can be solidified by a narrow focus on the some of the most recent works that just scratch the surface of the field.