In this volume I am presenting, explicating, and trying to move beyond Gould’s system for analyzing kinterm systems. In this preface I want to briefly discuss what attracted me in Gould’s work. Then I want, again briefly, to describe the background that I bring to this task.
Sydney H. Gould was a mathematician, and seems to have lived in a world of deductive proof. What initially impressed me was how clean, logical, and strongly predictive his system was – how much better it was than any of the competition (where I already had excellent analytic control over several other analytic approaches). Having worked a lot with his system since then I have that reaction even more strongly.
Since I have been working with (and within) his system in the present work, I simply, without really thinking about it, bought into his game. The game was to rigorously deduce solutions, and then see what in the empirical world these solutions fit. In developing the game itself, he went through a process of adapting the nature of its entities to improve the power of the deduced solutions. I do not know for sure, but I suspect the process by which over many years he evolved his system also involved some investigation of apparent non-fits, and thus of the process by which the target of analysis was refined (a process captured in Piaget’s subdivision of “adaptation” into “assimilation” [changing elements of the environment into a form in which the organism can incorporate them] and “accommodation” [in which the organism adjusts itself to enable its incorporation of new aliments]. For example, see his treatment of the Fanti situation in which FaSi is called “mother” but does not behave systemically as do Mo, MoSi, and other normally extended members of the “mother” category. So he removed FaSi from the system’s treatment of “mother”, and marked it off as a “concurrent” use of the “mother” term – that is, as a homonym. Since I was the ethnographer whose Fanti description Gould was relying on, I knew the system well, and, for my own reasons, had already recognized the anomaly represented by FaSi – even as I came up with a slightly clever distinctive feature approach for including it. My Fanti solution was emically ad hoc, and seemed to have no general application to other Cheyenne-type or Crow-type systems.
In sum, he treated terminological systems as systems, and aimed at designing an analytic approach (an analytic system) that accounted for (deduced from a few axiomatic first principles) these terminological systems. In doing so, he marked off parts of these terminologies that were systematic (in the sense of being regular) but not part of the system – such as the “brother”/“sister” distinction in English, which, while regular and sometimes semantically and socially important, did not affect the systemic relations (e.g., a “brother”’s son is equivalent to a “sister”’s son, and today both have become sometimes subsumed in a “sibling” category. Thus, he looked at what his analytic approach accounted for and what it did not – while trying to make his approach as widely applicable as possible and while having clear and consistent ways of demarcating what was grist for the approach from what was not.
Ptolemy and Copernicus both came up with formal systems that provided alternative descriptions of a particular natural system – our solar system. The decision between the two ultimately came down to which was better at encompassing new findings, which was simpler and more elegant, and which shed more light on the dynamics that drove the solar system. Like them (if you do not mind my reaching a bit) Gould was trying to describe a set of naturally occurring systems – kinship terminologies. At the time there was no real competition, though Lounsbury and Goodenough came closest and Kris Lehman shared some of the relevant insights. Subsequently Read (as well as I and maybe Keen) came up with emic systematic native language approaches, but these approaches were not so directly comparable across systems or for recognizing the traditional TYPEs. The question about Gould’s system then concerned whether it represented Copernican insight or Ptolemaic distraction. Switching metaphors, had Gould found some of nature’s joints?
As, in the intervening years, I worked with Gould’s system and applied it to terminologies I knew about, I became increasingly certain that he was on the Copernican side.
∵
In summary, my recognition of the potentialities offered by Gould’s analytic system evolved out of my background in linguistic and then cognitive anthropology, with some interest in computer science and early artificial intelligence (which morphed into modern cognitive sciences). The empirical loci of my work were kinship terminological systems (see Kronenfeld 2009), the semantics of everyday language (see Kronenfeld 1996), and the pragmatic system of culture (see Kronenfeld 2008, 2017).
In more detail: Coming off of an initial interest in physics and math, and some continuing interest in computer programming, my anthropology education in the 1960s involved structural and transformational linguistics and emphasized ethnoscience (which evolved into modern cognitive anthropology) and social organization. I liked puzzles and, early on, found that the analysis of kinship terminologies appealed to my puzzle-solving bent. My doctoral committee in the late 1960s (A. K. Romney, Joseph H. Greenberg, Roy G. D’Andrade, and Charles O. Frake), as well as O. Brent Berlin and Paul Kay, nicely encompassed my interests and provided the sources on which I subsequently built.
Linguistics was always a strong secondary interest. I attended the Linguistic Society of America’s Summer Institute in 1964 at Indiana University. For many years I chaired UCR’s Interdepartmental Program in Linguistics.
Kin terminologies, as a locus of research, are interesting and important in their joining of language with culture – along with the pragmatic systems that link the two. Language in this context particularly includes both semantic systems and the semantics of words – including the contrast relations and patterns of denotative and connotative extension built out of words, and the wider systems built out of combinations of words. Culture here includes the terms that embody focal social relations (including the obligations and privileges that these relations confer) and that provide the framework for much of social organization. Kinship terminologies also link to biology – via the important role that birth (including her links with her parents) plays in a child’s initial ascribed status and thus her placement in the wider social network. It is not surprising then that kinship terms provide a rich source of figurative extension across a great many cultures.
Given my history, it is not surprising that my first published contribution to kinship studies was a description of a computer program I wrote that analyzed a large class of kin terminologies – and that was published in Language, the journal of the Linguistic Society of America (Kronenfeld 1976).
Cognitive anthropology developed out of ethnoscience studies of specialized cultural vocabularies – especially the terminologies of kinship and of ethnobiology, but also in studies of the vocabularies associated with varieties of firewood, with the making and socially mediated consumption of food and drink, with traditional fishing techniques, and so forth. These systems of vocabulary were found to link to and reference pragmatic cultural knowledge systems. Cultural knowledge systems were found to be collectively held by members of relevant groups, differentially distributed across that membership, and totally or completely held by no one. Thus cognitive anthropologists backed into Durkheimian systems of collective representations – and then offered a new way of looking at and thinking about Durkheim’s concepts.
As I published my research, I was happy to find the overlap of my research interests with those of cognitive linguists – initially through my participation in two LAUD Symposia and then through my subsequent reading. Important to me here were the research concerns of Professor Rene Dirven and people associated with him – which for me especially included the work of Gary Palmer, Gitte Kristiansen, Farzad Sharifian, Dirk Geeraerts, Martin Pütz, and Peter Harder. It was by means of this reading that I particularly became aware of Sharifian’s and my shared interest in collective systems of differentially shared distributed cognition – which sharing led to my joining the editorial board of the International Journal of Language and Culture (IJoLC) in 2013, which he then edited.