Andy Blunden. June 2010

Hutchins on Trobriand reasoning

Response to Reasoning in Trobriand Discourse
See also “Culture and Inference. A Trobriand Case Study,” by Edwin Hutchins, Harvard University Press, 1980.

Hutchins makes the (to me) quite unsurprising claim that Trobriand Islanders routinely make inferences consistent with the use of syllogistic reason. The paper is important nonetheless because it refutes well-known claims to the contrary and is backed up with solid data. He sums up the relation between culture and reasoning in the conclusion to the book, as follows:

“The clear difference between cultures with respect to reasoning is in the representation of the world which is thought about rather than in the processes employed in doing the thinking. It is clear that Trobrianders cut the world into a different set of categories from those we entertain, and that those categories are linked together in unfamiliar structures. But the same types of logical relations underlie the connections of propositions in our conceptions and theirs, and the inferences that are apparent in their reasoning appear to be the same as the inferences we make.” (p. 128)

Hutchins makes important points about how the errors of previous investigators were possible. He also points to the difference between the Islanders often failing to understand the researcher’s project and the Islanders’ correct solution of projects defined within their own culture. This includes pointing to the difference between being able to use (what the researchers call) syllogistic reasoning, and conscious awareness [Russian: osoznanie] of syllogistic reasoning and being able to objectify forms of syllogistic reason and recognize them within projects foreign to their own culture, such as in psychological testing.

When we put these issues in the context of understanding Vygotsky’s ideas on concepts there are multiple layers of possible misunderstanding. Let’s leave Vygotsky and his comrades’’ misunderstandings about so-called “primitive” people and their ill-fated psychological testing in Uzbekistan out of the picture though, and confine ourselves to Vygotsky on concepts and Hutchins on syllogistic reasoning.

Firstly, it is only in the last pages of Hutchins paper that he touches on what we would call the question of concepts. Hutchins seems to take as the prototypical concept the mental image of some finite thing in the world, though he uses the word concept in passing to describe more developed thought-forms. In the litigation scenario Hutchins takes as his example, his attention is focussed on truth-value inferences between propositions, in other words he is concerned entirely with acts which could have been made by a computer, operations of the kind dealt with by set-theory and symbolic logic which in Vygotsky’s terms are relevant to the use of pseudoconcepts[1], not true concepts. The neat block diagrams he uses to illustrate the claim make the automaticity of the process explicit, and he seems to be of the view that to understand a mental process is to be able to model it with a computer program. The mind is a cluster of nested subroutines which turn atomistic perceptions into compound thoughts via schemata somehow residing the brain.

So if we look only at the points in the example that Hutchins asks us to consider, in relation to the question as to whether Trobrianders use “true” concepts, all that is proved is that the Trobrianders use pseudoconcepts effectively.

But in order to explain the setting, Hutchins tells us about a well-established system of normative relations and practices regulating property-ownership and the creation and transmission of words which are without a shadow of doubt carriers of true concepts. But Hutchins doesn’t even notice this; for him the concepts are just background information, the material on which reasoning works. I think this is because he himself, as a cognitivist, mistakes a pseudoconcept for a true concept, as does the entire philosophical tradition of Logical Positivism and its kin. Probably, Hutchins is just not very much concerned with what a concept is. They are just a kind of “abbreviation device” (p. 115).

Tupwa, Kasesila and Pokala[2] and so on are true concepts, created and nurtured in the institution of Trobriand law and transmitted through participation in its court system. Hutchins doesn’t even notice this or think it is worth mentioning. He treats these words as indicating relations that have various attributes, like pseudoconcepts. The institutionalised system of practices in which these concepts are realised he does not think worthy of discussion in the paper (though it is all explained in detail in the book), except as background information or context for his example. In the book, he outlines how ‘discourse about land can be explicitly represented’ as follows:

“In a propositional representation, semantic information is encoded as a network of concepts interconnected by relations. Concepts are specific instances in the domain of discourse. In this domain of discourse, for example, a particular person, say John, or a particular garden plot, say the garden named Riverbank, are represented as simple concepts. The relation ‘holds rights in’ can link concepts together to form the proposition ‘John holds rights in Riverbank’. The relation ‘holds rights in’ obviously obtains between many people and many units of land. In fact, this relation always links one instance from the range of concepts that are economically appropriable units of land. When a relation is stated in terms of such variable ranges as these, it is a schema. When the ranges are replaced by concepts (a process called instantiation because it is the assignment of specific instances to the relation) the schema becomes a proposition. A schema is then a form or a template from which an arbitrarily large number of propositions can be constructed. ... Basic concepts like ‘John’ either correspond to something in the world or they do not, but they are never true, false, likely or unlikely.”

So the question of whether Trobriand Islanders use syllogistic reasoning, the answer is in the affirmative, but is absolutely nothing to write home about. Meanwhile the more interesting question as to whether Trobriand Islander use true concepts, something which cannot be claimed for children in industrialised cultures, is not broached but established without noticing. The question of the context of reasoning, the problem over which previous investigators had stumbled, is resolved by observing reasoning in an indigenous context but this still leaves the problem of cross-cultural cognitive science unclarified.

1. A pseudoconcept is a thought-form identified by Vygotsky, which is basically commensurate with a set in Set Theory; that is, an object is deemed to belong to a set (i.e., is subsumed under the pseudoconcept) if it has/does not have attributes as defined for the set (“birds are animals with feathers which fly,” and so on) but which element-by-element correspond with those objects subsumed under a true concept, such that an adult does not notice that the child with whom they communicate is using different thought-forms to identify the same entities.

2. Categories of property law in the Trobriand culture. 40 or 50 of such concepts are outlined in the book, each affect-laden concepts through which relations between people, land and each other are understood in the context of Trobriand history and culture.