Space in semantic typology: The MesoSpace project

1. Semantic typology – Semantic typology is the crosslinguistic study of semantic categorization – the search for universals and dimensions of variation in the meanings expressed across languages. For example, suppose you see a bee flying into a house. A native speaker of Yucatec Maya might describe this event as follows: (1) Táantik uy=óok-ol hun-túul kàab ich le=nah=o’ RECENT it-enter-ASPECT one-CLASSIFIER.ANIMATE honey in the=house=DEIXIS ‘A bee just entered the house’ There is no reference to flying in (1). This could be added by expanding the sentence to include a second verb; but doing so would not be particularly idiomatic in Yucatec unless the speaker wanted to stress the fact that the bee flew instead of crawling into the house (Bohnemeyer & Stolz 2006). And the verb glossed ‘enter’ in (1) more literally means ‘become inside’; (1) would also be true as a description of an event in which somebody placed a toy house over a motionless bee (Bohnemeyer in press). There is no equivalent of the English past tense in (1) (Bohnemeyer 2002). There is an expression of temporal immediacy; but this is compatible with any reference point in the past, present, or future so (1) also translates ‘A bee had just entered the house’ and ‘A bee will have just entered the house’. The bee is referred to using the word for honey in combination with the numeral “classifier” túul indicating that that which is counted is animate. Replacing túul with the generic classifier for inanimate entities would yield the translation of ‘A beehive just entered the house’ (Lucy 1992). The preposition ich ‘in’ selects the inside of the house as a point of reference, but does not indicate whether this place marks the endpoint of a motion event, as in this case, or the beginning (“A bee flew out of the house”), some point in between (“A bee flew through the house”), or the location of an entity (“There is a bee in the house”) all of these meanings can be expressed with the same preposition (Bohnemeyer 2007). It appears, then, that the resources languages use to express the same idea can vary considerably. Even in contemporary linguistics there is widespread disagreement over the extent of this variation. Some view linguistic categorization as a mapping of a largely universal conceptual space into grammars and lexicons which vary only superficially across languages (e.g., Pinker 1994; Li & Gleitman 2002). Others assert that there is no crosslinguistic uniformity in linguistic categorization except perhaps at the most abstract levels of analysis (e.g., Levinson 2003; Evans & Levinson in press). The discrepancy between these positions is the result of sparseness of empirical evidence combined with the ideological struggle between universalists and relativists. Relativism is the idea that cognitive representations are to a significant extent culture-specific, learned, and social rather than individual. Conversely, universalism assumes that cognitive representations or at least core components of them – are culture-independent and possibly innate. Thus the relativism-universalism debate is one contemporary manifestation of the age-old nature-nurture debate. Along with cognitive psychology and the study of linguistic and cognitive development, semantic typology opens one of the few empirical windows onto the relativismuniversalism debate. Languages are engines of symbolic representations. These allow people to “share” their cognitive representations – inferences, beliefs, emotions, desires, etc. in ways that cognitive scientists have barely begun to understand. This amazing feat of sharing mental worlds in interaction makes human cognition far more social and cultural than any known form of animal cognition. A large portion of what we “know” or “assume” we do so neither on the basis of personal experience nor instinctively, because of something coded in our genes, but because this procedural knowledge is part of “our culture”. This