We examine the extent to which practices of language use may be diffused through language contact and areally shared, using data on spatial reference frame use by speakers of eight indigenous languages from in and around the Mesoamerican linguistic area and three varieties of Spanish. Regression models show that the frequency of L2-Spanish use by speakers of the indigenous languages predicts the use of relative reference frames in the L1 even when literacy and education levels are accounted for. A significant difference in frame use between the Mesoamerican and non-Mesoamerican indigenous languages further supports the contact diffusion analysis.

This article presents an overview of the goals and methods of semantic typology, the study of the distribution of semantic categories across languages. Results from this field inform theoretical accounts of syntax-semantics interface phenomena, as well as the nature of the relationship between language and cognition. This article discusses a variety of quantitative methods that represent recent efforts in semantic typology to (i) discover patterns in the distribution of independent variables and (ii) predict the distribution of dependent variables in relation to identified independent variables. Such methods include Multi-Dimensional Scaling, Hierarchical Cluster Analysis, and Generalized Linear Mixed Effects regression analyses. We identify and discuss notable published examples of these methods used in semantic typology.

This paper presents a cross-linguistic investigation of a constraint on the use on intrinsic frames of reference proposed by Levelt (1984, 1996). This proposed constraint claims that use of intrinsic frames when the ground object is in non-canonical position is blocked due to conflict with gravitational-based reference frames. Regression models of the data from Arabic, K’iche’, Spanish, Yucatec, and Zapotec suggest that this constraint is valid across languages. However, the strength at which the constraint operates is predicted by the frequency of canonical intrinsic frames in the particular language. The ratio of the incidence of intrinsic usage with canonical vs. non-canonical orientation appears to be remarkably uniform across languages, which suggests the possibility of a strong cognitive universal.

Numerous studies have found evidence of a speech community’s referential practices in discourse being predictive of its members’ behavior in nonverbal tasks. In this article, we discuss a series of exceptions to this alignment pattern, drawing on data from eleven populations of Asia, Europe, the Middle East, North and Central America, and Oceania. These exceptions have not been discussed in conjunction with one another and the striking commonalities across the findings of these studies have gone unnoticed: (a) in discourses referring to small-scale space, either intrinsic frame use is dominant or both relative and geocentric frames are used frequently in addition to intrinsic frames; and (b) in recall/recognition memory, geocentric coding is more common than egocentric coding (in tasks that involve stationary stimulus configurations) in nine of the populations, while in the remaining two, there is evidence of extensive intrinsic coding even in nonverbal cognition. We discuss these findings in light of Haun’s innate geocentrism hypothesis (Haun, D. B. M., C. Rapold, J. Call, G. Janzen & S. C. Levinson. 2006. Cognitive cladistics and cultural override in hominid spatial cognition. Proceedings of the National Academy of Sciences 103(46). 17568–17573). Our data offers partial support for this hypothesis, but simultaneously calls into question whether any extrinsic reference frames are available innately.