Speakers of many languages prefer allocentric frames of reference (FoRs) when talking about small‐scale space, using words like “east” or “downhill.” Ethnographic work has suggested that this preference is also reflected in how such speakers gesture. Here, we investigate this possibility with a field experiment in Juchitán, Mexico. In Juchitán, a preferentially allocentric language (Isthmus Zapotec) coexists with a preferentially egocentric one (Spanish). Using a novel task, we elicited spontaneous co‐speech gestures about small‐scale motion events (e.g., toppling blocks) in Zapotec‐dominant (...) speakers and in balanced Zapotec‐Spanish bilinguals. Consistent with prior claims, speakers’ spontaneous gestures reliably reflected either an egocentric or allocentric FoR. The use of the egocentric FoR was predicted—not by speakers’ dominant language or the language they used in the task—but by mastery of words for “right” and “left,” as well as by properties of the event they were describing. Additionally, use of the egocentric FoR in gesture predicted its use in a separate nonlinguistic memory task, suggesting a cohesive cognitive style. Our results show that the use of spatial FoRs in gesture is pervasive, systematic, and shaped by several factors. Spatial gestures, like other forms of spatial conceptualization, are thus best understood within broader ecologies of communication and cognition. (shrink)

Pointing is a cornerstone of human communication, but does it take the same form in all cultures? Manual pointing with the index finger appears to be used universally, and it is often assumed to be universally preferred over other forms. Non-manual pointing with the head and face has also been widely attested, but it is usually considered of marginal significance, both empirically and theoretically. Here, we challenge this assumed marginality. Using a novel communication task, we investigated pointing preferences in the (...) Yupno of Papua New Guinea and in U.S. undergraduates. Speakers in both groups pointed at similar rates, but form preferences differed starkly: The Yupno participants used non-manual pointing numerically more often than manual pointing, whereas the U.S. participants stuck unwaveringly to index-finger pointing. The findings raise questions about why groups differ in their pointing preferences and, more broadly, about why humans communicate in the ways they do. (shrink)