Introduction

The question I want to raise, ‘Where to orient oneself’, addresses two issues. First, it asks for the type of deictic field within which spatial information is transmitted. This issue is motivated by my attempts to understand what Belhare people mean when they tell you, for instance, to move something toba ‘up’; for it is by no means obvious where toba is. Secondly, the question looks for the domain in which spatial information is encoded and for the relation of this domain to grammar, semantics, and cognition. More specifically, I inquire into the effects that spatial deixis has on the grammar of Belhare, on the quality of different ‘senses’ in deixis (is there linguistically resolvable polysemy? or mere contextual vagueness?), and on the relation of linguistic deixis to other cognitive modalities that are basic to spatial orientation and manifest in cultural patterns and social behaviour. I address the cross-modal questions from a linguistic point of view, seeking for structural parallels in non-linguistic cognition.

The language and people I am concerned with are called Belhare (Nep. Belhālre or Belhārīya) or Belhare Rai, the term Rai (Nep. Rāī) being the collective ethnonym for a subgroup of the Kiranti (Nep. Kirãtī) people in Eastern Nepal (cf. Vikal & Rāī 2051, Bickel 1996). The language is spoken by some two thousand people. Virtually all speakers are bilingual, also speaking Nepali, the national Indo-Aryan lingua franca, but Belhare is still the preferred means of communication.