Linguistics of Temperature
Temperature phenomena are universal, relatively easily perceptible by humans and crucial for them, but their conceptualisation involves a complex interplay between external reality, bodily experience and evaluation of the relevant properties with regard to their functions in the human life. The meanings of temperature terms are, thus, both embodied and perspectival. Rather than reflecting the external world objectively, they offer a naïve picture of it, permeated with folk theories that are based on people’s experience and rooted in their culture (cultural models). Languages differ as to how many temperature terms they have and how these categorize the temperature domain in general Closely related languages can show remarkable differences in their uses of temperature adjectives, even when these are cognates to each other; conversely, temperature systems can show remarkable areal patterns. Temperature terms can belong to different word classes, even within one and the same language (adjectives – ”cold”, verbs – ”to freeze”, nouns – ”coldness”). Languages vary in their word-class attribution of temperature concepts: thus, for instance, many languages lack temperature adjectives. Word-class attribution and, further, lexicalization of temperature expressions and the possible syntactic constructions in which they can be used are sensitive to their semantics.
Temperature meanings are often semantically related to other meanings, either synchronically (within a polysemantic lexeme) or diachronically. Thus, temperature concepts often serve as source domains for various metaphors and are extended to other perceptional modalities (‘hot spices’, ‘warm colour’). Temperature meanings can also develop from others, e.g., ‘burn, fire’ >’hot’, or ’ice’ > ’cold’. Finally, the meanings of temperature terms can also change within the temperature domain itself, e.g. ‘warm, hot’ > ‘lukewarm’, as in Lat. tep- ‘warm’ vs. English tepid ‘lukewarm’. While some languages show extensive semantic derivation from the temperature domain, others lack it or use it to a limited degree. Languages vary as to which temperature term has predominantly positive associations in its extended use (cf. ‘cold’ in Wolof vs. ‘warm’ in the European languages), partly due to the different climatic conditions.
Temperature terms have, on the whole, received relatively little attention. Cross-linguistic research on temperature is mainly restricted to Sutrop (1998, 1999) and Plank (2003), which focus on how many basic temperature terms there are in a language and how they carve up the domain among themselves. There has been no cross-linguistic research on the grammatical behaviour of temperature expressions, apart from a few mentions.
In theoretical semantics, temperature adjectives have mainly figured in discussions of lexical fields, antonymy and linguistic scales (cf. Lehrer 1970, Cruse & Togia 1995, Sutrop 1998, cf. also Clausner & Croft 1999). Koptjevskaja-Tamm & Rakhilina 2006 suggest that linguistic categorization of the temperature domain is sensitive to several parameters, that are important and salient for humans and can be distinguishable by simple procedures relating to the human body. Within the Natural-Semantic Metalanguage, Goddard & Wierzbicka (2006) propose the general formula for describing the language-specific meanings of temperature terms via reference to fire.
Extended uses of temperature words have been studied indirectly in cognitive linguistics, primarily in research on the metaphors underlying emotions, e.g. AFFECTION IS WARMTH (Lakoff & Johnson 1997:50) and ANGER IS HEAT (Kövecses 1995, also Goossens 1998; cf. also Shindo 1998-99). An important question raised in Geeraerts & Grondelaers (1995) is to what degree such extensions reflect universal metaphorical patterns or are based on common cultural traditions. The current empirical evidence for the suggested metaphors is still relatively meagre.
This paper is an analysis of lexical categorisation of the temperature domain in modern Eastern Armenian. Compared to the vast research outline proposed in (Koptjevskaja-Tamm 2011), this paper has several important limitations. First, it is focused on non-derived, primary temperature terms (most of which happen to be adjectives or nouns, or both). Derived lexical items, as well as lexical items that apply to temperature phenomena only secondarily, are not considered. Second, it focuses on lexical rather than morphosyntactic categorisation, in the sense that more attention is paid, again, to lexical items than to the morphosyntactic patterns they are associated with. In a sense, we focus on elementary morphological units – dedicated temperature roots – rather than on the morphological and morphosyntactic patterns they are involved in.
As Eastern Armenian represents an elaborated system of temperature terms (some dozen lexical items), even under such restrictions the linguistic data presented below is well worth of analysis. To a certain extent, the issues of part-of-speech derivation are brought into consideration. It should also be noted that, as the examples in the paper show, there seem to be no sharp differences between e.g. morphosyntactic treatment of the subjects of tactile vs. ambient vs. personal temperatures, such as shown in Koptjevskaja-Tamm (2011) through a comparison of French, Finnish, Japanese and German examples. It is true that, as in many languages, in Armenian ambient temperature is often expressed by impersonal predication (cf. ex. 3 below). However, experiencers of personal temperature (as most experiencers in general), carriers of tactile temperature and meteorological phenomena expressed by explicit NPs (‘day’, ‘air’, ‘sun’) predicated by ambient terms are all treated as subjects. They are assigned nominative marking and control verbal agreement. There certainly may be finer morphosyntactic differences, but these are left for future research(ers).
The structure of the paper is as follows. Section 1 provides a brief genealogical, structural, sociolinguistic and historical account of the Eastern Armenian language and introduces the main source of the present study (Eastern Armenian National Corpus). Section 2 provides an overview of the temperature words of Eastern Armenian along the lines of the typological dimensions around which the present volume is centred, including the pivotal distinction between tactile and ambient temperatures. Personal temperature terms are in a sense independent from this main opposition, so that the discussion of personal temperatures goes alongside but apart from the main argument, both here and later in Section 6. Section 3 is an overview of metaphorical uses of the temperature terms. Section 4 considers part-of-speech properties and some aspects of derivational morphology of the Eastern Armenian temperature words. Section 5 is an account of the known etymologies of the terms. Section 6 introduces some data on relative textual frequencies of temperature terms. Section 7 is a summary of the paper.