Weather and Language Pa ˚ l K. Eriksen 1 , Seppo Kittila ¨ 2 * and Leena Kolehmainen 3 1 Norwegian University, 2 University of Helsinki and 3 University of Eastern 2 Abstract This paper summarizes current findings in the cross-linguistic study of meteorological construc- tions. It provides both a typology of weather events and a typology of encoding formats used for the expression of weather in and across languages. The discussion shows that there is a correlation between these two parameters: there are clear tendencies in the distribution of the various encod- ing types across the various event types. This gives rise to a typology of languages which explains linguistic variation in the encoding of meteorological events. 1. Introduction In this paper, we discuss the linguistic expression of weather in and across languages. Cross-linguistic research on weather sentences is still in its infancy. Research that is entirely devoted to the topic includes Bartens’ (1995) and Salo’s (2011) studies on meteo- rological expressions in Uralic languages, Saarinen’s (1997) research on weather in Euro- pean languages, Mettouchi and Tosco’s (2011) studies in Afroasiatic languages, Ruwet’s (1986) cross-linguistic paper on weather verbs and our own previous articles (Eriksen et al. 2010, forthcoming). The current review is based on the results of these papers. The scope and goals of the previous studies vary. Most studies concentrate either on a given language family (Bartens 1995, Salo 2011) or on languages in a given geographical area (Saarinen 1997, Mettouchi and Tosco 2011). In Eriksen et al. (2010, forthcoming), which the present study is primarily based on, a broader cross-linguistic view is adopted. The second major difference concerns the described phenomenon. While Saarinen (1997) and Mettouchi and Tosco (2011) focus on the coding of raining, the other studies cover the linguistic expression of all kinds of meteorological events including – in addition to rain – for instance the encoding of sunshine, snow, thunder and wind. All previous studies show that the expression of weather is an intriguing topic espe- cially with regard to argument structure. Salo (2011), for example, examines the varying valency of meteorological predicates. Her data demonstrate that although meteorological predicates are often regarded as prototypical impersonal predicates and although they also in Uralic languages frequently occur as impersonal atransitive predicates without any argument at all, this is only one possibility. According to her data, Uralic meteorological verbs exhibit a diversity of argument structure patterns ranging from atransitive verbs to intransitive verbs with a subject and to verbs with an object. The other studies, in turn, highlight the fact that studies of argument structure should take the denotation of the predicate and the argument into consideration, too. Despite of terminological differences, the approaches by Saarinen (1997), Eriksen et al. (2010, forth- coming) and Mettouchi and Tosco (2011) are largely compatible with each other. They show that it is possible to encode a meteorological event by the predicate, by the argu- ment or simultaneously by both of them. L N C 3 3 4 1 B Dispatch: 25.2.12 No. of pages: 20 CE: Vidhya Journal Name Manuscript No. Toc head: TYP & Languages of the World PE: Sharanya Language and Linguistics Compass (2012): 1–20, 10.1002/lnc3.341 ª 2012 The Authors Language and Linguistics Compass ª 2012 Blackwell Publishing Ltd 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49