Beth Levin of Stanford University has an interesting 2017 paper called Talking About the Weather: A Case Study of Precipitation Verbs which begins with a quote from Ronald Langacker:
In the eyes of linguists, such [=weather] expressions are nearly as problematic and ill-behaved as the weather itself: they not only have many special properties, but from one language to the next the same phenomenon is coded linguistically in ways that are lexically or grammatically quite distinct.
She proceeds to discuss “several challenging properties of weather events”:
— Identifiability: Depending on the metereological phenomenon, it can be difficult to identify any participants in the event (e.g., becoming dusk). Perhaps it is possible to recognize a single participant (e.g., snow, rain).
— Independence from the phenomenon: The participant, to the extent it is identifiable, is not independent from the phenomenon itself: snow and rain do not exist outside of the event of snowing or raining […].
— Selectional restrictions: When a participant is expressed, weather verbs impose fairly strict selectional restrictions on it: outside of metaphorical uses, only snow can snow, only rain can rain. […]
— Semantic role: It is hard to determine what semantic role to assign to this participant: when rain rains from the sky, is it acting or being affected?
She brings in Eriksen et al.’s suggestion that “weather event expressions fall into three major types according to which element in the sentence lexicalizes the ‘weather’ phenomenon”: the predicate type (It is raining), the argument type (Rain is falling), and the argument-predicate type (It is raining rain). She then goes on to her own proposed solution for encoding weather events; there is some Chomskyism, but not enough to set off my alarms and make me want to throw the computer against the wall. (We discussed precipitation expressions to some extent in last year’s word-order thread, starting here.) Thanks, Martin!
Recent Comments