| Sumario: | Some linguistic phenomena, such as egophoricity or vowel harmony, have a clearly restricted areal or genealogical distribution in the languages of the world. In typological studies of such phenomena, the sample languages are often selected among languages that are known, on the basis of earlier research, to exhibit the phenomenon under study, rather than using an extensive sample with a balanced world-wide representation of languages. The paper proposes to call this approach to language sampling phenomenon-based sampling. After a short overview of types of language sampling in typology, the notion of phenomenon-based language sampling is introduced and defined. A survey of relevant articles in the journal Linguistic Typology 1997-2024 is made to form a general picture of the kinds of phenomenon-based solutions to sampling that have been used in typological studies over the last thirty years. Attention is paid to the ways of identifying the languages that exhibit the phenomenon and to the ways of selecting the sample languages from among the languages thus identified. The various methodological solutions are compared, and interesting similarities and differences are taken up for discussion. It is recommended that instead of downplaying the role of sampling by characterizing their sample as a convenience sample, studies taking a phenomenon-based approach to sampling should explicate their sampling principles and describe the process in terms of phenomenon-based sampling.
|
|---|