Setningarleg og félagsleg skilyrði agnarfærslu í ensku

Félagsmálfræðingurinn Laurel McKenzie, dósent við New York University, heldur opinn fyrirlestur á vegum Málvísindastofnunar Háskóla Íslands í Eddu mánudaginn 6. maí kl. 15:00-17:00. 

Fyrirlesturinn nefnist Syntactic and social conditioning of the English particle verb alternation.

Um fyrirlesturinn

There is a large literature proposing or asserting that syntactic variation is unlikely to show social conditioning (see Levon and Buchstaller 2015 for a recent review). One commonly-cited counterexample to this is the English particle verb alternation. This refers to the word order alternation observed among a class of transitive verb-particle combinations, exemplified with the particle verb clean out in (1). The particle may remain joined with the verb, as in (1a): the verb clean and the particle out are linearly adjacent, with the object your gutters following. Alternatively, the particle verb may be split by its object, as in (1b): the verb clean and the particle out are linearly separated from each other by the gutters.

(1) a. Joined: Look, just clean out your gutters and increase the grading away from your house

      b. Split: A lot of people didn’t get up to clean the gutters out before winter

In a small study of radio speech, Kroch and Small (1978) find that radio show hosts and their in-studio guests are more likely to use the joined variant of the particle verb alternation than are listeners calling in to the show. They argue that this finding can be explained by in-studio speakers showing increased sensitivity to standard language norms, which prescribe use of the joined variant.

Kroch and Small were careful to point out the need to replicate their study with more data and with multivariate analysis. We take advantage of the RadioTalk Corpus (Beeferman et al. 2019, 2021), a 2.8 billion-word corpus of US talk radio broadcasts, to attempt to replicate Kroch and Small’s findings. Our data set consists of 15,000+ observations of transitive particle verbs with non-pronominal objects.

First, we examine language-internal effects on the particle verb alternation, where we confirm some well-known findings (e.g. an effect of direct object length) and introduce some undocumented ones (e.g. an effect of particle prosody). Most notably, we find that previous researchers’ coding of a particle verb’s semantic compositionality can be improved on when particle verbs are instead coded for their membership in syntactic classes.

Second, we turn to the social factor of a speaker's role in the radio broadcast (in-studio host/guest vs. listeners calling in over the phone). Here, however, Kroch and Small’s result fails to replicate: in-studio speakers and callers show no difference in particle verb variant choice. We conclude by discussing possible reasons for this, ranging from diachronic change in the particle verb alternation over the 20th century to changes in talk radio norms since the 1970s.

References

Beeferman, Doug, William Brannon, and Deb Roy. 2019. RadioTalk: A large-scale corpus of talk radio transcripts. Proceedings of the 20th Conference of the International Speech Communication Association (INTERSPEECH 2019).

Beeferman, Doug, William Brannon, and Deb Roy. 2021. RadioTalk GitHub Repository. Laboratory for Social Machines. https://github.com/mit-ccc/RadioTalk

Kroch, Anthony, and Cathy Small. 1978. Grammatical ideology and its effect on speech. In Linguistic Variation: Models and Methods, ed. David Sankoff, 45–55. New York: Academic Press.

Levon, Erez, and Isabelle Buchstaller. 2015. Perception, cognition, and linguistic structure: The effect of linguistic modularity and cognitive style on sociolinguistic processing. Language Variation and Change 27(3): 319–348.