Acquisition of Verbal Morphology in Russian with Attention to Defective Verbs

• In Event: Poster Session 10
  In Poster Session: PS 10 Topic - Language, Communication

Abstract

Learning of morpho-phonological alternations has been claimed to be affected by many factors beyond phonological markedness: the scope and regularity of the patterns (Albright 2009, Yang 2012), type frequencies of a pattern and of lexemes involved, and a range of remote-base effects such as Paradigm Uniformity and Lexical Conservatism (Steriade, 2019). In this work we investigate the following hypotheses about how morpho-phonological alternations are learned: (a) children are biased to produce forms faithful to a surface base-form (Hayes 2004, Do 2013), (b) rules that have wider scope are learned better than rules with narrower scope, (c) remote-bases containing the same alternation significantly aid learning, and rules with lack of support from remote bases can lead to ineffability (Pertsova, 2016). We conducted production experiments with Russian speaking children and adults, testing their ability to produce alternations of existing and pseudo-verbs of several types. We focused on two types of alternations, labial and dental. The labial alternation involves ‘l’-epenthesis between a stem-final labial consonant and the inflectional suffix in 1sg pres. The dental alternations involve mutations of stem-final dental consonants that are not synchronically predictable by a single phonological rule and, thus, create several narrow sub-rules. The stimuli also include dental verbs known or suspected to be defective (lacking 1sg forms).
We used a production experiment in which subjects were exposed to existing and pseudo-verbs in Russian with a picture describing the action, and had to produce 3sg and 1sg forms of the verb. Verbs in each group were normalized for frequency, length, and aspect. The participants were 16 younger children (ages 3-4), 22 older children (ages 4.4 - 5.5) and 30 adults.
Adults performed at ceiling on all existing verbs except dental defective verbs — expected alternations were produced 75% of the time, which is consistent with previous findings (Pertsova & Kuznetzova, 2016). Adults also performed close to ceiling on all pseudo-verbs except those with dental alternations, supporting the hypothesis that these alternations are not learned as well because they have narrower scope. Older children performed best on existing verbs without alternations and labial verbs, followed by regular dental verbs, variation verbs, and worst of all on dental defective verbs. Younger children had a similar pattern, but with the order of regular dental and variation verbs reversed. Older children generalized alternations to pseudo-verbs more than 50% of the time for all verb types, while younger children only generalized to novel labial and no-alternation verbs. For the other pseudo-verbs, they most commonly refused to respond or produced incorrect forms without alternations. For free-variation verbs, adults preferred the variant with alternation. Young children, on the other hand, preferred non-alternating variant if they produced a correct response, with older children producing both variants equally often. Overall, we find effect of rule-scope, and support from remote bases. Defectiveness arises early on, and persists into adulthood.

Authors

• Olga Dmitrievna Vedenina, National Research University Higher School of Economics
  Presenting Author

• Katya Pertsova, University of North Carolina at Chapel Hill
  Non-Presenting Author

• Anastasiya Lopukhina, HSE University
  Non-Presenting Author