The productivity of velar insertion in Spanish verbs ending in stem-final /s/
Document Type
Paper Presentation
Presenter Language
English
Research Area
Phonology, corpus linguistics, experimental linguistics
Location
MBSC Gallery Room 308
Start Date
17-10-2024 5:00 PM
End Date
17-10-2024 5:30 PM
Abstract
Some, but not all, of the -er- 2nd and -ir- 3rd conjugation Spanish verbs exhibit a [k] or [ɡ] between the stem and the suffix in the 1SG.PRS.IND and all PRS.SBJV forms (e.g.[kɾes-k-o] / *[kɾes-o] ‘I grow’, yet not in [mes-o] / *[mes-k-o] ‘I rock’), as shown in Table 1.
The verbs with these added velar stops have traditionally been considered irregular because no exceptionless phonological generalization has been identified which can predict which verbs exhibit the velar alternation (Harris 1972; Lloyd 1987). However, these claims have been made without quantitative evidence, based on individually selected examples. The purpose of this study is to challenge this notion by presenting the results of two separate corpus analyses, one which is based on an adult lexicon, and one which is based on a child one.
The first empirical fact that my corpus analyses show is that there are just a few exceptions to the generalization that [k,ɡ] are inserted between stem-final sonorants or /s/ and suffixes beginning with a non-front vowel. More precisely, 117 out of 166 (70.48%) of verbs ending /r, ɾ, l, n, s/ exhibit velar stop insertion, where 112 of those verbs end in /s/, shown in Table 2.Given the proportion of verbs which do exhibit velar insertion, the puzzle is the following: is this an exceptional, irregular class of verbs, or a productive one?
The second important result of this analysis is that it invokes use of a learning algorithm, the Tolerance Principle, to offer a precise, quantitative measure of productivity that allows for a limited number of exceptions (TP; Yang 2016). Specifically, I show that application of the TP predicts velar insertion in Spanish to be productive with 2nd and 3rd conjugation verb stems ending in /s/, while those ending in the sonorants /r, ɾ, l, n/ need to be memorized. A series of psycholinguistic nonce-probe tasks on Spanish and Italian verbs exhibiting velar insertion also suggests that it is a productive class (Bybee & Pardo 1981; Cappellaro et al. 2023), supporting the prediction made by the TP.
Selected references
Bybee, J. L., & Pardo, E. 1981. On lexical and morphological conditioning of alternations: A nonce-probe experiment with Spanish verbs. https://doi.org/10.1515/ling.1981.19.9-10.937
Cappellaro, C., Dumrukcic, N., Fritz, I., Franzon, F., & Maiden, M. 2023. The cognitive reality of morphomes. Evidence from Italian. Morphology, 1-39. https://doi.org/10.1007/s11525-023-09419-2
Harris, J. W. 1972. Five classes of irregular verbs in Spanish. Generative studies in Romance languages, 247-271.
Lloyd, P. M. 1987. From Latin to Spanish: Historical phonology and morphology of the Spanish language (Vol. 173). American Philosophical Society.
Yang, C. 2016. The price of productivity. Manuscript, University of Pennsylvania.
Full version of the abstract, including Tables 1 and 2.
HLS abstract - Tables 1 and 2.pdf (143 kB)
Tables 1 and 2.
The productivity of velar insertion in Spanish verbs ending in stem-final /s/
MBSC Gallery Room 308
Some, but not all, of the -er- 2nd and -ir- 3rd conjugation Spanish verbs exhibit a [k] or [ɡ] between the stem and the suffix in the 1SG.PRS.IND and all PRS.SBJV forms (e.g.[kɾes-k-o] / *[kɾes-o] ‘I grow’, yet not in [mes-o] / *[mes-k-o] ‘I rock’), as shown in Table 1.
The verbs with these added velar stops have traditionally been considered irregular because no exceptionless phonological generalization has been identified which can predict which verbs exhibit the velar alternation (Harris 1972; Lloyd 1987). However, these claims have been made without quantitative evidence, based on individually selected examples. The purpose of this study is to challenge this notion by presenting the results of two separate corpus analyses, one which is based on an adult lexicon, and one which is based on a child one.
The first empirical fact that my corpus analyses show is that there are just a few exceptions to the generalization that [k,ɡ] are inserted between stem-final sonorants or /s/ and suffixes beginning with a non-front vowel. More precisely, 117 out of 166 (70.48%) of verbs ending /r, ɾ, l, n, s/ exhibit velar stop insertion, where 112 of those verbs end in /s/, shown in Table 2.Given the proportion of verbs which do exhibit velar insertion, the puzzle is the following: is this an exceptional, irregular class of verbs, or a productive one?
The second important result of this analysis is that it invokes use of a learning algorithm, the Tolerance Principle, to offer a precise, quantitative measure of productivity that allows for a limited number of exceptions (TP; Yang 2016). Specifically, I show that application of the TP predicts velar insertion in Spanish to be productive with 2nd and 3rd conjugation verb stems ending in /s/, while those ending in the sonorants /r, ɾ, l, n/ need to be memorized. A series of psycholinguistic nonce-probe tasks on Spanish and Italian verbs exhibiting velar insertion also suggests that it is a productive class (Bybee & Pardo 1981; Cappellaro et al. 2023), supporting the prediction made by the TP.
Selected references
Bybee, J. L., & Pardo, E. 1981. On lexical and morphological conditioning of alternations: A nonce-probe experiment with Spanish verbs. https://doi.org/10.1515/ling.1981.19.9-10.937
Cappellaro, C., Dumrukcic, N., Fritz, I., Franzon, F., & Maiden, M. 2023. The cognitive reality of morphomes. Evidence from Italian. Morphology, 1-39. https://doi.org/10.1007/s11525-023-09419-2
Harris, J. W. 1972. Five classes of irregular verbs in Spanish. Generative studies in Romance languages, 247-271.
Lloyd, P. M. 1987. From Latin to Spanish: Historical phonology and morphology of the Spanish language (Vol. 173). American Philosophical Society.
Yang, C. 2016. The price of productivity. Manuscript, University of Pennsylvania.