TY - JOUR
T1 - How the input shapes the acquisition of verb morphology
T2 - Elicited production and computational modelling in two highly inflected languages
AU - Engelmann, Felix
AU - Granlund, Sonia
AU - Kolak, Joanna
AU - Szreder, Marta
AU - Ambridge, Ben
AU - Pine, Julian
AU - Theakston, Anna
AU - Lieven, Elena
N1 - Funding Information:
This work was supported by the International Centre for Language and Communicative Development (LuCiD). The support of the Economic and Social Research Council [ES/L008955/1] is gratefully acknowledged. We are very grateful for the help of Minna Kirjavainen, who collected, transcribed and curated the Kirjavainen-Max Planck Finnish corpus, and Viljami Venekoski, who created the animated stimuli for the Finnish experiment. This work benefited greatly from extensive discussions with Grzegorz Krajewski and Jeffrey Elman. Finally, we thank the children, schools, nurseries, parents and teachers who made this research possible. Appendix A
Publisher Copyright:
© 2019
PY - 2019/5
Y1 - 2019/5
N2 - The aim of the present work was to develop a computational model of how children acquire inflectional morphology for marking person and number; one of the central challenges in language development. First, in order to establish which putative learning phenomena are sufficiently robust to constitute a target for modelling, we ran large-scale elicited production studies with native learners of Finnish (N = 77; 35–63 months) and Polish (N = 81; 35–59 months), using a novel method that, unlike previous studies, allows for elicitation of all six person/number forms in the paradigm (first, second and third person; singular and plural). We then proceeded to build and test a connectionist model of the acquisition of person/number marking which not only acquires near adult-like mastery of the system (including generalisation to unseen items), but also yields all of the key phenomena observed in the elicited-production studies; specifically, effects of token frequency and phonological neighbourhood density of the target form, and a pattern whereby errors generally reflect the replacement of low frequency targets by higher-frequency forms of the same verb, or forms with the same person/number as the target, but with a suffix from an inappropriate conjugation class. The findings demonstrate that acquisition of even highly complex systems of inflectional morphology can be accounted for by a theoretical model that assumes rote storage and phonological analogy, as opposed to formal symbolic rules.
AB - The aim of the present work was to develop a computational model of how children acquire inflectional morphology for marking person and number; one of the central challenges in language development. First, in order to establish which putative learning phenomena are sufficiently robust to constitute a target for modelling, we ran large-scale elicited production studies with native learners of Finnish (N = 77; 35–63 months) and Polish (N = 81; 35–59 months), using a novel method that, unlike previous studies, allows for elicitation of all six person/number forms in the paradigm (first, second and third person; singular and plural). We then proceeded to build and test a connectionist model of the acquisition of person/number marking which not only acquires near adult-like mastery of the system (including generalisation to unseen items), but also yields all of the key phenomena observed in the elicited-production studies; specifically, effects of token frequency and phonological neighbourhood density of the target form, and a pattern whereby errors generally reflect the replacement of low frequency targets by higher-frequency forms of the same verb, or forms with the same person/number as the target, but with a suffix from an inappropriate conjugation class. The findings demonstrate that acquisition of even highly complex systems of inflectional morphology can be accounted for by a theoretical model that assumes rote storage and phonological analogy, as opposed to formal symbolic rules.
KW - Computational modelling
KW - Cross-linguistic
KW - Elicited production
KW - Language acquisition
KW - Morphology
KW - Neural networks
KW - Verb marking
UR - http://www.scopus.com/inward/record.url?scp=85061623033&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85061623033&partnerID=8YFLogxK
U2 - 10.1016/j.cogpsych.2019.02.001
DO - 10.1016/j.cogpsych.2019.02.001
M3 - Article
C2 - 30782514
AN - SCOPUS:85061623033
SN - 0010-0285
VL - 110
SP - 30
EP - 69
JO - Cognitive Psychology
JF - Cognitive Psychology
ER -