TY - JOUR
T1 - Evaluating models of robust word recognition with serial reproduction
AU - Meylan, Stephan C.
AU - Nair, Sathvik
AU - Griffiths, Thomas L.
N1 - Funding Information:
We thank the Computational Cognitive Science Lab at U.C. Berkeley for valuable feedback. This material is based on work supported by the National Science Foundation (Graduate Research Fellowship to S.C.M. under Grant No. DGE-1106400 ), the U.S. Air Force Office of Scientific Research ( FA9550-13-1-0170 to T.L.G), and the Defense Advanced Research Projects Agency (Next Generation Social Sciences to T.L.G).
Publisher Copyright:
© 2021 Elsevier B.V.
PY - 2021/5
Y1 - 2021/5
N2 - Spoken communication occurs in a “noisy channel” characterized by high levels of environmental noise, variability within and between speakers, and lexical and syntactic ambiguity. Given these properties of the received linguistic input, robust spoken word recognition—and language processing more generally—relies heavily on listeners' prior knowledge to evaluate whether candidate interpretations of that input are more or less likely. Here we compare several broad-coverage probabilistic generative language models in their ability to capture human linguistic expectations. Serial reproduction, an experimental paradigm where spoken utterances are reproduced by successive participants similar to the children's game of “Telephone,” is used to elicit a sample that reflects the linguistic expectations of English-speaking adults. When we evaluate a suite of probabilistic generative language models against the yielded chains of utterances, we find that those models that make use of abstract representations of preceding linguistic context (i.e., phrase structure) best predict the changes made by people in the course of serial reproduction. A logistic regression model predicting which words in an utterance are most likely to be lost or changed in the course of spoken transmission corroborates this result. We interpret these findings in light of research highlighting the interaction of memory-based constraints and representations in language processing.
AB - Spoken communication occurs in a “noisy channel” characterized by high levels of environmental noise, variability within and between speakers, and lexical and syntactic ambiguity. Given these properties of the received linguistic input, robust spoken word recognition—and language processing more generally—relies heavily on listeners' prior knowledge to evaluate whether candidate interpretations of that input are more or less likely. Here we compare several broad-coverage probabilistic generative language models in their ability to capture human linguistic expectations. Serial reproduction, an experimental paradigm where spoken utterances are reproduced by successive participants similar to the children's game of “Telephone,” is used to elicit a sample that reflects the linguistic expectations of English-speaking adults. When we evaluate a suite of probabilistic generative language models against the yielded chains of utterances, we find that those models that make use of abstract representations of preceding linguistic context (i.e., phrase structure) best predict the changes made by people in the course of serial reproduction. A logistic regression model predicting which words in an utterance are most likely to be lost or changed in the course of spoken transmission corroborates this result. We interpret these findings in light of research highlighting the interaction of memory-based constraints and representations in language processing.
KW - Generative language models
KW - Iterated learning
KW - Noisy-channel communication
KW - Sentence processing
KW - Serial reproduction
KW - Spoken word recognition
UR - http://www.scopus.com/inward/record.url?scp=85100156542&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85100156542&partnerID=8YFLogxK
U2 - 10.1016/j.cognition.2020.104553
DO - 10.1016/j.cognition.2020.104553
M3 - Article
C2 - 33482474
AN - SCOPUS:85100156542
SN - 0010-0277
VL - 210
JO - Cognition
JF - Cognition
M1 - 104553
ER -