TY - JOUR
T1 - Even deeper problems with neural network models of language
AU - Bever, Thomas G.
AU - Chomsky, Noam
AU - Fong, Sandiway
AU - Piattelli-Palmarin, Massimo
N1 - Publisher Copyright:
© 2023 Cambridge University Press. All rights reserved.
PY - 2023/12/6
Y1 - 2023/12/6
N2 - We recognize today's deep neural network (DNN) models of language behaviors as engineering achievements. However, what we know intuitively and scientifically about language shows that what DNNs are and how they are trained on bare texts, makes them poor models of mind and brain for language organization, as it interacts with infant biology, maturation, experience, unique principles, and natural law.
AB - We recognize today's deep neural network (DNN) models of language behaviors as engineering achievements. However, what we know intuitively and scientifically about language shows that what DNNs are and how they are trained on bare texts, makes them poor models of mind and brain for language organization, as it interacts with infant biology, maturation, experience, unique principles, and natural law.
UR - http://www.scopus.com/inward/record.url?scp=85178850648&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85178850648&partnerID=8YFLogxK
U2 - 10.1017/S0140525X23001619
DO - 10.1017/S0140525X23001619
M3 - Review article
C2 - 38054358
AN - SCOPUS:85178850648
SN - 0140-525X
VL - 46
JO - Behavioral and Brain Sciences
JF - Behavioral and Brain Sciences
M1 - e387
ER -