Are we there yet? Encoder-decoder neural networks as cognitive models of English past tense inflection

Corkery, Maria; Matusevych, Yevgen; Goldwater, Sharon

doi:10.48550/arxiv.1906.01280

Cited by 2 publications

(2 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most previous models of morphology acquisition have focused on capturing (1) the cognitive mechanism that allows for limited generalization, that is, applying the past tense -ed to regular forms and not for irregular forms and/or (2) a possible stage of overgeneralization in which irregular verbs are occasionally produced with the regular inflection, for example, goed (Corkery, Matusevych, & Goldwater, 2019;Cottrell & Plunkett, 1994;Hare & Elman, 1995;Hoeffner, 1992;Kirov & Cotterell, 2018;Legate & Yang, 2007;MacWhinney & Leinbach, 1991;O'Donnell, 2015;Pinker & Prince, 1988;Plunkett & Juola, 1999;Rumelhart & McClelland, 1986;Yang, 2016). Those that have tried to capture the variable production of the regular -ed suffix and the bare form have either assumed that children's grammars contain incorrect form-meaning mappings, that is that they erroneously associate past tense to bare forms (Legate & Yang, 2007) or have made predictions about form production without distinguishing whether or not children are trying to specifically produce past tense meanings in these cases (Freudenthal, Gobet, & Pine, 2023;Freudenthal, Pine, & Gobet, 2010Freudenthal, Pine, Jones, & Gobet, 2015;Freudenthal, Ramscar, Leonard, & Pine, 2021).…”

Section: The Acquisition Of Past Tense Regular Formmentioning

confidence: 99%

When Children's Production Deviates From Observed Input: Modeling the Variable Production of the English Past Tense

Barak,

Harmon,

Feldman

et al. 2023

Cognitive Science

View full text Add to dashboard Cite

As children gradually master grammatical rules, they often go through a period of producing form‐meaning associations that were not observed in the input. For example, 2‐ to 3‐year‐old English‐learning children use the bare form of verbs in settings that require obligatory past tense meaning while already starting to produce the grammatical –ed inflection. While many studies have focused on overgeneralization errors, fewer studies have attempted to explain the root of this earlier stage of rule acquisition. In this work, we use computational modeling to replicate children's production behavior prior to the generalization of past tense production in English. We illustrate how seemingly erroneous productions emerge in a model, without being licensed in the grammar and despite the model aiming at conforming to grammatical forms. Our results show that bare form productions stem from a tension between two factors: (1) trying to produce a less frequent meaning (the past tense) and (2) being unable to restrict the production of frequent forms (the bare form) as learning progresses. Like children, our model goes through a stage of bare form production and then converges on adult‐like production of the regular past tense, showing that these different stages can be accounted for through a single learning mechanism.

show abstract

Section: The Acquisition Of Past Tense Regular Formmentioning

confidence: 99%

When Children's Production Deviates From Observed Input: Modeling the Variable Production of the English Past Tense

Barak,

Harmon,

Feldman

et al. 2023

Cognitive Science

View full text Add to dashboard Cite

show abstract

“…The pairs of cues I investigated are those included in Table 1. To be more certain of the model's behavior in general (Corkery et al, 2019), I repeat the experiment (training a model and extracting its perceptual distances for the Garner paradigm) with 70 different random seeds, which determine the weight initializations.…”

Section: Training and Experimentalmentioning

confidence: 99%

Probing a Neural Network Model of Sound Change for Perceptual Integration

Hughes

2023

AMP

View full text Add to dashboard Cite

The cross-linguistic tendency for contrast shifts to occur between some cues more than others has been investigated typologically and experimentally (Yang 2019), but with less attention in computational modeling. This paper adapts a human experimental paradigm (Kingston et al. 2008) to the speech perception component of a neural network model of sound change (Beguš 2020) to better understand how it processes acoustic cues in the context of Yang’s proposal that auditory dimensions affect which cues are more likely to undergo contrast shift. Piloting this neural network probing technique, I find evidence that the model integrates different pairs of English stop voicing cues than humans do, suggesting that further amendments to the model are necessary to implement Yang (2019)’s account. In general, these results highlight potential acoustic processing differences between humans and the model under investigation, a Convolutional Neural Network, which is commonly used in spoken language applications.

show abstract

Are we there yet? Encoder-decoder neural networks as cognitive models of English past tense inflection

Cited by 2 publications

References 0 publications

When Children's Production Deviates From Observed Input: Modeling the Variable Production of the English Past Tense

When Children's Production Deviates From Observed Input: Modeling the Variable Production of the English Past Tense

Probing a Neural Network Model of Sound Change for Perceptual Integration

Contact Info

Product

Resources

About