Introduction to the special issue emergence of speech and language from prediction error: error-driven language models

Nixon, Jessie S.; Tomaschek, Fabian

doi:10.1080/23273798.2023.2197650

Cited by 3 publications

(3 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…the cue's predictive value. If a more salient cue is present, while an expected outcome fails to occur, this results in a higher amount of prediction error (Nixon and Tomaschek, 2023). From this, a stronger unlearning effect follows.…”

Section: Discussionmentioning

confidence: 99%

“…This is called cue competition (Hoppe et al, 2022, Nixon, 2020, Ramscar et al, 2010, Rescorla and Wagner, 1972. Activation from a cue to an outcome increases in proportion to the prediction error present during a learning event (Nixon and Tomaschek, 2023). If a cue often predicts an outcome, there is less uncertainty about whether the outcome follows, and, as a result, there is also less prediction error remaining.…”

Section: Learning Theorymentioning

confidence: 99%

See 1 more Smart Citation

One cue’s loss is another cue’s gain – Learning morphophonology through unlearning

Mujezinović,

Kapatsinski,

van de Vijver

2023

Preprint

View full text Add to dashboard Cite

A word often expresses many different morphological functions. Which part of a form contributes to which part of the overall meaning is often unclear, which raises the question as to how such functions are learned. While linguistic studies tacitly assume cooccurrence of cues and outcomes to suffice in learning these functions, recent studies suggest contingency rather than contiguity to be crucial (Nixon, 2020, Ramscar et al., 2010). Within error-driven learning theory, cues gain association strength if they predict a certain outcome, and lose strength if the outcome is absent. This reduction of association strength is called unlearning. So far, it is unclear if such unlearning has consequences for cue-outcome associations beyond the ones which get reduced. To test for such consequences of unlearning, we taught participants morphophonological patterns in an artificial language learning experiment. In one block, the cues to two functions – plural and diminutive – co-occurred within the same word forms. In another block, a single cue to only one of these two functions was presented in a different set of word forms. We wanted to find out, if participants unlearn this cue’s association to the incorrect function, and if this results in a higher relative weight to this function for the absent cue. Our results show that participants learned that the absent cue predicts the absent function better, if unlearning was possible. This effect was stronger, in case the present cue was more salient. This shows that unlearning takes place, even if no alternative cues to an absent outcome are provided, highlighting that learners take both positive and negative evidence into account – as predicted by domain general error driven learning.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Learning Theorymentioning

confidence: 99%

One cue’s loss is another cue’s gain – Learning morphophonology through unlearning

Mujezinović,

Kapatsinski,

van de Vijver

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Because models that employ the error-driven learning rules are easily interpreted, a number of recent psycholinguistic studies have made use of the learning equations proposed by Rescorla and Wagner (1972) (see also Widrow & Hoff, 1960;Rosenblatt, 1962) to formalize learning problems and derive predictions about participants' learning behavior. Various models, including the dual-path model (Chang, Dell, & Bock, 2006, Chang, 2002 and reinforcement learning (e.g., Harmon, Idemaru, & Kapatsinski, 2019), as well as the Rescorla-Wagner model, have also been used to successfully predict learning behavior (see Nixon & Tomaschek, 2023, and the contributions therein for recent developments in error-driven learning of language using various types of models). This work has included studies in domains such as: word learning (Nixon, 2020, Ramscar et al, 2010Ramscar, Dye, Popick, & O'Donnell-McCarthy, 2011, Ramscar et al, 2013a, morphological structure learning in children (Ramscar et al, 2010, Ramscar et al, 2011, Ramscar et al, 2013b, Ramscar et al, 2013a and adults (Arnon & Ramscar, 2012, Nixon, 2020, Ramscar, 2013, Vujović, Ramscar, & Wonnacott, 2021; morphological processing in adults (Baayen, Milin, Durdevic, Hendrix, & Marelli, 2011, Nieder, Tomaschek, Cohrs, & de Vijver, 2021, Seyfarth & Myslin, 2014, Tomaschek, Plag, Ernestus, & Baayen, 2019; lexical selection in production (Harmon & Kapatsinski, 2020); auditory comprehension and recognition (Arnold, Tomaschek, Sering, Lopez, & Baayen, 2017, Shafaei-Bajestan & Baayen, 2018; effects of the lexicon on articulation (Schmitz, Plag, Baer-Henney, & Stein, 2021, Tucker, Sims, & Baayen, 2019, Tomaschek et al,...…”

Section: Discriminative Learningmentioning

confidence: 99%

The Keys to the Future? An Examination of Statistical Versus Discriminative Accounts of Serial Pattern Learning

Tomaschek,

Ramscar,

Nixon

2024

Cognitive Science

Self Cite

View full text Add to dashboard Cite

Sequence learning is fundamental to a wide range of cognitive functions. Explaining how sequences—and the relations between the elements they comprise—are learned is a fundamental challenge to cognitive science. However, although hundreds of articles addressing this question are published each year, the actual learning mechanisms involved in the learning of sequences are rarely investigated. We present three experiments that seek to examine these mechanisms during a typing task. Experiments 1 and 2 tested learning during typing single letters on each trial. Experiment 3 tested for “chunking” of these letters into “words.” The results of these experiments were used to examine the mechanisms that could best account for them, with a focus on two particular proposals: statistical transitional probability learning and discriminative error‐driven learning. Experiments 1 and 2 showed that error‐driven learning was a better predictor of response latencies than either n‐gram frequencies or transitional probabilities. No evidence for chunking was found in Experiment 3, probably due to interspersing visual cues with the motor response. In addition, learning occurred across a greater distance in Experiment 1 than Experiment 2, suggesting that the greater predictability that comes with increased structure leads to greater learnability. These results shed new light on the mechanism responsible for sequence learning. Despite the widely held assumption that transitional probability learning is essential to this process, the present results suggest instead that the sequences are learned through a process of discriminative learning, involving prediction and feedback from prediction error.

show abstract

One Cue's Loss Is Another Cue's Gain—Learning Morphophonology Through Unlearning

Mujezinović,

Kapatsinski,

van de Vijver

2024

Cognitive Science

View full text Add to dashboard Cite

A word often expresses many different morphological functions. Which part of a word contributes to which part of the overall meaning is not always clear, which raises the question as to how such functions are learned. While linguistic studies tacitly assume the co‐occurrence of cues and outcomes to suffice in learning these functions (Baer‐Henney, Kügler, & van de Vijver, 2015; Baer‐Henney & van de Vijver, 2012), error‐driven learning suggests that contingency rather than contiguity is crucial (Nixon, 2020; Ramscar, Yarlett, Dye, Denny, & Thorpe, 2010). In error‐driven learning, cues gain association strength if they predict a certain outcome, and they lose strength if the outcome is absent. This reduction of association strength is called unlearning. So far, it is unclear if such unlearning has consequences for cue–outcome associations beyond the ones that get reduced. To test for such consequences of unlearning, we taught participants morphophonological patterns in an artificial language learning experiment. In one block, the cues to two morphological outcomes—plural and diminutive—co‐occurred within the same word forms. In another block, a single cue to only one of these two outcomes was presented in a different set of word forms. We wanted to find out, if participants unlearn this cue's association with the outcome that is not predicted by the cue alone, and if this allows the absent cue to be associated with the absent outcome. Our results show that if unlearning was possible, participants learned that the absent cue predicts the absent outcome better than if no unlearning was possible. This effect was stronger if the unlearned cue was more salient. This shows that unlearning takes place even if no alternative cues to an absent outcome are provided, which highlights that learners take both positive and negative evidence into account—as predicted by domain general error‐driven learning.

show abstract

Introduction to the special issue emergence of speech and language from prediction error: error-driven language models

Cited by 3 publications

References 46 publications

One cue’s loss is another cue’s gain – Learning morphophonology through unlearning

One cue’s loss is another cue’s gain – Learning morphophonology through unlearning

The Keys to the Future? An Examination of Statistical Versus Discriminative Accounts of Serial Pattern Learning

One Cue's Loss Is Another Cue's Gain—Learning Morphophonology Through Unlearning

Contact Info

Product

Resources

About