Morpho-Phonetic Effects in Speech Production: Modeling the Acoustic Duration of English Derived Words With Linear Discriminative Learning

Abstract

Recent evidence for the influence of morphological structure on the phonetic output goes unexplained by established models of speech production and by theories of the morphology-phonology interaction. Linear discriminative learning (LDL) is a recent computational approach in which such effects can be expected. We predict the acoustic duration of 4,530 English derivative tokens with the morphological functions DIS, NESS, LESS, ATION, and IZE in natural speech data by using predictors derived from a linear discriminative learning network. We find that the network is accurate in learning speech production and comprehension, and that the measures derived from it are successful in predicting duration. For example, words are lengthened when the semantic support of the word's predicted articulatory path is stronger. Importantly, differences between morphological categories emerge naturally from the network, even when no morphological information is provided. The results imply that morphological effects on duration can be explained without postulating theoretical units like the morpheme, and they provide further evidence that LDL is a promising alternative for modeling speech production.

Understanding the Phonetic Characteristics of Speech Under Uncertainty—Implications of the Representation of Linguistic Knowledge in Learning and Processing

“…Critically, we find that when inflectional functions of [ɐ] serve as inputs to the learning network, uncertainty associated with these functions obtained from the network is a better statistical predictor for [ɐ]'s phonetic characteristics than when inflectional functions serve as outputs. Accordingly, the present study contributes to a line of research that investigates how uncertainty affects speech production through a combination of computational modeling of learning and an examination of the predictions of these models for the phonetic characteristics of actual speech (for example Baayen et al, 2019 ; Tomaschek et al, 2019 ; Tucker et al, 2019 ; Stein and Plag, 2021 ; Schmitz et al, 2021b in the present special issue).…”

How trial-to-trial learning shapes mappings in the mental lexicon: Modelling Lexical Decision with Linear Discriminative Learning

“…The representations and learning mechanisms that we are using in the present study have been found to be useful to predict behavioural data in previous work (e.g. Chuang et al, 2020b;Stein and Plag, 2021;Schmitz et al, 2021). We compare the measures extracted from the DLM with classical psycholinguistic predictors such as orthographic neighbourhood density.…”

“…This setup has been successful at predicting a range of behavioural data related to lexical processing (e.g. Chuang et al, 2020b;Stein and Plag, 2021;Schmitz et al, 2021;Chuang et al, 2022a;Gahl and Baayen, 2022;Cassani et al, 2019;Heitmeier and Baayen, 2020). To model trial-to-trial learning we make use of the Widrow-Hoff learning rule (Widrow and Hoff, 1960) which allows learning of real-valued semantic vectors.…”

“…The DLM posits simple mappings between representations of words' forms and their meanings (Baayen et al, 2018(Baayen et al, , 2019. The DLM has been successful both in modelling different morphological systems across a range of languages, such as Latin, English, German, Estonian, Korean and Maltese (Baayen et al, 2018(Baayen et al, , 2019Chuang et al, 2020aChuang et al, , 2022aNieder et al, 2021), but at the same time also at modelling a range of behavioural data (Cassani et al, 2019;Heitmeier and Baayen, 2020;Chuang et al, 2020b;Shafaei-Bajestan et al, 2021;Heitmeier et al, 2021;Stein and Plag, 2021;Schmitz et al, 2021). It implements learning using an error-driven learning rule for continuous data (Widrow and Hoff, 1960) which is closely related to the later developed Rescorla-Wagner rule.…”

The keys to the future? An examination of associative versus discriminative accounts of Serial Pattern Learning

“…Discriminative learning process thus capture incremental, implicit learning, rather than explicit changes in perception or beliefs based on logic or reasoning (see e.g., Ramscar et al, 2013a;Nixon, Poelstra and van Rij, 2022), and can be formalised by training neural networks using virtually any computational learning algorithm that takes into account error during learning (though whether a model is discriminative or associative depends on the representation of cues and outcomes in the model, see Bröker and Ramscar, 2020). Because they are relatively simple to implement, and models based on them are easily interpreted, a number of recent psycholinguistic studies have made use of the learning equations proposed by Rescorla and Wagner (1972) to formalise learning problems and derive predictions about participants' behaviour during experiments in domains such as: word learning (Nixon, 2020;Ramscar et al, 2010;Ramscar, Dye, Popick and 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 and adults (Arnon and Ramscar, 2012;Ramscar, 2013;Vujović, Ramscar and Wonnacott, 2021); morphological processing in adults (Baayen, Milin, Durdevic, Hendrix and Marelli, 2011;Tomaschek, Plag, Ernestus and Baayen, 2019;Nieder, Tomaschek, Cohrs and de Vijver, 2021); auditory comprehension and recognition (Arnold, Tomaschek, Sering, Lopez and Baayen, 2017;Shafaei-Bajestan and Baayen, 2018); effects of the lexicon on articulation (Tucker, Sims and Baayen, 2019;Tomaschek et al, 2019;Schmitz, Plag, Baer-Henney and Stein, 2021;Stein and Plag, 2021;Tomaschek and Ramscar, 2022); phonetic learning (Nixon, 2018(Nixon, , 2020Tomaschek, 2020, 2021); and the trial-by-trial changes in the neural signal that occur with learning…”