Presentations

Abstract

Speaking requires translating concepts into a sequence of sounds. Contemporary models of language production assume that this translation involves a series of steps: from selecting the concepts to be expressed, to phonetic and articulatory encoding of the words. In addition, speakers monitor their planned output using sensorimotor predictive mechanisms. The current work concerns phonetic encoding and the speaker's monitoring of articulation. Specifically, we test whether monitoring is sensitive to the frequency of syllable-sized representations.
We run a series of immediate and delayed syllable production experiments (repetition and reading). We exploit the syllable-frequency effect: in immediate naming, high-frequency syllables are produced faster than low-frequency syllables. The effect is thought to reflect the stronger automatization of motor plan retrieval of high-frequency syllables during phonetic encoding. We predict distinct ERP and spatiotemporal patterns for high- vs. low-frequency syllables. Following articulation, we analyse auditory-evoked N1 responses that – among other features – reflect the suppression of one's own speech. Low-frequency syllables are expected to require more close monitoring, and therefore smaller N1/P2 amplitudes. The results can be important as effects of syllable frequency stand to inform us about the tradeoff between stored versus assembled representations for setting sensory targets in the production of speech.

Abstract

Speaking requires translating concepts into a sequence of sounds. Contemporary models of language production assume that this translation involves a series of steps: from selecting the concepts to be expressed, to phonetic and articulatory encoding of the words. In addition, speakers monitor their planned output using sensorimotor predictive mechanisms. The current work concerns phonetic encoding and the speaker's monitoring of articulation. Specifically, we test whether monitoring is sensitive to the frequency of syllable-sized representations.
We run a series of immediate and delayed syllable production experiments (repetition and reading). We exploit the syllable-frequency effect: in immediate naming, high-frequency syllables are produced faster than low-frequency syllables. The effect is thought to reflect the stronger automatization of motor plan retrieval of high-frequency syllables during phonetic encoding. We predict distinct ERP and spatiotemporal patterns for high- vs. low-frequency syllables. Following articulation, we analyse auditory-evoked N1 responses that – among other features – reflect the suppression of one's own speech. Low-frequency syllables are expected to require more close monitoring, and therefore smaller N1/P2 amplitudes. The results can be important as effects of syllable frequency stand to inform us about the tradeoff between stored versus assembled representations for setting sensory targets in the production of speech.

Abstract

Producing spontaneous speech is challenging. It often contains disfluencies like repetitions, prolongations, silent pauses or filled pauses. Previous research has largely focused on the language-based factors (e.g., planning difficulties) underlying the production of these disfluencies. But research has also shown that some speakers are more disfluent than others. What cognitive mechanisms underlie this difference? We reanalyzed a behavioural dataset of 112 participants, who were assessed on a battery of tasks testing linguistic knowledge, processing speed, non-verbal IQ, working memory, and basic production skills and also produced six 1-minute samples of spontaneous speech (Hintz et al., 2020). We assessed the length and lexical diversity of participants’ speech and determined how often they produced silent pauses and filled pauses. We used network analysis, factor analysis and non-parametric regressions to investigate the relationship between these variables and individual differences in particular cognitive skills. We found that individual differences in linguistic knowledge or processing speed were not related to the production of disfluencies. In contrast, the proportion of filled pauses (relative to all words in the 1-minute narratives) correlated negatively with working memory capacity.

Abstract

Convergence broadly refers to interlocutors’ tendency to progressively sound more like each other over time. Recent empirical work has used various experimental paradigms to observe convergence in voice fundamental frequency (f0). One study used stable mean f0 over trials in a synchronous speech task with manipulated (i.e., high and low) f0 conditions (Bradshaw & McGettigan, 2021). Here, we attempted to replicate this study in Dutch. First, in a reading task, participants read 40 sentences at their own pace to establish f0 baselines. Later, in a synchronous speech task, participants read 80 sentences in synchrony with a speaker whose voice was manipulated ±2st above or below (i.e., for the high and low f0 conditions, respectively) a reference mean f0 value. The reference mean f0 value and the manipulation size were obtained across multiple pre-tests. Our results revealed that the f0 manipulation significantly predicted f0 convergence in both high f0 and low f0 conditions. Furthermore, the proportion of convergers in the sample was larger than those reported by Bradshaw & McGettigan, highlighting the benefits of stimulus optimization. Our study thus provides stronger evidence that the pitch of two talkers tends to converge as they speak together.

Abstract

The semantic interference effect in picture naming entails longer naming latencies for pictures presented with semantically related versus unrelated distractors. One factor suggested to influence the effect is word category. However, results have been inconclusive. Momma et al. (2020) used a sentence-picture interference paradigm where the sentence context (“her singing” or “she’s singing”) disambiguated the word category (noun or verb, respectively) of distractor and target, manipulating their word category match/mismatch. Semantic interference was only found when distractor and target belonged to the same word category, suggesting that syntactic category constrains lexical competition during sentence production. Considering this important theoretical conclusion, we conducted a preregistered replication study with Dutch participants, mirroring the design of the original study. In each of 2 experiments, 60 native speakers read sentences containing sentence-final distractor words that had to be interpreted as nouns or verbs, depending on the sentence context. Subsequently, they named target action pictures as either verbs (experiment 1) or nouns (experiment 2). Results of Experiment 1 showed a main effect of relatedness, suggesting a semantic interference effect regardless of word category. We discuss differences between the original and current study results with cross-linguistic differences in (de)compositional processing and frequency of distractor forms.

Abstract

The name-picture verification task is often used to assess the difficulty of prelexical processes (object recognition and semantic access) during picture naming. However, whether to use responses from word-picture match or from mismatch trials to index the difficulty of pre-lexical processes is debated. Levelt (2002) argued for the use of mismatch trials because on match trials the printed object name might facilitate picture recognition. However, in a study with speakers of Spanish Stadthagen-Gonzalez et al. (2009) showed that visual and conceptual properties of objects only correlated with the latencies of match responses but not with those of mismatch responses and therefore advocated the use of match responses. The present study aimed to replicate Stadthagen- Gonzalez et al. (2009) findings using native British English speakers and English norms for non-lexical and lexical variables. We replicated the finding that non-lexical variables affected the speed of match, but not mismatch responses. However, in addition, we found that lexical variables also affected the speed of match responses, which means that these latencies need to be interpreted with caution. In other words, neither match nor mismatch responses seem ideally suited to assess the difficulty of pre-lexical processes in picture naming. Levelt, W. J. M. (2002). Picture naming and word frequency. Language and Cognitive Processes, 17, 663–671. Stadthagen-Gonzalez, H., Damian, M. F., Pérez, M. A., Bowers, J. S., & Marín, J. (2009). Name-picture verification as a control measure for object naming: A task analysis and norms for a

Abstract

As the popularity of sending messages electronically increased, so did the necessity of conveying messages more efficiently. A popular way of increasing efficiency is to use shortcuts such as gr8 “great” or cu “see you”. The aim of the study was to investigate whether such abbreviations have their own lexical representations or whether they are recorded into the words they stand for. We used associative masked and overt priming in lexical decision tasks. Primes were text shortcuts and their corresponding words. Targets were words that were associatively related to the prime (e.g., cu/see you – GOODBYE), associatively related to a part of the prime (e.g., cu/see you – LOOK) or unrelated to the prime and any part of the prime (e.g., 4u/for you – GOODBYE; 4u/for you – LOOK). In both tasks, responses were faster to targets preceded by related compared to unrelated primes (shortcuts and words). In the overt priming task, we in addition found faster responses when the target was related to a part of the prime. However, this effect was present only for word but not for shortcut primes. These results indicate that shortcuts have their own lexical representations that are independent of the representations of the individual words they replace.

Abstract

Ik laat zien dat het gebruik van sms- en msn-afkortingen geen schadelijke gevolgen heeft op het begrip van geschreven taal zoals blijkt uit een reeks van experimenten. Sterker nog, afkortingen hebben een eigen representatie in ons mentale woordenboek (bijvoorbeeld ‘iig’) en worden niet vertaald naar woorden waaruit ze bestaan (bijvoorbeeld ‘in ieder geval’). Het gebruik van korte tekstberichten via mobiele telefoons en e-mail neemt steeds meer toe en daardoor worden steeds meer afkortingen en acroniemen in de taal gebruikt. Momenteel worden er wereldwijd naar schatting 2000 miljard sms-jes per jaar verstuurd. In Nederland en België sms’t ongeveer 80% van de jongeren regelmatig, en ongeveer 70% van hen gebruikt sms-taal (Daniels, 2009). Het gebruik van afkortingen heeft de schuld gekregen van de verloedering van de Nederlandse taal, iets dat men terug zou zien in de kwaliteit van het Nederlands in e-mails en brieven. Verder gebruik van afkortingen zou de communicatie verslechteren tussen de gebruikers van afkortingen en niet gebruikers, bijvoorbeeld tussen studenten en docenten. De vraag is of dat echt zo is. Uit mijn onderzoek blijkt dat het begrijpen van een tekst niet leidt onder het gebruik van afkortingen in de zinnen

Abstract

Earlier research has shown that speakers naming object pairs can retrieve their names in parallel, but often fail to do so. The conditions for the occurrence of parallel object processing are largely unknown. We examined how associations between objects affected the speakers’ processing strategies. In Experiment 1, participants named object triplets in a left-right-bottom order. During the saccade towards the right object (interloper), it was replaced by a new object (target). We varied the relationship between interloper and target (conceptually related or unrelated) and between interloper and left object (categorically related, associated, unrelated). Target gaze durations were shorter after related than unrelated interlopers. However, this preview effect was independent of the interloper-left object relationship, suggesting that the left and right object were processed sequentially. In Experiments 2 and 3, participants named pairs of associated or unrelated objects. The left objects were repeated several times in each test block to facilitate their processing. Gaze durations for the left objects were longer in associated than unrelated pairs, consistent with parallel processing of the associated objects and interference among them. We will discuss how the difficulty of processing the objects and the relationship between them jointly determine speakers’ processing strategy.

Abstract

Individuals differ greatly on how efficiently they can retrieve words from the mental lexicon in spoken word production. Little is known about how this individual difference might be related to individual difference in other cognitive abilities. The present study investigated the relationship between individuals’ working memory capacity and their efficiency of lexical access. Speeded naming of drawings of actions and objects was used to measure the efficiency of lexical access. To assess working memory capacity, we used the Operation Span (OSPAN) task (Turner & Engle, 1989), in which participants memorize unrelated words while making judgments on simple mathematic operations. We found that action naming response time (RT) was significantly longer than object naming RT. More importantly, there was a significant negative correlation between the action naming RT and the OSPAN score. There was no correlation between object naming RT and the OSPAN score. Ex-Gaussian analyses of the RT distributions revealed that the negative relationship between OSPAN and RT was evident throughout the distributions (i.e., reflected in the Gaussian part of the distribution). These results suggest that when words are difficult to select (our verbcondition) people with larger working memory capacity outperform people with lower working memory capacity. Turner, M. L., & Engle. R. W. (1989). Is working memory capacity task dependent? Journal of Memory and Language, 28, 127-154.