Publications

Abstract

Iconic words and signs are characterized by a perceived resemblance between aspects of their form and aspects of their meaning. For example, in English, iconic words include peep and crash, which mimic the sounds they denote, and wiggle and zigzag, which mimic motion. As a semiotic property of words and signs, iconicity has been demonstrated to play a role in word learning, language processing, and language evolution. This paper presents the results of a large-scale norming study for more than 14,000 English words conducted with over 1400 American English speakers. We demonstrate the utility of these ratings by replicating a number of existing findings showing that iconicity ratings are related to age of acquisition, sensory modality, semantic neighborhood density, structural markedness, and playfulness. We discuss possible use cases and limitations of the rating dataset, which is made publicly available.

Abstract

Vocal learning is a behavioral trait in which the social and acoustic environment shapes the vocal repertoire of individuals. Over the past century, the study of vocal learning has progressed at the intersection of ecology, physiology, neuroscience, molecular biology, genomics, and evolution. Yet, despite the complexity of this trait, vocal learning is frequently described as a binary trait, with species being classified as either vocal learners or vocal non-learners. As a result, studies have largely focused on a handful of species for which strong evidence for vocal learning exists. Recent studies, however, suggest a continuum in vocal learning capacity across taxa. Here, we further suggest that vocal learning is a multi-component behavioral phenotype comprised of distinct yet interconnected modules. Discretizing the vocal learning phenotype into its constituent modules would facilitate integration of findings across a wider diversity of species, taking advantage of the ways in which each excels in a particular module, or in a specific combination of features. Such comparative studies can improve understanding of the mechanisms and evolutionary origins of vocal learning. We propose an initial set of vocal learning modules supported by behavioral and neurobiological data and highlight the need for diversifying the field in order to disentangle the complexity of the vocal learning phenotype.

Abstract

It is unknown whether modality affects the efficiency with which humans learn novel word forms and their meanings, with previous studies reporting both written and auditory advantages. The current study implements controls whose absence in previous work likely offers explanation for such contradictory findings. In two novel word learning experiments, participants were trained and tested on pseudoword - novel object pairs, with controls on: modality of test, modality of meaning, duration of exposure and transparency of word form. In both experiments word forms were presented in either their written or spoken form, each paired with a pictorial meaning (novel object). Following a 20-minute filler task, participants were tested on their ability to identify the picture-word form pairs on which they were trained. A between subjects design generated four participant groups per experiment 1) written training, written test; 2) written training, spoken test; 3) spoken training, written test; 4) spoken training, spoken test. In Experiment 1 the written stimulus was presented for a time period equal to the duration of the spoken form. Results showed that when the duration of exposure was equal, participants displayed a written training benefit. Given words can be read faster than the time taken for the spoken form to unfold, in Experiment 2 the written form was presented for 300 ms, sufficient time to read the word yet 65% shorter than the duration of the spoken form. No modality effect was observed under these conditions, when exposure to the word form was equivalent. These results demonstrate, at least for proficient readers, that when exposure to the word form is controlled across modalities the efficiency with which word form-meaning associations are learnt does not differ. Our results therefore suggest that, although we typically begin as aural-only word learners, we ultimately converge on developing learning mechanisms that learn equally efficiently from both written and spoken materials.

Abstract

This study aimed to increase our understanding on the relationship between reading and listening comprehension. Both in comprehension theory and in educational practice, reading and listening comprehension are often seen as interchangeable, overlooking modality-specific aspects of them separately. Three questions were addressed. First, it was examined to what extent reading and listening comprehension comprise modality-specific, distinct skills or an overlapping, domain-general skill in terms of the amount of explained variance in one comprehension type by the opposite comprehension type. Second, general and modality-unique subskills of reading and listening comprehension were sought by assessing the contributions of the foundational skills word reading fluency, vocabulary, memory, attention, and inhibition to both comprehension types. Lastly, the practice of using either listening comprehension or vocabulary as a proxy of general comprehension was investigated. Reading and listening comprehension tasks with the same format were assessed in 85 second and third grade children. Analyses revealed that reading comprehension explained 34% of the variance in listening comprehension, and listening comprehension 40% of reading comprehension. Vocabulary and word reading fluency were found to be shared contributors to both reading and listening comprehension. None of the other cognitive skills contributed significantly to reading or listening comprehension. These results indicate that only part of the comprehension process is indeed domain-general and not influenced by the modality in which the information is provided. Especially vocabulary seems to play a large role in this domain-general part. The findings warrant a more prominent focus of modality-specific aspects of both reading and listening comprehension in research and education.

Abstract

For bilinguals, speaking in a second language (L2) compared to the native language (L1) is usually more difficult. In this study we asked whether the difficulty in L2 production reflects increased demands imposed on domain-general or core language mechanisms. We compared the brain response to speech production in L1 and L2 within two functionally-defined networks in the brain: the Multiple Demand (MD) network and the language network. We found that speech production in L2 was linked to a widespread increase of brain activity in the domain-general MD network. The language network did not show a similarly robust differences in processing speech in the two languages, however, we found increased response to L2 production in the language-specific portion of the left inferior frontal gyrus (IFG). To further explore our results, we have looked at domain-general and language-specific response within the brain structures postulated to form a Bilingual Language Control (BLC) network. Within this network, we found a robust increase in response to L2 in the domain-general, but also in some language-specific voxels including in the left IFG. Our findings show that L2 production strongly engages domain-general mechanisms, but only affects language sensitive portions of the left IFG. These results put constraints on the current model of bilingual language control by precisely disentangling the domain-general and language-specific contributions to the difficulty in speech production in L2.

Abstract

When bilingual speakers switch back to speaking in their native language (L1) after having used their second language (L2), they often experience difficulty in retrieving words in their L1. This phenomenon is referred to as the L2 after-effect. We used the L2 after-effect as a lens to explore the neural bases of bilingual language control mechanisms. Our goal was twofold: first, to explore whether bilingual language control draws on domain-general or language-specific mechanisms; second, to investigate the precise mechanism(s) that drive the L2 after-effect. We used a precision fMRI approach based on functional localizers to measure the extent to which the brain activity that reflects the L2 after-effect overlaps with the language network (Fedorenko et al., 2010) and the domain-general multiple demand network (Duncan, 2010), as well as three task-specific networks that tap into interference resolution, lexical retrieval, and articulation. Forty-two Polish–English bilinguals participated in the study. Our results show that the L2 after-effect reflects increased engagement of domain-general but not language-specific resources. Furthermore, contrary to previously proposed interpretations, we did not find evidence that the effect reflects increased difficulty related to lexical access, articulation, and the resolution of lexical interference. We propose that difficulty of speech production in the picture naming paradigm—manifested as the L2 after-effect—reflects interference at a nonlinguistic level of task schemas or a general increase of cognitive control engagement during speech production in L1 after L2.

Abstract

The language network of the human brain has core components in the inferior frontal cortex and superior/middle temporal cortex, with left-hemisphere dominance in most people. Functional specialization and interconnectivity of these neocortical regions is likely to be reflected in their molecular and cellular profiles. Excitatory connections between cortical regions arise and innervate according to layer-specific patterns. Here we generated a new gene expression dataset from human postmortem cortical tissue samples from core language network regions, using spatial transcriptomics to discriminate gene expression across cortical layers. Integration of these data with existing single-cell expression data identified 56 genes that showed differences in laminar expression profiles between frontal and temporal language cortex together with upregulation in layer II/III and/or layer V/VI excitatory neurons. Based on data from large-scale genome-wide screening in the population, DNA variants within these 56 genes showed set-level associations with inter-individual variation in structural connectivity between left-hemisphere frontal and temporal language cortex, and with predisposition to dyslexia. The axon guidance genes SLIT1 and SLIT2 were consistently implicated. These findings identify region-specific patterns of laminar gene expression as a feature of the brain’s language network.

Abstract

Tokenization significantly influences language models (LMs)’ performance. This paper traces the evolution of tokenizers from word-level to subword-level, analyzing how they balance tokens and types to enhance model adaptability while controlling complexity. Despite subword tokenizers like Byte Pair Encoding (BPE) overcoming many word tokenizer limitations, they encounter difficulties in handling non-Latin languages and depend heavily on extensive training data and computational resources to grasp the nuances of multiword expressions (MWEs). This article argues that tokenizers, more than mere technical tools, should drawing inspiration from the cognitive science about human language processing. This study then introduces the “Principle of Least Effort” from cognitive science, that humans naturally seek to reduce cognitive effort, and discusses the benefits of this principle for tokenizer development. Based on this principle, the paper proposes that the Less-is-Better (LiB) model could be a new approach for LLM tokenizer. The LiB model can autonomously learn an integrated vocabulary consisting of subwords, words, and MWEs, which effectively reduces both the numbers of tokens and types. Comparative evaluations show that the LiB tokenizer outperforms existing word and BPE tokenizers, presenting an innovative method for tokenizer development, and hinting at the possibility of future cognitive science-based tokenizers being more efficient.

Abstract

Early word learning involves mapping individual words to their meanings and building organized semantic representations among words. Previous corpus-based studies (e.g., using text from websites, newspapers, child-directed speech corpora) demonstrated that linguistic information such as word co-occurrence alone is sufficient to build semantically organized word knowledge. The present study explored two new research directions to advance understanding of how infants acquire semantically organized word knowledge. First, infants in the real world hear words surrounded by contextual information. Going beyond inferring semantic knowledge merely from language input, we examined the role of extra-linguistic contextual information in learning semantic knowledge. Second, previous research relies on large amounts of linguistic data to demonstrate in-principle learning, which is unrealistic compared with the input children receive. Here, we showed that incorporating extra-linguistic information provides an efficient mechanism through which semantic knowledge can be acquired with a small amount of data infants perceive in everyday learning contexts, such as toy play.

Abstract

There is substantial evidence that infants prefer infant-directed speech (IDS) to adult-directed speech (ADS). The strongest evidence for this claim has come from two large-scale investigations: i) a community-augmented meta-analysis of published behavioral studies and ii) a large-scale multi-lab replication study. In this paper, we aim to improve our understanding of the IDS preference and its boundary conditions by combining and comparing these two data sources across key population and design characteristics of the underlying studies. Our analyses reveal that both the meta-analysis and multi-lab replication show moderate effect sizes (d ≈ 0.35 for each estimate) and that both of these effects persist when relevant study-level moderators are added to the models (i.e., experimental methods, infant ages, and native languages). However, while the overall effect size estimates were similar, the two sources diverged in the effects of key moderators: both infant age and experimental method predicted IDS preference in the multi-lab replication study, but showed no effect in the meta-analysis. These results demonstrate that the IDS preference generalizes across a variety of experimental conditions and sampling characteristics, while simultaneously identifying key differences in the empirical picture offered by each source individually and pinpointing areas where substantial uncertainty remains about the influence of theoretically central moderators on IDS preference. Overall, our results show how meta-analyses and multi-lab replications can be used in tandem to understand the robustness and generalizability of developmental phenomena.

Abstract

Word learning happens in everyday contexts with many words and many potential referents for those words in view at the same time. It is challenging for young learners to find the correct referent upon hearing an unknown word at the moment. This problem of referential uncertainty has been deemed as the crux of early word learning (Quine, 1960). Recent empirical and computational studies have found support for a statistical solution to the problem termed cross-situational learning. Cross-situational learning allows learners to acquire word meanings across multiple exposures, despite each individual exposure is referentially uncertain. Recent empirical research shows that infants, children and adults rely on cross-situational learning to learn new words (Smith & Yu, 2008; Suanda, Mugwanya, & Namy, 2014; Yu & Smith, 2007). However, researchers have found evidence supporting two very different theoretical accounts of learning mechanisms: Hypothesis Testing (Gleitman, Cassidy, Nappa, Papafragou, & Trueswell, 2005; Markman, 1992) and Associative Learning (Frank, Goodman, & Tenenbaum, 2009; Yu & Smith, 2007). Hypothesis Testing is generally characterized as a form of learning in which a coherent hypothesis regarding a specific word-object mapping is formed often in conceptually constrained ways. The hypothesis will then be either accepted or rejected with additional evidence. However, proponents of the Associative Learning framework often characterize learning as aggregating information over time through implicit associative mechanisms. A learner acquires the meaning of a word when the association between the word and the referent becomes relatively strong. In this chapter, we consider these two psychological theories in the context of cross-situational word-referent learning. By reviewing recent empirical and cognitive modeling studies, our goal is to deepen our understanding of the underlying word learning mechanisms by examining and comparing the two theoretical learning accounts.

Abstract

This present study identified an optimal model representing the relationship between orthography and phonology in Chinese handwritten production using dynamic causal modeling, and further explored how this model was modulated by word frequency and syllable frequency. Each model contained five volumes of interest in the left hemisphere (angular gyrus [AG], inferior frontal gyrus [IFG], middle frontal gyrus [MFG], superior frontal gyrus [SFG], and supramarginal gyrus [SMG]), with the IFG as the driven input area. Results showed the superiority of a model in which both the MFG and the AG connected with the IFG, supporting the orthography autonomy hypothesis. Word frequency modulated the AG → SFG connection (information flow from the orthographic lexicon to the orthographic buffer), and syllable frequency affected the IFG → MFG connection (information transmission from the semantic system to the phonological lexicon). This study thus provides new insights into the connectivity architecture of neural substrates involved in writing.

Abstract

Electrophysiological brain activity has been shown to synchronize with the quasi-regular repetition of grammatical phrases in connected speech—so-called phrase-rate neural tracking. Current debate centers around whether this phenomenon is best explained in terms of the syntactic properties of phrases or in terms of syntax-external information, such as the sequential repetition of parts of speech. As these two factors were confounded in previous studies, much of the literature is compatible with both accounts. Here, we used electroencephalography (EEG) to determine if and when the brain is sensitive to both types of information. Twenty native speakers of Mandarin Chinese listened to isochronously presented streams of monosyllabic words, which contained either grammatical two-word phrases (e.g., catch fish, sell house) or non-grammatical word combinations (e.g., full lend, bread far). Within the grammatical conditions, we varied two structural factors: the position of the head of each phrase and the type of attachment. Within the non-grammatical conditions, we varied the consistency with which parts of speech were repeated. Tracking was quantified through evoked power and inter-trial phase coherence, both derived from the frequency-domain representation of EEG responses. As expected, neural tracking at the phrase rate was stronger in grammatical sequences than in non-grammatical sequences without syntactic structure. Moreover, it was modulated by both attachment type and head position, revealing the structure-sensitivity of phrase-rate tracking. We additionally found that the brain tracks the repetition of parts of speech in non-grammatical sequences. These data provide an integrative perspective on the current debate about neural tracking effects, revealing that the brain utilizes regularities computed over multiple levels of linguistic representation in guiding rhythmic computation.

Abstract

In sentences like “the mouse that chased the cat was hungry”, the syntactically correct interpretation (the mouse chases the cat) is contradicted by semantic and pragmatic knowledge. Previous research has shown that L1 speakers sometimes base sentence interpretation on this type of knowledge (so-called “shallow” or “good-enough” processing). We made use of both behavioural and ERP measurements to investigate whether L2 learners differ from native speakers in the extent to which they engage in “shallow” syntactic processing. German learners of Dutch as well as Dutch native speakers read sentences containing relative clauses (as in the example above) for which the plausible thematic roles were or were not reversed, and made plausibility judgments. The results show that behaviourally, L2 learners had more difficulties than native speakers to discriminate plausible from implausible sentences. In the ERPs, we replicated the previously reported finding of a “semantic P600” for semantic reversal anomalies in native speakers, probably reflecting the effort to resolve the syntax-semantics conflict. In L2 learners, though, this P600 was largely attenuated and surfaced only in those trials that were judged correctly for plausibility. These results generally point at a more prevalent, but not exclusive occurrence of shallow syntactic processing in L2 learners.

Abstract

Statistical learning (SL) is postulated to play an important role in the process of language acquisition as well as in other cognitive functions. It was found to enable learning of various types of statistical patterns across different sensory modalities. However, few studies have distinguished distributional SL (DSL) from sequential and spatial SL, or examined DSL across modalities using comparable tasks. Considering the relevance of such findings to the nature of SL, the current study investigated the modality- and stimulus-specificity of DSL. Using a within-subject design we compared DSL performance in auditory and visual modalities. For each sensory modality, two stimulus types were used: linguistic versus non-linguistic auditory stimuli and temporal versus spatial visual stimuli. In each condition, participants were exposed to stimuli that varied in their length as they were drawn from two categories (short versus long). DSL was assessed using a categorization task and a production task. Results showed that learners’ performance was only correlated for tasks in the same sensory modality. Moreover, participants were better at categorizing the temporal signals in the auditory conditions than in the visual condition, where in turn an advantage of the spatial condition was observed. In the production task participants exaggerated signal length more for linguistic signals than non-linguistic signals. Together, these findings suggest that DSL is modality- and stimulus-sensitive.

Abstract

Prosodic prominence can enhance memory for the prominent words. This mnemonic benefit has been linked to listeners’ allocation of attention and deeper processing, which leads to more robust semantic representations. We investigated whether, in addition to the well-established effect at the semantic level, there was a memory benefit for prominent words at the phonological level. To do so, participants (48 native speakers of Dutch), first performed an accent judgement task, where they had to discriminate accented from unaccented words, and accented from unaccented pseudowords. All stimuli were presented in lists. They then performed an old/new recognition task for the stimuli. Accuracy in the accent judgement task was equally high for words and pseudowords. In the recognition task, performance was, as expected, better for words than pseudowords. More importantly, there was an interaction of accent with word type, with a significant advantage for accented compared to unaccented words, but not for pseudowords. The results confirm the memory benefit for accented compared to unaccented words seen in earlier studies, and they are consistent with the view that prominence primarily affects the semantic encoding of words. There was no evidence for an additional memory benefit arising at the phonological level.

Abstract

Semantic unification during sentence comprehension has been associated with amplitude change of the N400 in event-related potential (ERP) studies, and activation in the left inferior frontal gyrus (IFG) in functional magnetic resonance imaging (fMRI) studies. However, the specificity of this activation to semantic unification remains unknown. To more closely examine the brain processes involved in semantic unification, we employed simultaneous EEG-fMRI to time-lock the semantic unification related N400 change, and integrated trial-by-trial variation in both N400 and BOLD change beyond the condition-level BOLD change difference measured in traditional fMRI analyses. Participants read sentences in which semantic unification load was parametrically manipulated by varying cloze probability. Separately, ERP and fMRI results replicated previous findings, in that semantic unification load parametrically modulated the amplitude of N400 and cortical activation. Integrated EEG-fMRI analyses revealed a different pattern in which functional activity in the left IFG and bilateral supramarginal gyrus (SMG) was associated with N400 amplitude, with the left IFG activation and bilateral SMG activation being selective to the condition-level and trial-level of semantic unification load, respectively. By employing the EEG-fMRI integrated analyses, this study among the first sheds light on how to integrate trial-level variation in language comprehension.

Abstract

Research into the role of brain oscillations in basic perceptual and cognitive functions has suggested that the alpha rhythm reflects functional inhibition while the beta rhythm reflects neural ensemble (re)activation. However, little is known regarding the generalization of these proposed fundamental operations to linguistic processes, such as speech comprehension and production. Here, we recorded magnetoencephalography in participants performing a novel rule-switching paradigm. Specifically, Dutch native speakers had to produce an alternative exemplar from the same category or a feature of a given target word embedded in spoken sentences (e.g., for the word “tuna”, an exemplar from the same category—“seafood”—would be “shrimp”, and a feature would be “pink”). A cue indicated the task rule—exemplar or feature—either before (pre-cue) or after (retro-cue) listening to the sentence. Alpha power during the working memory delay was lower for retro-cue compared with that for pre-cue in the left hemispheric language-related regions. Critically, alpha power negatively correlated with reaction times, suggestive of alpha facilitating task performance by regulating inhibition in regions linked to lexical retrieval. Furthermore, we observed a different spatiotemporal pattern of beta activity for exemplars versus features in the right temporoparietal regions, in line with the proposed role of beta in recruiting neural networks for the encoding of distinct categories. Overall, our study provides evidence for the generalizability of the role of alpha and beta oscillations from perceptual to more “complex, linguistic processes” and offers a novel task to investigate links between rule-switching, working memory, and word production.

Abstract

Swedish morphemes are classified as prosodically specified or prosodically unspecified, depending on lexical or phonological stress, respectively. Here, we investigate the allomorphy of the suffix -(i)sk, which indicates the distinction between lexical and phonological stress; if attached to a lexically stressed morpheme, it takes a non-syllabic form (-sk), whereas if attached to a phonologically stressed morpheme, an epenthetic vowel is inserted (-isk). Using mismatch negativity (MMN), we explored the neural processing of this allomorphy across lexically stressed and phonologically stressed morphemes. In an oddball paradigm, participants were occasionally presented with congruent and incongruent derivations, created by the suffix -(i)sk, within the repetitive presentation of their monomorphemic stems. The results indicated that the congruent derivation of the lexically stressed stem elicited a larger MMN than the incongruent sequences of the same stem and the derivational suffix, whereas after the phonologically stressed stem a non-significant tendency towards an opposite pattern was observed. We argue that the significant MMN response to the congruent derivation in the lexical stress condition is in line with lexical MMN, indicating a holistic processing of the sequence of lexically stressed stem and derivational suffix. The enhanced MMN response to the incongruent derivation in the phonological stress condition, on the other hand, is suggested to reflect combinatorial processing of the sequence of phonologically stressed stem and derivational suffix. These findings bring a new aspect to the dual-system approach to neural processing of morphologically complex words, namely the specification of word stress.

Abstract

In Swedish, words are associated with either of two pitch contours known as Accent 1 and Accent 2. Using a psychometric test, we investigated how listeners judge pitch accent violations while interpreting discourse. Forty native speakers of Central Swedish were presented with auditory dialogues, where test words were appropriately or inappropriately accented in a given context, and asked to judge the correctness of sentences containing the test words. Data indicated a statistically significant effect of wrong accent pattern on the correctness judgment. Both Accent 1 and Accent 2 violations interfered with the coherent interpretation of discourse and were judged as incorrect by the listeners. Moreover, there was a statistically significant difference in the perceived correctness between the accent patterns. Accent 2 violations led to a lower correctness score compared to Accent 1 violations, indicating that the listeners were more sensitive to pitch accent violations in Accent 2 words than in Accent 1 words. This result is in line with the notion that Accent 2 is marked and lexically represented in Central Swedish. Taken together, these findings indicate that listeners use both Accent 1 and Accent 2 to arrive at the correct interpretation of the linguistic input, while assigning varying degrees of relevance to them depending on their markedness.

Abstract

Speakers remember their own utterances better than those of their interlocutors, suggesting that language production is beneficial to memory. This may be partly explained by a generation effect: The act of generating a word is known to lead to a memory advantage (Slamecka & Graf, 1978). In earlier work, we showed a generation effect for recognition of images (Zormpa, Brehm, Hoedemaker, & Meyer, 2019). Here, we tested whether the recognition of their names would also benefit from name generation. Testing whether picture naming improves memory for words was our primary aim, as it serves to clarify whether the representations affected by generation are visual or conceptual/lexical. A secondary aim was to assess the influence of processing time on memory. Fifty-one participants named pictures in three conditions: after hearing the picture name (identity condition), backward speech, or an unrelated word. A day later, recognition memory was tested in a yes/no task. Memory in the backward speech and unrelated conditions, which required generation, was superior to memory in the identity condition, which did not require generation. The time taken by participants for naming was a good predictor of memory, such that words that took longer to be retrieved were remembered better. Importantly, that was the case only when generation was required: In the no-generation (identity) condition, processing time was not related to recognition memory performance. This work has shown that generation affects conceptual/lexical representations, making an important contribution to the understanding of the relationship between memory and language.

Abstract

The production effect (better memory for words read aloud than words read silently) and the picture superiority effect (better memory for pictures than words) both improve item memory in a picture naming task (Fawcett, J. M., Quinlan, C. K., & Taylor, T. L. (2012). Interplay of the production and picture superiority effects: A signal detection analysis. Memory (Hove, England), 20(7), 655–666. doi:10.1080/09658211.2012.693510). Because picture naming requires coming up with an appropriate label, the generation effect (better memory for generated than read words) may contribute to the latter effect. In two forced-choice memory experiments, we tested the role of generation in a picture naming task on later recognition memory. In Experiment 1, participants named pictures silently or aloud with the correct name or an unreadable label superimposed. We observed a generation effect, a production effect, and an interaction between the two. In Experiment 2, unreliable labels were included to ensure full picture processing in all conditions. In this experiment, we observed a production and a generation effect but no interaction, implying the effects are dissociable. This research demonstrates the separable roles of generation and production in picture naming and their impact on memory. As such, it informs the link between memory and language production and has implications for memory asymmetries between language production and comprehension.