Exemplar by feature applicability matrices and other Dutch normative data for semantic concepts

The impact of ChatGPT on human data collection: A case study involving typicality norming data

Tools like ChatGPT, which allow people to unlock the potential of large language models (LLMs), have taken the world by storm. ChatGPT's ability to produce written output of remarkable quality has inspired, or forced, academics to consider its consequences for both research and education. In particular, the question of what constitutes authorship, and how to evaluate (scientific) contributions has received a lot of attention. However, its impact on (online) human data collection has mostly flown under the radar. The current paper examines how ChatGPT can be (mis)used in the context of generating norming data. We found that ChatGPT is able to produce sensible output, resembling that of human participants, for a typicality rating task. Moreover, the test-retest reliability of ChatGPT's ratings was similar to that of human participants tested 1 day apart. We discuss the relevance of these findings in the context of (online) human data collection, focusing both on opportunities (e.g., (risk-)free pilot data) and challenges (e.g., data fabrication).

Semantic representations in inferior frontal and lateral temporal cortex during picture naming, reading, and repetition

Reading, naming, and repetition are classical neuropsychological tasks widely used in the clinic and psycholinguistic research. While reading and repetition can be accomplished by following a direct or an indirect route, pictures can be named only by means of semantic mediation. By means of fMRI multivariate pattern analysis, we evaluated whether this well-established fundamental difference at the cognitive level is associated at the brain level with a difference in the degree to which semantic representations are activated during these tasks. Semantic similarity between words was estimated based on a word association model. Twenty subjects participated in an event-related fMRI study where the three tasks were presented in pseudo-random order. Linear discriminant analysis of fMRI patterns identified a set of regions that allow to discriminate between words at a high level of word-specificity across tasks. Representational similarity analysis was used to determine whether semantic similarity was represented in these regions and whether this depended on the task performed. The similarity between neural patterns of the left Brodmann area 45 (BA45) and of the superior portion of the left supramarginal gyrus correlated with the similarity in meaning between entities during picture naming. In both regions, no significant effects were seen for repetition or reading. The semantic similarity effect during picture naming was significantly larger than the similarity effect during the two other tasks. In contrast, several regions including left anterior superior temporal gyrus and left ventral BA44/frontal operculum, among others, coded for semantic similarity in a task-independent manner. These findings provide new evidence for the dynamic, task-dependent nature of semantic representations in the left BA45 and a more task-independent nature of the representational activation in the lateral temporal cortex and ventral BA44/frontal operculum.

Semantic feature production norms for manipulable objects

Feature generation tasks and feature databases are important for understanding how knowledge is organized in semantic memory, as they reflect not only the kinds of information that individuals hold about objects but also how objects are conceptually represented. Traditionally, semantic norms focus on a variety of object categories and, as a result, have a small number of concepts per semantic category. Here, our main goal is to create a more fine-grained feature database exclusively for one category of objects-manipulable objects. This database contributes to the understanding of within-category, content-specific processing. To achieve this, we asked 130 participants to freely generate features for 80 manipulable objects and another group of 32 participants to generate action features for the same objects. We then compared our databases with other published semantic norms and found high similarity between them. In our databases, we calculated the similarity between objects in terms of visual, functional, encyclopaedic, and action feature types using Spearman correlation, Baker's gamma index, and cophenetic correlation. We discovered that objects were grouped in a distinctive and meaningful way according to feature type. Finally, we tested the validity of our databases by asking three groups of participants to perform a feature verification experiment while manipulating production frequency. Our results demonstrate that participants can recognize and associate the features of our databases with specific manipulable objects. Participants were faster to verify high-frequency features than low-frequency features. Overall, our data provide important insights into how we process manipulable objects and can be used to further inform cognitive and neural theories of object processing and identification.

Independency of Coding for Affective Similarities and for Word Co-occurrences in Temporal Perisylvian Neocortex

Word valence is one of the principal dimensions in the organization of word meaning. Co-occurrence-based similarities calculated by predictive natural language processing models are relatively poor at representing affective content, but very powerful in their own way. Here, we determined how these two canonical but distinct ways of representing word meaning relate to each other in the human brain both functionally and neuroanatomically. We re-analysed an fMRI study of word valence. A co-occurrence-based model was used and the correlation with the similarity of brain activity patterns was compared to that of affective similarities. The correlation between affective and co-occurrence-based similarities was low (r = 0.065), confirming that affect was captured poorly by co-occurrence modelling. In a whole-brain representational similarity analysis, word embedding similarities correlated significantly with the similarity between activity patterns in a region confined to the superior temporal sulcus to the left, and to a lesser degree to the right. Affective word similarities correlated with the similarity in activity patterns in this same region, confirming previous findings. The affective similarity effect extended more widely beyond the superior temporal cortex than the effect of co-occurrence-based similarities did. The effect of co-occurrence-based similarities remained unaltered after partialling out the effect of affective similarities (and vice versa). To conclude, different aspects of word meaning, derived from affective judgements or from word co-occurrences, are represented in superior temporal language cortex in a neuroanatomically overlapping but functionally independent manner.

Delayed Lexical Access and Cascading Effects on Spreading Semantic Activation During Spoken Word Recognition in Children With Hearing Aids and Cochlear Implants: Evidence From Eye-Tracking

OBJECTIVE

The objective of this study was to characterize the dynamics of real-time lexical access, including lexical competition among phonologically similar words, and spreading semantic activation in school-age children with hearing aids (HAs) and children with cochlear implants (CIs). We hypothesized that developing spoken language via degraded auditory input would lead children with HAs or CIs to adapt their approach to spoken word recognition, especially by slowing down lexical access.

DESIGN

Participants were children ages 9- to 12-years old with normal hearing (NH), HAs, or CIs. Participants completed a Visual World Paradigm task in which they heard a spoken word and selected the matching picture from four options. Competitor items were either phonologically similar, semantically similar, or unrelated to the target word. As the target word unfolded, children's fixations to the target word, cohort competitor, rhyme competitor, semantically related item, and unrelated item were recorded as indices of ongoing lexical access and spreading semantic activation.

RESULTS

Children with HAs and children with CIs showed slower fixations to the target, reduced fixations to the cohort competitor, and increased fixations to the rhyme competitor, relative to children with NH. This wait-and-see profile was more pronounced in the children with CIs than the children with HAs. Children with HAs and children with CIs also showed delayed fixations to the semantically related item, although this delay was attributable to their delay in activating words in general, not to a distinct semantic source.

CONCLUSIONS

Children with HAs and children with CIs showed qualitatively similar patterns of real-time spoken word recognition. Findings suggest that developing spoken language via degraded auditory input causes long-term cognitive adaptations to how listeners recognize spoken words, regardless of the type of hearing device used. Delayed lexical access directly led to delays in spreading semantic activation in children with HAs and CIs. This delay in semantic processing may impact these children's ability to understand connected speech in everyday life.

The effect of an aphasia ID card on the processing of language produced by a speaker with nonfluent aphasia

INTRODUCTION

We tested whether aphasia self-disclosure via an aphasia ID card impacts (1) how non-aphasic listeners initially process language produced by a speaker with aphasia and (2) learning of the speaker's error patterns over time.

METHODS

In this eye-tracking experiment, 27 young adults followed instructions recorded by a speaker with nonfluent aphasia while viewing a target picture and a distractor. The Card group (n = 14) was shown a simulated aphasia ID card for the speaker and the No Card group (n = 13) was not. The task was divided into Pre-Observation and Post-Observation blocks. Between blocks, participants observed the speaker making semantic paraphasias. Eye-tracking analyses compared the time course of target advantage (reflecting competition from the distractor picture) and workspace advantage (reflecting attention to task) between groups and blocks.

RESULTS

Pre-Observation, the Card group had a higher target advantage than the No Card group in the post-response window (i.e., after participants had responded), indicating sustained attention to the speaker's language. Across blocks, there was evidence that the Card group (but not the No Card group) learned that the speaker makes semantic paraphasias.

CONCLUSIONS

Aphasia ID cards impacted listeners' processing of language produced by a speaker with nonfluent aphasia. Increased patience and attentiveness may underlie both the Card group's sustained attention to the speaker as well as learning of the speaker's error patterns. Further research should address whether these changes impact communication success between PWA and new conversation partners.

Visual and Affective Multimodal Models of Word Meaning in Language and Mind

One of the main limitations of natural language‐based approaches to meaning is that they do not incorporate multimodal representations the way humans do. In this study, we evaluate how well different kinds of models account for people's representations of both concrete and abstract concepts. The models we compare include unimodal distributional linguistic models as well as multimodal models which combine linguistic with perceptual or affective information. There are two types of linguistic models: those based on text corpora and those derived from word association data. We present two new studies and a reanalysis of a series of previous studies. The studies demonstrate that both visual and affective multimodal models better capture behavior that reflects human representations than unimodal linguistic models. The size of the multimodal advantage depends on the nature of semantic representations involved, and it is especially pronounced for basic‐level concepts that belong to the same superordinate category. Additional visual and affective features improve the accuracy of linguistic models based on text corpora more than those based on word associations; this suggests systematic qualitative differences between what information is encoded in natural language versus what information is reflected in word associations. Altogether, our work presents new evidence that multimodal information is important for capturing both abstract and concrete words and that fully representing word meaning requires more than purely linguistic information. Implications for both embodied and distributional views of semantic representation are discussed.

New Spanish semantic feature production norms for older adults

Semantic feature production norms are a useful tool for researchers to have empirically collected data about the semantic representations of a particular population. As older adults have been shown to have certain differences in their semantic knowledge organization in comparison with younger adults, it is relevant for them to have their own normative data. Thus we present here the first Spanish semantic feature production norms for older adults. They contain information about the feature composition of 400 concrete concepts. We also provide information about some feature and concept variables as well as comparisons between young and old adults on these variables.

Similarity Judgment Within and Across Categories: A Comprehensive Model Comparison

Similarity is one of the most important relations humans perceive, arguably subserving category learning and categorization, generalization and discrimination, judgment and decision making, and other cognitive functions. Researchers have proposed a wide range of representations and metrics that could be at play in similarity judgment, yet have not comprehensively compared the power of these representations and metrics for predicting similarity within and across different semantic categories. We performed such a comparison by pairing nine prominent vector semantic representations with seven established similarity metrics that could operate on these representations, as well as supervised methods for dimensional weighting in the similarity function. This approach yields a factorial model structure with 126 distinct representation-metric pairs, which we tested on a novel dataset of similarity judgments between pairs of cohyponymic words in eight categories. We found that cosine similarity and Pearson correlation were the overall best performing unweighted similarity functions, and that word vectors derived from free association norms often outperformed word vectors derived from text (including those specialized for similarity). Importantly, models that used human similarity judgments to learn category-specific weights on dimensions yielded substantially better predictions than all unweighted approaches across all types of similarity functions and representations, although dimension weights did not generalize well across semantic categories, suggesting strong category context effects in similarity judgment. We discuss implications of these results for cognitive modeling and natural language processing, as well as for theories of the representations and metrics involved in similarity.

Distinctive responses in anterior temporal lobe and ventrolateral prefrontal cortex during categorization of semantic information

Semantic categorization is a fundamental ability in language as well as in interaction with the environment. However, it is unclear what cognitive and neural basis generates this flexible and context dependent categorization of semantic information. We performed behavioral and fMRI experiments with a semantic priming paradigm to clarify this. Participants conducted semantic decision tasks in which a prime word preceded target words, using names of animals (mammals, birds, or fish). We focused on the categorization of unique marine mammals, having characteristics of both mammals and fish. Behavioral experiments indicated that marine mammals were semantically closer to fish than terrestrial mammals, inconsistent with the category membership. The fMRI results showed that the left anterior temporal lobe was sensitive to the semantic distance between prime and target words rather than category membership, while the left ventrolateral prefrontal cortex was sensitive to the consistency of category membership of word pairs. We interpreted these results as evidence of existence of dual processes for semantic categorization. The combination of bottom-up processing based on semantic characteristics in the left anterior temporal lobe and top-down processing based on task and/or context specific information in the left ventrolateral prefrontal cortex is required for the flexible categorization of semantic information.

A model-based analysis of the impairment of semantic memory

We use cognitive models to evaluate three theories of the change in semantic memory caused by Alzheimer's disease. We use data from 14,096 clinical assessments of 3602 Alzheimer's patients and their caregivers. Each patient completed a semantic memory task involving the odd-one-out comparison of animal names. Each patient was also independently evaluated to determine their level of impairment. Our cognitive models assume a feature-based representation of the animals and odd-one-out choice probabilities based on common-feature similarities. We find no evidence for the restructured representation hypothesis, which claims that impairment causes changes in the features used to represent stimuli. We also find no evidence for the attention change hypothesis, which claims that impairment causes greater attention to be given to concrete features at the expense of more abstract features. We do find evidence for the noisy access hypothesis, which claims that odd-one-out choices become less determined by semantic similarity and more prone to the simple response strategy of choosing the last option. We conclude that the noisy access hypothesis provides a simple account of odd-one-out choice behavior throughout the progression of Alzheimer's disease. More elaborate theories involving changes to underlying mental representations and attention processes need to provide evidence they are superior to the noisy access account.

Lexicosemantic, affective, and distributional norms for 1,000 Dutch adjectives

The research of the word is still very much the research of the noun. Adjectives have been largely overlooked, despite being the second-largest word class in many languages and serving an important communicative function, because of the rich, nuanced qualifications they afford. Adjectives are also ideally suited to study the interface between cognition and emotion, as they naturally cover the entire range of lexicosemantic variables such as imageability (infinite-green), and affective variables such as valence (sad-happy). We illustrate this by showing how the centrality of words in the mental lexicon varies as a function of the words' affective dimensions, using newly collected norms for 1,000 Dutch adjectives. The norms include the lexicosemantic variables age of acquisition, familiarity, concreteness, and imageability; the affective variables valence, arousal, and dominance; and a variety of distributional variables, including network statistics resulting from a large-scale word association study. The norms are freely available from https://osf.io/nyg8v/, for researchers studying adjectives specifically or for whom adjectives constitute convenient stimuli to study other topics, such as vagueness, inference, spatial cognition, or affective word processing.

Semantic similarity and associated abstractness norms for 630 French word pairs

The representation of abstract concepts remains a challenge, justifying the need for further experimental investigation. To that end, we introduce a normative database for 630 semantically similar French word pairs and associated levels of abstractness for 1260 isolated words based on data from 900 subjects. The semantic similarity and abstractness norms were obtained in two studies using 7-point scales. The database is organised according to word-pair semantic similarity, abstractness, and associated lexical variables such as word length (in number of letters), word frequency, and other lexical variables to allow for matching of experimental material. The associated variables were obtained by cross-referencing our database with other known psycholinguistic databases including Lexique (New et al., 2004), the French Lexicon Project (Ferrand et al., 2010), Wordlex (Gimenes & New, 2016), and MEGALEX (Ferrand et al., 2018). We introduced sufficient diversity to allow researchers to select pairs with varying levels of semantic similarity and abstractness. In addition, it is possible to use these data as continuous or discrete variables. The full data are available in the supplementary materials as well as on OSF ( https://osf.io/qsd4v/ ).

Informational content of cosine and other similarities calculated from high-dimensional Conceptual Property Norm data

To study concepts that are coded in language, researchers often collect lists of conceptual properties produced by human subjects. From these data, different measures can be computed. In particular, inter-concept similarity is an important variable used in experimental studies. Among possible similarity measures, the cosine of conceptual property frequency vectors seems to be a de facto standard. However, there is a lack of comparative studies that test the merit of different similarity measures when computed from property frequency data. The current work compares four different similarity measures (cosine, correlation, Euclidean and Chebyshev) and five different types of data structures. To that end, we compared the informational content (i.e., entropy) delivered by each of those 4 × 5 = 20 combinations, and used a clustering procedure as a concrete example of how informational content affects statistical analyses. Our results lead us to conclude that similarity measures computed from lower-dimensional data fare better than those calculated from higher-dimensional data, and suggest that researchers should be more aware of data sparseness and dimensionality, and their consequences for statistical analyses.

Representation of associative and affective semantic similarity of abstract words in the lateral temporal perisylvian language regions

The examination of semantic cognition has traditionally identified word concreteness as well as valence as two of the principal dimensions in the representation of conceptual knowledge. More recently, corpus-based vector space models as well as graph-theoretical analysis of large-scale task-related behavioural responses have revolutionized our insight into how the meaning of words is structured. In this fMRI study, we apply representational similarity analysis to investigate the conceptual representation of abstract words. Brain activity patterns were related to a cued-association based graph as well as to a vector-based co-occurrence model of word meaning. Twenty-six subjects (19 females and 7 males) performed an overt repetition task during fMRI. First, we performed a searchlight classification procedure to identify regions where activity is discriminable between abstract and concrete words. These regions were left inferior frontal gyrus, the upper and lower bank of the superior temporal sulcus bilaterally, posterior middle temporal gyrus and left fusiform gyrus. Representational Similarity Analysis demonstrated that for abstract words, the similarity of activity patterns in the cortex surrounding the superior temporal sulcus bilaterally and in the left anterior superior temporal gyrus reflects the similarity in word meaning. These effects were strongest for semantic similarity derived from the cued association-based graph and for affective similarity derived from either of the two models. The latter effect was mainly driven by positive valence words. This research highlights the close neurobiological link between the information structure of abstract and affective word content and the similarity in activity pattern in the lateral and anterior temporal language system.

Core features: measures and characterization for different languages

According to the feature-based view of semantic representation, concepts can be represented as distributed networks of semantic features, which contribute with different weights to determine the overall meaning of a concept. The study of semantic features, typically collected in property generation tasks, is enriched with measures indicating the informativeness and distinctiveness of a given feature for the related concepts. However, while these measures have been provided in several languages (e.g. Italian, Spanish and English), they have hardly been applied comparatively across languages. The purpose of this paper is to investigate language-related differences and similarities emerging from the semantic representation of aggregated core features. Features with higher salience for a set of concrete concepts are identified and described in terms of their feature type. Then, comparisons are made between domains (natural vs. artefacts) and languages (Italian, Spanish and English) and descriptive statistics are provided. These results show that the characterization of concrete concepts is overall fairly stable across languages, although interesting cross-linguistic differences emerged. We will discuss the implications of our findings in relation to the theoretical paradigm of semantic feature norms, as well as in relation to speakers' mutual understanding in multilingual settings.

A practical primer on processing semantic property norm data

Semantic property listing tasks require participants to generate short propositions (e.g., [Formula: see text], [Formula: see text]) for a specific concept (e.g., DOG). This task is the cornerstone of the creation of semantic property norms which are essential for modeling, stimuli creation, and understanding similarity between concepts. Despite the wide applicability of semantic property norms for a large variety of concepts across different groups of people, the methodological aspects of the property listing task have received less attention, even though the procedure and processing of the data can substantially affect the nature and quality of the measures derived from them. The goal of this paper is to provide a practical primer on how to collect and process semantic property norms. We will discuss the key methods to elicit semantic properties and compare different methods to derive meaningful representations from them. This will cover the role of instructions and test context, property preprocessing (e.g., lemmatization), property weighting, and relationship encoding using ontologies. With these choices in mind, we propose and demonstrate a processing pipeline that transparently documents these steps, resulting in improved comparability across different studies. The impact of these choices will be demonstrated using intrinsic (e.g., reliability, number of properties) and extrinsic measures (e.g., categorization, semantic similarity, lexical processing). This practical primer will offer potential solutions to several long-standing problems and allow researchers to develop new property listing norms overcoming the constraints of previous studies.

The Cognitive Neuroscience of Stable and Flexible Semantic Typicality

Typicality effects are among the most well-studied phenomena in the study of concepts. The classical notion of typicality is that typical concepts share many features with category co-members and few features with members of contrast categories. However, this notion was challenged by evidence that typicality is highly context dependent and not always dependent on central tendency. Dieciuc and Folstein (2019) argued that there is strong evidence for both views and that the two types of typicality effects might depend on different mechanisms. A recent theoretical framework, the controlled semantic cognition framework (Lamdon Ralph et al., 2017) strongly emphasizes the classical view, but includes mechanisms that could potentially account for both kinds of typicality. In contrast, the situated cognition framework (Barsalou, 2009b) articulates the context-dependent view. Here, we review evidence from cognitive neuroscience supporting the two frameworks. We also briefly evaluate the ability of computational models associated with the CSC to account for phenomena supporting SitCog (Rogers and McClelland, 2004). Many predictions of both frameworks are borne out by recent cognitive neuroscience evidence. While the CSC framework can at least potentially account for many of the typicality phenomena reviewed, challenges remain, especially with regard to ad hoc categories.

Redefining the resolution of semantic knowledge in the brain: Advances made by the introduction of models of semantics in neuroimaging

The boundaries of our understanding of conceptual representation in the brain have been redrawn since the introduction of explicit models of semantics. These models are grounded in vast behavioural datasets acquired in healthy volunteers. Here, we review the most important techniques which have been applied to detect semantic information in neuroimaging data and argue why semantic models are possibly the most valuable addition to the research of semantics in recent years. Using multivariate analysis, predictions based on patient lesion data have been confirmed during semantic processing in healthy controls. Secondly, this new method has given rise to new research avenues, e.g. the detection of semantic processing outside of the temporal cortex. As a future line of work, the same research strategy could be useful to study neurological conditions such as the semantic variant of primary progressive aphasia, which is characterized by pathological semantic processing.

Can prediction-based distributional semantic models predict typicality?

Recent advances in the field of computational linguistics have led to the development of various prediction-based models of semantics. These models seek to infer word representations from large text collections by predicting target words from neighbouring words (or vice versa). The resulting representations are vectors in a continuous space, collectively called word embeddings. Although psychological plausibility was not a primary concern for the developers of predictive models, it has been the topic of several recent studies in the field of psycholinguistics. That is, word embeddings have been linked to similarity ratings, word associations, semantic priming, word recognition latencies, and so on. Here, we build on this work by investigating category structure. Throughout seven experiments, we sought to predict human typicality judgements from two languages, Dutch and English, using different semantic spaces. More specifically, we extracted a number of predictor variables, and evaluated how well they could capture the typicality gradient of common categories (e.g., birds, fruit, vehicles, etc.). Overall, the performance of predictive models was rather modest and did not compare favourably with that of an older count-based model. These results are somewhat disappointing given the enthusiasm surrounding predictive models. Possible explanations and future directions are discussed.

Brain reading and behavioral methods provide complementary perspectives on the representation of concepts

The advent of brain reading techniques has enabled new approaches to the study of concept representation, based on the analysis of multivoxel activation patterns evoked by the contemplation of individual concepts such as animal concepts. The present fMRI study characterized the representation of 30 animal concepts. Dimensionality reduction of the multivoxel activation patterns underlying the individual animal concepts indicated that the semantic building blocks of the brain's representations of the animals corresponded to intrinsic animal properties (e.g. fierceness, intelligence, size). These findings were compared to behavioral studies of concept representation, which have typically collected pairwise similarity ratings between two concepts (e.g. Henley, 1969). Behavioral similarity judgments, by contrast, indicated that the animals were organized into taxonomically defined groups (e.g. canine, feline, equine). The difference in the results between the brain reading and behavioral approaches might derive from differences in cognitive processing during judging similarities versus contemplating one animal at a time. Brain reading approaches may have an advantage in describing thoughts about an individual concept, owing to the ability to decode brain activation patterns elicited by the brief consideration of a single concept (e.g. word reading) without a complex cognitive or behavioral task (e.g. similarity judgments). On the other hand, some behavioral tasks may tend to evoke a concept from numerous perspectives, yielding a representation of the breadth and sophistication of the concept knowledge. These results suggest that neural and behavioral measures offer complementary perspectives that together characterize the content and structure of concept representations.

Exploring the Organization of Semantic Memory through Unsupervised Analysis of Event-related Potentials

Modern multivariate methods have enabled the application of unsupervised techniques to analyze neurophysiological data without strict adherence to predefined experimental conditions. We demonstrate a multivariate method that leverages priming effects on the evoked potential to perform hierarchical clustering on a set of word stimuli. The current study focuses on the semantic relationships that play a key role in the organization of our mental lexicon of words and concepts. The N400 component of the event-related potential is considered a reliable neurophysiological response that is indicative of whether accessing one concept facilitates subsequent access to another (i.e., one "primes" the other). To further our understanding of the organization of the human mental lexicon, we propose to utilize the N400 component to drive a clustering algorithm that can uncover, given a set of words, which particular subsets of words show mutual priming. Such a scheme requires a reliable measurement of the amplitude of the N400 component without averaging across many trials, which was here achieved using a recently developed multivariate analysis method based on beamforming. We validated our method by demonstrating that it can reliably detect, without any prior information about the nature of the stimuli, a well-known feature of the organization of our semantic memory: the distinction between animate and inanimate concepts. These results motivate further application of our method to data-driven exploration of disputed or unknown relationships between stimuli.

Representation of Semantic Similarity in the Left Intraparietal Sulcus: Functional Magnetic Resonance Imaging Evidence

According to a recent study, semantic similarity between concrete entities correlates with the similarity of activity patterns in left middle IPS during category naming. We examined the replicability of this effect under passive viewing conditions, the potential role of visuoperceptual similarity, where the effect is situated compared to regions that have been previously implicated in visuospatial attention, and how it compares to effects of object identity and location. Forty-six subjects participated. Subjects passively viewed pictures from two categories, musical instruments and vehicles. Semantic similarity between entities was estimated based on a concept-feature matrix obtained in more than 1,000 subjects. Visuoperceptual similarity was modeled based on the HMAX model, the AlexNet deep convolutional learning model, and thirdly, based on subjective visuoperceptual similarity ratings. Among the IPS regions examined, only left middle IPS showed a semantic similarity effect. The effect was significant in hIP1, hIP2, and hIP3. Visuoperceptual similarity did not correlate with similarity of activity patterns in left middle IPS. The semantic similarity effect in left middle IPS was significantly stronger than in the right middle IPS and also stronger than in the left or right posterior IPS. The semantic similarity effect was similar to that seen in the angular gyrus. Object identity effects were much more widespread across nearly all parietal areas examined. Location effects were relatively specific for posterior IPS and area 7 bilaterally. To conclude, the current findings replicate the semantic similarity effect in left middle IPS under passive viewing conditions, and demonstrate its anatomical specificity within a cytoarchitectonic reference frame. We propose that the semantic similarity effect in left middle IPS reflects the transient uploading of semantic representations in working memory.

Feature taxonomy: What type of features do children associate with categories and how do they fare in predicting category judgments?

The present study investigates category intension in school-aged children and adults at two different levels of abstraction (i.e., superordinate and basic level) for two category types (i.e., artefacts and natural kinds). We addressed two critical questions: what kind of features do children and adults generate to define semantic categories and which features predict category membership judgment best at each abstraction level? Overall, participants generated relatively more entity features for natural kinds categories, compared to artefact categories, as well as for basic level categories, compared to superordinate categories. Furthermore, the results showed that older children and adults generated relatively more entity features than younger children. Finally, situation features play the most important role in the prediction of category judgments at both levels of abstraction. Theoretical implications and comparable results from previous studies are described in detail.

Dutch taboo norms

This article provides norms for general taboo, personal taboo, insult, valence, and arousal for 672 Dutch words, including 202 taboo words. Norms were collected using a 7-point Likert scale and based on ratings by psychology students from the Erasmus University Rotterdam in The Netherlands. The sample consisted of 87 psychology students (58 females, 29 males). We obtained high reliability based on split-half analyses. Our norms show high correlations with arousal and valence ratings collected by another Dutch word-norms study (Moors et al.,, Behavior Research Methods, 45, 169–177, 2013). Our results show that the previously found quadratic relation (i.e., U-shaped pattern) between valence and arousal also holds when only taboo words are considered. Additionally, words rated high on taboo tended to be rated low on valence, but some words related to sex rated high on both taboo and valence. Words that rated high on taboo rated high on insult, again with the exception of words related to sex many of which rated low on insult. Finally, words rated high on taboo and insult rated high on arousal. The Dutch Taboo Norms (DTN) database is a useful tool for researchers interested in the effects of taboo words on cognitive processing. The data associated with this paper can be accessed via the Open Science Framework (https://osf.io/vk782/).

The Typicality Ranking Task: A New Method to Derive Typicality Judgments from Children

An alternative method for deriving typicality judgments, applicable in young children that are not familiar with numerical values yet, is introduced, allowing researchers to study gradedness at younger ages in concept development. Contrary to the long tradition of using rating-based procedures to derive typicality judgments, we propose a method that is based on typicality ranking rather than rating, in which items are gradually sorted according to their typicality, and that requires a minimum of linguistic knowledge. The validity of the method is investigated and the method is compared to the traditional typicality rating measurement in a large empirical study with eight different semantic concepts. The results show that the typicality ranking task can be used to assess children's category knowledge and to evaluate how this knowledge evolves over time. Contrary to earlier held assumptions in studies on typicality in young children, our results also show that preference is not so much a confounding variable to be avoided, but that both variables are often significantly correlated in older children and even in adults.

Semantic similarity: normative ratings for 185 Spanish noun triplets

The present study introduces the first Spanish database with normative ratings of semantic similarity for 185 word triplets. Each word triplet is constituted by a target word (e.g., guisante [pea]) and two semantically related and nonassociatively related words: a word highly related in meaning to the target (e.g., judía [bean]), and a word less related in meaning to the target (e.g., patata [potato]). The degree of meaning similarity was assessed by 332 participants by using a semantic similarity rating task on a 9-point scale. Pairs having a value of semantic similarity ranging from 5 to 9 were classified as being more semantically related, whereas those with values ranging from 2 to 4.99 were considered as being less semantically related. The relative distance between the two pairs for the same target ranged from 0.48 to 5.07 points. Mean comparisons revealed that participants rated the more similar words as being significantly more similar in meaning to the target word than were the less similar words. In addition to the semantic similarity norms, values of concreteness and familiarity of each word in a triplet are provided. The present database can be a very useful tool for scientists interested in designing experiments to examine the role of semantics in language processing. Since the variable of semantic similarity includes a wide range of values, it can be used as either a continuous or a dichotomous variable. The full database is available in the supplementary materials.

Using the speeded word fragment completion task to examine semantic priming

The present research investigates semantic priming with an adapted version of the word fragment completion task. In this task, which we refer to as the speeded word fragment completion task, participants need to complete words such as lett_ce (lettuce), from which one letter was omitted, as quickly as possible. This paradigm has some interesting qualities in comparison with the traditionally used lexical decision task. That is, it requires no pseudowords, it is more engaging for participants, and most importantly, it allows for a more fine-grained investigation of semantic activation. In two studies, we found that words were completed faster when the preceding trial comprised a semantically related fragment such as tom_to (tomato) than when it comprised an unrelated fragment such as guit_r (guitar). A third experiment involved a lexical decision task, to compare both paradigms. The results showed that the magnitude of the priming effect was similar, but item-level priming effects were inconsistent over tasks. Crucially, the speeded word fragment completion task obtained strong priming effects for highly frequent, central words, such as work, money, and warm, whereas the lexical decision task did not. In a final experiment featuring only short, highly frequent words, the lexical decision task failed to find a priming effect, whereas the fragment completion task did obtain a robust effect. Taken together, these results suggest that the speeded word fragment completion task may prove a viable alternative for examining semantic priming.

Masked translation priming asymmetry in Chinese-English bilinguals: Making sense of the Sense Model

Masked translation priming asymmetry is the robust finding that priming from a bilingual's first language (L1) to their second language (L2) is stronger than priming from L2 to L1. This asymmetry has been claimed to be task dependent. The Sense Model proposed by Finkbeiner, Forster, Nicol, and Nakamura (2004) claims that the asymmetry is reduced in semantic categorization relative to lexical decision due to a category filtering mechanism that limits the features considered in categorization decisions to dominant, category-relevant features. This paper reports two pairs of semantic categorization and lexical decision tasks designed to test the Sense Model's predictions. The experiments replicated the finding of Finkbeiner et al. that L2-L1 priming is somewhat stronger in semantic categorization than lexical decision, selectively for exemplars of the category. However, the direct comparison of L2-L1 and L1-L2 translation priming across tasks failed to confirm the Sense Model's central prediction that translation priming asymmetry is significantly reduced in semantic categorization. The data therefore fail to support the category filtering account of translation priming asymmetry. Rather, they suggest that pre-activation of conceptual features of the target category provides feedback to lexical forms that compensates for the weak connections between the lexical and conceptual representations of L2 words.

Problems of reliability and validity with similarity derived from category fluency

This study aims to assess the reliability and the validity of exemplar similarity derived from category fluency tasks. A homogeneous sample of 21 healthy participants completed a category fluency task twice with an interval of one week. They also rated pairs comprised of the most frequently generated exemplars in terms of similarity. Similarities were derived from the fluency data by determining the average distance between generated exemplars and correcting it for repetitions and response sequence length. We calculated the correlation between the similarities derived from the two sessions of the fluency task and between the derived similarities and the directly rated similarities. Spatial representations of the similarities were constructed using multidimensional scaling to visualize the differences between both sessions of the fluency task and the pairwise rating task. We find that the derived similarities are not stable in time and show little correspondence with directly rated similarities. The differences between similarities derived from category fluency tasks in healthy participants, indicate that similar differences between healthy controls and patients with mental disorders, do not necessarily point to a semantic impairment of the latter, but rather reflect the unreliability of the data.

Noun and knowledge retrieval for biological and non-biological entities following right occipitotemporal lesions

We investigated the critical contribution of right ventral occipitotemporal cortex to knowledge of visual and functional-associative attributes of biological and non-biological entities and how this relates to category-specificity during confrontation naming. In a consecutive series of 7 patients with lesions confined to right ventral occipitotemporal cortex, we conducted an extensive assessment of oral generation of visual-sensory and functional-associative features in response to the names of biological and nonbiological entities. Subjects also performed a confrontation naming task for these categories. Our main novel finding related to a unique case with a small lesion confined to right medial fusiform gyrus who showed disproportionate naming impairment for nonbiological versus biological entities, specifically for tools. Generation of visual and functional-associative features was preserved for biological and non-biological entities. In two other cases, who had a relatively small posterior lesion restricted to primary visual and posterior fusiform cortex, retrieval of visual attributes was disproportionately impaired compared to functional-associative attributes, in particular for biological entities. However, these cases did not show a category-specific naming deficit. Two final cases with the largest lesions showed a classical dissociation between biological versus nonbiological entities during naming, with normal feature generation performance. This is the first lesion-based evidence of a critical contribution of the right medial fusiform cortex to tool naming. Second, dissociations along the dimension of attribute type during feature generation do not co-occur with category-specificity during naming in the current patient sample.

Acute psychosocial stress and children's memory

We investigated whether children's performance on working memory (WM) and delayed retrieval (DR) tasks decreased after stress exposure, and how physiological stress responses related to performance under stress. About 158 children (83 girls; Mage = 10.61 years, SD = 0.52) performed two WM tasks (WM forward and WM backward) and a DR memory task first during a control condition, and 1 week later during a stress challenge. Salivary alpha-amylase (sAA) and cortisol were assessed during the challenge. Only WM backward performance declined over conditions. Correlations between physiological stress responses and performance within the stress challenge were present only for WM forward and DR. For WM forward, higher cortisol responses were related to better performance. For DR, there was an inverted U-shape relation between cortisol responses and performance, as well as a cortisol × sAA interaction, with concurrent high or low responses related to optimal performance. This emphasizes the importance of including curvilinear and interaction effects when relating physiology to memory.

Similarity of fMRI activity patterns in left perirhinal cortex reflects semantic similarity between words

How verbal and nonverbal visuoperceptual input connects to semantic knowledge is a core question in visual and cognitive neuroscience, with significant clinical ramifications. In an event-related functional magnetic resonance imaging (fMRI) experiment we determined how cosine similarity between fMRI response patterns to concrete words and pictures reflects semantic clustering and semantic distances between the represented entities within a single category. Semantic clustering and semantic distances between 24 animate entities were derived from a concept-feature matrix based on feature generation by >1000 subjects. In the main fMRI study, 19 human subjects performed a property verification task with written words and pictures and a low-level control task. The univariate contrast between the semantic and the control task yielded extensive bilateral occipitotemporal activation from posterior cingulate to anteromedial temporal cortex. Entities belonging to a same semantic cluster elicited more similar fMRI activity patterns in left occipitotemporal cortex. When words and pictures were analyzed separately, the effect reached significance only for words. The semantic similarity effect for words was localized to left perirhinal cortex. According to a representational similarity analysis of left perirhinal responses, semantic distances between entities correlated inversely with cosine similarities between fMRI response patterns to written words. An independent replication study in 16 novel subjects confirmed these novel findings. Semantic similarity is reflected by similarity of functional topography at a fine-grained level in left perirhinal cortex. The word specificity excludes perceptually driven confounds as an explanation and is likely to be task dependent.

Aligning Mental Representations

This work introduces a framework that implements asymmetric communication theory proposed by Sperber and Wilson [1]. The framework applies a generalization model known as the Bayesian model of generalization (BMG) [2] for aligning knowledge possessed by two communicating parties. The work focuses on the application of the BMG to publicly available datasets, the Leuven natural concept database [3] representing semantic structures of domain knowledge possessed by individual subjects [3]. Results indicate that the BMG is potentially a model applicable to simulating the alignment of domain knowledge from the information receiver’s viewpoint.

Right fusiform response patterns reflect visual object identity rather than semantic similarity

We previously reported the neuropsychological consequences of a lesion confined to the middle and posterior part of the right fusiform gyrus (case JA) causing a partial loss of knowledge of visual attributes of concrete entities in the absence of category-selectivity (animate versus inanimate). We interpreted this in the context of a two-step model that distinguishes structural description knowledge from associative-semantic processing and implicated the lesioned area in the former process. To test this hypothesis in the intact brain, multi-voxel pattern analysis was used in a series of event-related fMRI studies in a total of 46 healthy subjects. We predicted that activity patterns in this region would be determined by the identity of rather than the conceptual similarity between concrete entities. In a prior behavioral experiment features were generated for each entity by more than 1000 subjects. Based on a hierarchical clustering analysis the entities were organised into 3 semantic clusters (musical instruments, vehicles, tools). Entities were presented as words or pictures. With foveal presentation of pictures, cosine similarity between fMRI response patterns in right fusiform cortex appeared to reflect both the identity of and the semantic similarity between the entities. No such effects were found for words in this region. The effect of object identity was invariant for location, scaling, orientation axis and color (grayscale versus color). It also persisted for different exemplars referring to a same concrete entity. The apparent semantic similarity effect however was not invariant. This study provides further support for a neurobiological distinction between structural description knowledge and processing of semantic relationships and confirms the role of right mid-posterior fusiform cortex in the former process, in accordance with previous lesion evidence.