Opposing effects of semantic diversity in lexical and semantic relatedness decisions

Diversity of narrative context disrupts the early stage of learning the meanings of novel words

High quality lexical representations develop through repeated exposures to words in different contexts. This preregistered experiment investigated how diversity of narrative context affects the earliest stages of word learning via reading. Adults (N = 100) learned invented meanings for eight pseudowords, which each occurred in five written paragraphs either within a single coherent narrative context or five different narrative contexts. The words' semantic features were controlled across conditions to avoid influences from polysemy (lexical ambiguity). Posttests included graded measures of word-form recall (spelling accuracy) and recognition (multiple choice), and word-meaning recall (number of semantic features). Diversity of narrative context did not affect word-form learning, but more semantic features were correctly recalled for words trained in a single context. These findings indicate that learning the meanings of novel words is initially boosted by anchoring them to a single coherent narrative discourse.

No evidence for high inflexible precision of prediction errors in autism during lexical processing

Research has shown that information processing differences associated with autism could impact on language and literacy development. This study tested an approach to autistic cognition that suggests learning occurs via prediction errors, and autistic people have very precise and inflexible predictions that result in more sensitivity to meaningless signal errors than non-autistic readers. We used this theoretical background to investigate whether differences in prediction coding influence how orthographic (Experiment 1) and semantic information (Experiment 2) is processed by autistic readers. Experiment 1 used a lexical decision task to test whether letter position information was processed less flexibly by autistic than non-autistic readers. Three types of letter strings: words, transposed letter and substituted letters nonwords were presented. Experiment 2 used a semantic relatedness task to test whether autistic readers processed words with high and low semantic diversity differently to non-autistic readers. Results showed similar transposed letter and semantic diversity effects for all readers; indicating that orthographic and semantic information are processed similarly by autistic and non-autistic readers; and therefore, differences in prediction coding were not evident for these lexical processing tasks.

Semantic diversity is best measured with unscaled vectors: Reply to Cevoli, Watkins and Rastle (2020)

Semantic diversity refers to the degree of semantic variability in the contexts in which a particular word is used. We have previously proposed a method for measuring semantic diversity based on latent semantic analysis (LSA). In a recent paper, Cevoli et al. (2020) attempted to replicate our method and obtained different semantic diversity values. They suggested that this discrepancy occurred because they scaled their LSA vectors by their singular values, while we did not. Using their new results, they argued that semantic diversity is not related to ambiguity in word meaning, as we originally proposed. In this reply, we demonstrate that the use of unscaled vectors provides better fits to human semantic judgements than scaled ones. Thus we argue that our original semantic diversity measure should be preferred over the Cevoli et al. version. We replicate Cevoli et al.'s analysis using the original semantic diversity measure and find (a) our original measure is a better predictor of word recognition latencies than the Cevoli et al. equivalent and (b) that, unlike Cevoli et al.'s measure, our semantic diversity is reliably associated with a measure of polysemy based on dictionary definitions. We conclude that the Hoffman et al. semantic diversity measure is better-suited to capturing the contextual variability among words and that words appearing in a more diverse set of contexts have more variable semantic representations. However, we found that homonyms did not have higher semantic diversity values than non-homonyms, suggesting that the measure does not capture this special case of ambiguity.

Distributional Measures of Semantic Abstraction

This article provides an in-depth study of distributional measures for distinguishing between degrees of semantic abstraction. Abstraction is considered a “central construct in cognitive science” and a “process of information reduction that allows for efficient storage and retrieval of central knowledge”. Relying on the distributional hypothesis, computational studies have successfully exploited measures of contextual co-occurrence and neighbourhood density to distinguish between conceptual semantic categorisations. So far, these studies have modeled semantic abstraction across lexical-semantic tasks such as ambiguity; diachronic meaning changes; abstractness vs. concreteness; and hypernymy. Yet, the distributional approaches target different conceptual types of semantic relatedness, and as to our knowledge not much attention has been paid to apply, compare or analyse the computational abstraction measures across conceptual tasks. The current article suggests a novel perspective that exploits variants of distributional measures to investigate semantic abstraction in English in terms of the abstract–concrete dichotomy (e.g., glory–banana) and in terms of the generality–specificity distinction (e.g., animal–fish), in order to compare the strengths and weaknesses of the measures regarding categorisations of abstraction, and to determine and investigate conceptual differences. In a series of experiments we identify reliable distributional measures for both instantiations of lexical-semantic abstraction and reach a precision higher than 0.7, but the measures clearly differ for the abstract–concrete vs. abstract–specific distinctions and for nouns vs. verbs. Overall, we identify two groups of measures, (i) frequency and word entropy when distinguishing between more and less abstract words in terms of the generality–specificity distinction, and (ii) neighbourhood density variants (especially target–context diversity) when distinguishing between more and less abstract words in terms of the abstract–concrete dichotomy. We conclude that more general words are used more often and are less surprising than more specific words, and that abstract words establish themselves empirically in semantically more diverse contexts than concrete words. Finally, our experiments once more point out that distributional models of conceptual categorisations need to take word classes and ambiguity into account: results for nouns vs. verbs differ in many respects, and ambiguity hinders fine-tuning empirical observations.

What is semantic diversity and why does it facilitate visual word recognition?

Previous research has speculated that semantic diversity and lexical ambiguity may be closely related constructs. Our research sought to test this claim in respect of the semantic diversity measure proposed by Hoffman et al. (2013). To this end, we replicated the procedure described by Hoffman et al., Behavior Research Methods, 45(3), 718–730 (2013) for computing multidimensional representations of contextual information using Latent Semantic Analysis, and from these we derived semantic diversity values for 28,555 words. We then replicated the facilitatory effect of semantic diversity on word recognition using existing data resources and observed this effect to be greater for low-frequency words. Yet, we found no relationship between this measure and lexical ambiguity effects in word recognition. Further analysis of the LSA-based contextual representations used to compute Hoffman et al. (2013) measure of semantic diversity revealed that they do not capture the distinct meanings of ambiguous words. Instead, these contextual representations appear to capture general information about the topics and types of written material in which words occur. These analyses suggest that the semantic diversity metric previously proposed by Hoffman et al. (2013) facilitates word recognition because high-diversity words are likely to have been encountered no matter what one has read, whereas many participants may not have encountered lower-diversity words simply because the topics and types of written material in which they occur are more restricted.

Evidence for preferential attachment: Words that are more well connected in semantic networks are better at acquiring new links in paired-associate learning

Here, we view the mental lexicon as a semantic network where words are connected if they are semantically related. Steyvers and Tenenbaum (Cognitive Science, 29, 41–78, 2005) proposed that the growth of semantic networks follows preferential attachment, the observation that new nodes are more likely to connect to preexisting nodes that are more well connected (i.e., the rich get richer). If this is the case, well-connected known words should be better at acquiring new links than poorly connected words. We tested this prediction in three paired-associate learning (PAL) experiments in which participants memorized arbitrary cue–response word pairs. We manipulated the semantic connectivity of the cue words, indexed by the words’ free associative degree centrality. Experiment 1 is a reanalysis of the PAL data from Qiu and Johns (Psychonomic Bulletin & Review, 27, 114–121, 2020), in which young adults remembered 40 cue–response word pairs (e.g., nature–chain) and completed a cued recall task. Experiment 2 is a preregistered replication of Qiu and Johns. Experiment 3 addressed some limitations in Qiu and Johns’s design by using pseudowords as the response items (e.g., boot–arruity). The three experiments converged to show that cue words of higher degree centrality facilitated the recall/recognition of the response items, providing support for the notion that better-connected words have a greater ability to acquire new links (i.e., the rich do get richer). Importantly, while degree centrality consistently accounted for significant portions of variance in PAL accuracy, other psycholinguistic variables (e.g., concreteness, contextual diversity) did not, suggesting that degree centrality is a distinct variable that affects the ease of verbal associative learning.

Barking up the right tree: Univariate and multivariate fMRI analyses of homonym comprehension

Homonyms are a critical test case for investigating how the brain resolves ambiguity in language and, more generally, how context influences semantic processing. Previous neuroimaging studies have associated processing of homonyms with greater engagement of regions involved in executive control of semantic processing. However, the precise role of these areas and the involvement of semantic representational regions in homonym comprehension remain elusive. We addressed this by combining univariate and multivariate fMRI analyses of homonym processing. We tested whether multi-voxel activation patterns could discriminate between presentations of the same homonym in different contexts (e.g., bark following tree vs. bark following dog). The ventral anterior temporal lobe, implicated in semantic representation but not previously in homonym comprehension, showed this meaning-specific coding, despite not showing increased mean activation for homonyms. Within inferior frontal gyrus (IFG), a key site for semantic control, there was a dissociation between pars orbitalis, which also showed meaning-specific coding, and pars triangularis, which discriminated more generally between semantically related and unrelated word pairs. IFG effects were goal-dependent, only occurring when the task required semantic decisions, in line with a top-down control function. Finally, posterior middle temporal cortex showed a hybrid pattern of responses, supporting the idea that it acts as an interface between semantic representations and the control system. The study provides new evidence for context-dependent coding in the semantic system and clarifies the role of control regions in processing ambiguity. It also highlights the importance of combining univariate and multivariate neuroimaging data to fully elucidate the role of a brain region in semantic cognition.

Working memory load affects early affective responses to concrete and abstract words differently: Evidence from ERPs

Early posterior negativity (EPN) is an early-occurring, event-related, potential that is elicited by pictures and words that have highly arousing characteristics. Whilst EPN has been found with words presented in isolation several times, different types of words have shown quite different effects across different types of tasks. One possible reason for this is that memory and attentional demands may affect the way semantic features of words are processed, and this may modulate EPN. This was investigated in a silent reading task using abstract and concrete words of negative and neutral valence and a dual phonological working memory task to manipulate memory load. The results showed that abstract but not concrete words elicited EPN, and this may have affected downstream processing. Further analyses examining alpha desynchronization showed that negative concrete words appeared to be significantly affected by the memory load manipulation, unlike negative abstract words. These results provide evidence that the processing of features in negative concrete words is more affected by working memory and attentional demands than the processing of features in negative abstract words, and this may be responsible for the failure of negative concrete words to elicit EPN in this study. Thus, the extent to which words elicit EPN appears to be dependent on both their semantic representations and competing cognitive processes. These results provide a potential explanation for some of the differences that have been reported in previous experiments as well as insight into how memory and attention can affect the processing of the semantic features of words.

Divergent effects of healthy ageing on semantic knowledge and control: Evidence from novel comparisons with semantically impaired patients

Effective use of semantic knowledge requires a set of conceptual representations and control processes which ensure that currently relevant aspects of this knowledge are retrieved and selected. It is well-established that levels of semantic knowledge increase across the lifespan. However, the effects of ageing on semantic control processes have not been assessed. I addressed this issue by comparing the performance profiles of young and older people on a verbal comprehension test. Two sets of variables were used to predict accuracy and RT in each group: (1) the psycholinguistic properties of words probed in each trial and (2) the performance on each trial by two groups of semantically impaired neuropsychological patients. Young people demonstrated poor performance for low-frequency and abstract words, suggesting that they had difficulty processing words with intrinsically weak semantic representations. Indeed, performance in this group was strongly predicted by the performance of patients with semantic dementia, who suffer from degradation of semantic knowledge. In contrast, older adults performed poorly on trials where the target semantic relationship was weak and distractor relationships strong - conditions which require high levels of controlled processing. Their performance was not predicted by the performance of semantic dementia patients, but was predicted by the performance of patients with semantic control deficits. These findings indicate that the effects of ageing on semantic cognition are more complex than has previously been assumed. While older people have larger stores of knowledge than young people, they appear to be less skilled at exercising control over the activation of this knowledge.

Concepts, control, and context: A connectionist account of normal and disordered semantic cognition

Semantic cognition requires conceptual representations shaped by verbal and nonverbal experience and executive control processes that regulate activation of knowledge to meet current situational demands. A complete model must also account for the representation of concrete and abstract words, of taxonomic and associative relationships, and for the role of context in shaping meaning. We present the first major attempt to assimilate all of these elements within a unified, implemented computational framework. Our model combines a hub-and-spoke architecture with a buffer that allows its state to be influenced by prior context. This hybrid structure integrates the view, from cognitive neuroscience, that concepts are grounded in sensory-motor representation with the view, from computational linguistics, that knowledge is shaped by patterns of lexical co-occurrence. The model successfully codes knowledge for abstract and concrete words, associative and taxonomic relationships, and the multiple meanings of homonyms, within a single representational space. Knowledge of abstract words is acquired through (a) their patterns of co-occurrence with other words and (b) acquired embodiment, whereby they become indirectly associated with the perceptual features of co-occurring concrete words. The model accounts for executive influences on semantics by including a controlled retrieval mechanism that provides top-down input to amplify weak semantic relationships. The representational and control elements of the model can be damaged independently, and the consequences of such damage closely replicate effects seen in neuropsychological patients with loss of semantic representation versus control processes. Thus, the model provides a wide-ranging and neurally plausible account of normal and impaired semantic cognition.

Semantic ambiguity effects on traditional Chinese character naming: A corpus-based approach

Words are considered semantically ambiguous if they have more than one meaning and can be used in multiple contexts. A number of recent studies have provided objective ambiguity measures by using a corpus-based approach and have demonstrated ambiguity advantages in both naming and lexical decision tasks. Although the predictive power of objective ambiguity measures has been examined in several alphabetic language systems, the effects in logographic languages remain unclear. Moreover, most ambiguity measures do not explicitly address how the various contexts associated with a given word relate to each other. To explore these issues, we computed the contextual diversity (Adelman, Brown, & Quesada, Psychological Science, 17; 814-823, 2006) and semantic ambiguity (Hoffman, Lambon Ralph, & Rogers, Behavior Research Methods, 45; 718-730, 2013) of traditional Chinese single-character words based on the Academia Sinica Balanced Corpus, where contextual diversity was used to evaluate the present semantic space. We then derived a novel ambiguity measure, namely semantic variability, by computing the distance properties of the distinct clusters grouped by the contexts that contained a given word. We demonstrated that semantic variability was superior to semantic diversity in accounting for the variance in naming response times, suggesting that considering the substructure of the various contexts associated with a given word can provide a relatively fine scale of ambiguity information for a word. All of the context and ambiguity measures for 2,418 Chinese single-character words are provided as supplementary materials.

Modeling the Structure and Dynamics of Semantic Processing

The contents and structure of semantic memory have been the focus of much recent research, with major advances in the development of distributional models, which use word co-occurrence information as a window into the semantics of language. In parallel, connectionist modeling has extended our knowledge of the processes engaged in semantic activation. However, these two lines of investigation have rarely been brought together. Here, we describe a processing model based on distributional semantics in which activation spreads throughout a semantic network, as dictated by the patterns of semantic similarity between words. We show that the activation profile of the network, measured at various time points, can successfully account for response times in lexical and semantic decision tasks, as well as for subjective concreteness and imageability ratings. We also show that the dynamics of the network is predictive of performance in relational semantic tasks, such as similarity/relatedness rating. Our results indicate that bringing together distributional semantic networks and spreading of activation provides a good fit to both automatic lexical processing (as indexed by lexical and semantic decisions) as well as more deliberate processing (as indexed by ratings), above and beyond what has been reported for previous models that take into account only similarity resulting from network structure.

Examining the role of context variability in memory for items and associations

What properties of a word make it easy or difficult to remember? Word frequency and context variability are separate, closely related word properties that have disparate influences on memorability. The influence of word frequency changes depending on the memory task, with high-frequency words tending to be recalled better and low-frequency words to be recognized better. Conversely, low-context-variability words tend to be remembered better across tasks. One proposed explanation for the low-variability advantage is that low-variability words are easier to associate with the experimental context, given that they are associated with fewer extra-experimental contexts. On the basis of this explanation, it has been suggested that the formation of interitem associations during encoding should interfere with the formation of item-to-context associations, attenuating the low-variability advantage. Across experiments, we tested whether focusing on interitem associations disrupted the low-variability advantage, by manipulating encoding tasks, test expectancy, final test condition, word frequency, and context variability. Focusing on interitem associations did not harm performance for low-variability words. Words low in both frequency and variability were recognized better, but word pairs composed of high-frequency, low-variability words were recognized better in associative recognition. On the basis of the data, we suggest that focusing on interitem associations does not come at the expense of item-to-context associations. Moreover, the data further support the idea that frequency and variability are distinct properties.

The influence of contextual diversity on word learning

In a series of analyses over mega datasets, Jones, Johns, and Recchia (Canadian Journal of Experimental Psychology, 66(2), 115-124, 2012) and Johns et al. (Journal of the Acoustical Society of America, 132:2, EL74-EL80, 2012) found that a measure of contextual diversity that takes into account the semantic variability of a word's contexts provided a better fit to both visual and spoken word recognition data than traditional measures, such as word frequency or raw context counts. This measure was empirically validated with an artificial language experiment (Jones et al.). The present study extends the empirical results with a unique natural language learning paradigm, which allows for an examination of the semantic representations that are acquired as semantic diversity is varied. Subjects were incidentally exposed to novel words as they rated short selections from articles, books, and newspapers. When novel words were encountered across distinct discourse contexts, subjects were both faster and more accurate at recognizing them than when they were seen in redundant contexts. However, learning across redundant contexts promoted the development of more stable semantic representations. These findings are predicted by a distributional learning model trained on the same materials as our subjects.

The Paradox of Abstraction: Precision Versus Concreteness

We introduce a novel measure of abstractness based on the amount of information of a concept computed from its position in a semantic taxonomy. We refer to this measure as precision. We propose two alternative ways to measure precision, one based on the path length from a concept to the root of the taxonomic tree, and another one based on the number of direct and indirect descendants. Since more information implies greater processing load, we hypothesize that nouns higher in precision will have a processing disadvantage in a lexical decision task. We contrast precision to concreteness, a common measure of abstractness based on the proportion of sensory-based information associated with a concept. Since concreteness facilitates cognitive processing, we predict that while both concreteness and precision are measures of abstractness, they will have opposite effects on performance. In two studies we found empirical support for our hypothesis. Precision and concreteness had opposite effects on latency and accuracy in a lexical decision task, and these opposite effects were observable while controlling for word length, word frequency, affective content and semantic diversity. Our results support the view that concepts organization includes amodal semantic structures which are independent of sensory information. They also suggest that we should distinguish between sensory-based and amount-of-information-based abstractness.

Nurturing a lexical legacy: reading experience is critical for the development of word reading skill

The scientific study of reading has taught us much about the beginnings of reading in childhood, with clear evidence that the gateway to reading opens when children are able to decode, or 'sound out' written words. Similarly, there is a large evidence base charting the cognitive processes that characterise skilled word recognition in adults. Less understood is how children develop word reading expertise. Once basic reading skills are in place, what factors are critical for children to move from novice to expert? This paper outlines the role of reading experience in this transition. Encountering individual words in text provides opportunities for children to refine their knowledge about how spelling represents spoken language. Alongside this, however, reading experience provides much more than repeated exposure to individual words in isolation. According to the lexical legacy perspective, outlined in this paper, experiencing words in diverse and meaningful language environments is critical for the development of word reading skill. At its heart is the idea that reading provides exposure to words in many different contexts, episodes and experiences which, over time, sum to a rich and nuanced database about their lexical history within an individual's experience. These rich and diverse encounters bring about local variation at the word level: a lexical legacy that is measurable during word reading behaviour, even in skilled adults.

Semantic Richness Effects in Syntactic Classification: The Role of Feedback

Words with richer semantic representations are recognized faster across a range of lexical processing tasks. The most influential account of this finding is based on the idea that semantic richness effects are mediated by feedback from semantic-level to lower-level representations. In an earlier lexical decision study, Yap et al. (2015) tested this claim by examining the joint effects of stimulus quality and four semantic richness dimensions (imageability, number of features, semantic neighborhood density, semantic diversity). The results of that study showed that joint effects of stimulus quality and richness were generally additive, consistent with the idea that semantic feedback does not typically reach the earliest levels of representation in lexical decision. The present study extends this earlier work by investigating the joint effects of stimulus quality and the same four semantic richness dimensions on syntactic classification performance (is this a noun or verb?), which places relatively more emphasis on semantic processing. Additive effects of stimulus quality and richness were found for two of the four targeted dimensions (concreteness, number of features) while semantic neighborhood density and semantic diversity did not seem to influence syntactic classification response times. These findings provide further support against the view that semantic information reaches early letter-level processes.

It's all in the delivery: Effects of context valence, arousal, and concreteness on visual word processing

Prior research has examined how distributional properties of contexts (number of unique contexts or their informativeness) influence the effort of word recognition. These properties do not directly interrogate the semantic properties of contexts. We evaluated the influence of average concreteness, valence (positivity) and arousal of the contexts in which a word occurs on response times in the lexical decision task, age of acquisition of the word, and word recognition memory performance. Using large corpora and norming mega-studies we quantified semantics of contexts for thousands of words and demonstrated that contextual factors were predictive of lexical representation and processing above and beyond the influence shown by concreteness, valence and arousal of the word itself. Our findings indicate that lexical representations are influenced not only by how diverse the word's contexts are, but also by the embodied experiences they elicit.

Is More Always Better for Verbs? Semantic Richness Effects and Verb Meaning

We examined how several semantic richness variables contribute to verb meaning, across a number of tasks. Because verbs can vary in tense, and the manner in which tense is coded (i.e., regularity), we also examined how these factors moderated the effects of semantic richness. In Experiment 1 we found that age of acquisition (AoA), valence, arousal and embodiment predicted faster response times in LDT. In Experiment 2 we examined a particular semantic richness variable, verb embodiment, and found that it was moderated by tense and regularity. In Experiment 3a we found that AoA predicted faster response times in verb reading. Finally, in Experiment 3b, semantic diversity predicted response times in a past tense generation task, either facilitating or inhibiting responses for regular or irregular verbs, respectively. These results demonstrate that semantic richness variables contribute to verb meaning even when verbs are presented in isolation, and that these effects depend on several factors unique to verbs.

The meaning of 'life' and other abstract words: Insights from neuropsychology

There are a number of long‐standing theories on how the cognitive processing of abstract words, like ‘life’, differs from that of concrete words, like ‘knife’. This review considers current perspectives on this debate, focusing particularly on insights obtained from patients with language disorders and integrating these with evidence from functional neuroimaging studies. The evidence supports three distinct and mutually compatible hypotheses. (1) Concrete and abstract words differ in their representational substrates, with concrete words depending particularly on sensory experiences and abstract words on linguistic, emotional, and magnitude‐based information. Differential dependence on visual versus verbal experience is supported by the evidence for graded specialization in the anterior temporal lobes for concrete versus abstract words. In addition, concrete words have richer representations, in line with better processing of these words in most aphasic patients and, in particular, patients with semantic dementia. (2) Abstract words place greater demands on executive regulation processes because they have variable meanings that change with context. This theory explains abstract word impairments in patients with semantic‐executive deficits and is supported by neuroimaging studies showing greater response to abstract words in inferior prefrontal cortex. (3) The relationships between concrete words are governed primarily by conceptual similarity, while those of abstract words depend on association to a greater degree. This theory, based primarily on interference and priming effects in aphasic patients, is the most recent to emerge and the least well understood. I present analyses indicating that patterns of lexical co‐occurrence may be important in understanding these effects.