Insights into failed lexical retrieval from network science

Obsolescence effects in second language phonological networks

Phonological networks are representations of word forms and their phonological relationships with other words in a given language lexicon. A principle underlying the growth (or evolution) of those networks is preferential attachment, or the "rich-gets-richer" mechanisms, according to which words with many phonological neighbors (or links) are the main beneficiaries of future growth opportunities. Due to their limited number of words, language lexica constitute node-constrained networks where growth cannot keep increasing in a linear way; hence, preferential attachment is likely mitigated by certain factors. The present study investigated obsolescence effects (i.e., a word's finite timespan of being active in terms of growth) in an evolving phonological network of English as a second language. It was found that phonological neighborhoods are constructed by one large initial lexical spurt, followed by sublinear growth spurts that eventually lead to very limited growth in later lexical spurts during network evolution. First-language-given neighborhood densities are rarely reached even by the most advanced language learners. An analysis of the strength of phonological relationships between phonological word forms revealed a tendency to incorporate phonetically more distant phonological neighbors at earlier acquisition stages. Overall, the findings suggest an obsolescence effect in growth that favors younger words. Implications for the second-language lexicon include leveraged learning mechanisms and learning bouts focused on a smaller range of phonological segments, and involve questions concerning lexical processing in aging networks.

Identifying the phonological backbone in the mental lexicon

Previous studies used techniques from network science to identify individual nodes and a set of nodes that were "important" in a network of phonological word-forms from English. In the present study we used a network simplification process-known as the backbone-that removed redundant edges to extract a subnetwork of "important" words from the network of phonological word-forms. The backbone procedure removed 68.5% of the edges in the original network to extract a backbone with a giant component containing 6,211 words. We compared psycholinguistic and network measures of the words in the backbone to the words that did not survive the backbone extraction procedure. Words in the backbone occurred more frequently in the language, were shorter in length, were similar to more phonological neighbors, and were closer to other words than words that did not survive the backbone extraction procedure. Words in the backbone of the phonological network might form a "kernel lexicon"-a small but essential set of words that allows one to communicate in a wide-range of situations-and may provide guidance to clinicians and researchers on which words to focus on to facilitate typical development, or to accelerate rehabilitation efforts. The backbone extraction method may also prove useful in other applications of network science to the speech, language, hearing and cognitive sciences.

Representing melodic relationships using network science

Music is a complex system consisting of many dimensions and hierarchically organized information-the organization of which, to date, we do not fully understand. Network science provides a powerful approach to representing such complex systems, from the social networks of people to modelling the underlying network structures of different cognitive mechanisms. In the present research, we explored whether network science methodology can be extended to model the melodic patterns underlying expert improvised music. Using a large corpus of transcribed improvisations, we constructed a network model in which 5-pitch sequences were linked depending on consecutive occurrences, constituting 116,403 nodes (sequences) and 157,429 edges connecting them. We then investigated whether mathematical graph modelling relates to musical characteristics in real-world listening situations via a behavioral experiment paralleling those used to examine language. We found that as melodic distance within the network increased, participants judged melodic sequences as less related. Moreover, the relationship between distance and reaction time (RT) judgements was quadratic: participants slowed in RT up to distance four, then accelerated; a parallel finding to research in language networks. This study offers insights into the hidden network structure of improvised tonal music and suggests that humans are sensitive to the property of melodic distance in this network. More generally, our work demonstrates the similarity between music and language as complex systems, and how network science methods can be used to quantify different aspects of its complexity.

Using Network Science and Psycholinguistic Megastudies to Examine the Dimensions of Phonological Similarity

Network science was used to examine different dimensions of phonological similarity in English. Data from a phonological associate task and an identification of words in noise task were used to create a phonological association network and a misperception network. These networks were compared to a network formed by a computational metric widely used to assess phonological similarity (i.e., one-phoneme metric). The phonological association network and the misperception network were topographically more similar to each other than either were to the one-phoneme metric network, but there were several network features in common between the one-phoneme metric network and the phonological association network. To assess the influence of network structure on processing, we compared the influence of degree (i.e., neighborhood density) from each of the networks on visual and auditory lexical decision reaction times obtained from two psycholinguistic megastudies. The effect of degree differed across network types and tasks. We discuss the use of each approach to determine phonological similarity and a possible direction forward for language research through the use of multiplex networks.

Editors' Introduction to Networks of the Mind: How Can Network Science Elucidate Our Understanding of Cognition?

Thinking is complex. Over the years, several types of methods and paradigms have developed across the psychological, cognitive, and neural sciences to study such complexity. A rapidly growing multidisciplinary quantitative field of network science offers quantitative methods to represent complex systems as networks, or graphs, and study the network properties of these systems. While the application of network science to study the brain has greatly advanced our understanding of the brains structure and function, the application of these tools to study cognition has been done to a much lesser account. This topic is a collection of papers that discuss the fruitfulness of applying network science to study cognition across a wide scope of research areas from generalist accounts of memory and encoding, to individual differences, to communities, and finally to cultural and individual change.

Semantic maturation during the comprehension-expression gap in late and typical talkers

This study investigates the influence of semantic maturation on early lexical development by examining the impact of contextual diversity-known to influence semantic development-on word promotion from receptive to productive vocabularies (i.e., comprehension-expression gap). Study 1 compares the vocabularies of 3685 American-English-speaking typical talkers (TTs) and late talkers (LTs; 16-30 months old; 1257 females, 1021 gender unknown; ethnicity unknown; data downloaded in 2018) and finds that LTs, with a longer preverbal phase, produced nouns with lower contextual diversity (R²  = .80), but verbs with higher contextual diversity (R²  = .13). Study 2 compares computational network growth models of semantic maturation and finds that verbs require more semantic maturation than nouns, and TTs produce words that are more semantically mature than LTs.

Multilayer networks: An untapped tool for understanding bilingual neurocognition

Cross-linguistic similarity is a term so broad and multi-faceted that it is not easily defined. The degree of overlap between languages is known to affect lexical competition during online processing and production, and its relevance for second language acquisition has also been established. Nevertheless, determining what makes two languages similar (or not) increases in complexity when multiple levels of the language hierarchy (e.g., phonology, syntax) are considered. How can we feasibly account for the patterns of convergence and divergence at each level of representation, as well as the interactions between them? The growing field of network science brings new methodologies to bear on this longstanding question. Below, we summarize current network science approaches to modeling language structure and discuss implications for understanding various linguistic processes. Critically, we stress the particular value of multilayer techniques, unique and powerful in their ability to simultaneously accommodate an array of node-to-node (or word-to-word) relationships.

Is the Mind a Network? Maps, Vehicles, and Skyhooks in Cognitive Network Science

Cognitive researchers often carve cognition up into structures and processes. Cognitive processes operate on structures, like vehicles driving over a map. Language alongside semantic and episodic memory are proposed to have structure, as are perceptual systems. Over these structures, processes operate to construct memory and solve problems by retrieving and manipulating information. Network science offers an approach to representing cognitive structures and has made tremendous inroads into understanding the nature of cognitive structure and process. But is the mind a network? If so, what kind? In this article, we briefly review the main metaphors, assumptions, and pitfalls prevalent in cognitive network science (maps and vehicles; one network/process to rule them all), highlight the need for new metaphors that elaborate on the map-and-vehicle framework (wormholes, skyhooks, and generators), and present open questions in studying the mind as a network (the challenge of capturing network change, what should the edges of cognitive networks be made of, and aggregated vs. individual-based networks). One critical lesson of this exercise is that the richness of the mind as network approach makes it a powerful tool in its own right; it has helped to make our assumptions more visible, generating new and fascinating questions, and enriching the prospects for future research. A second lesson is that the mind as a network-though useful-is incomplete. The mind is not a network, but it may contain them.

Beauty and Wellness in the Semantic Memory of the Beholder

Beauty and wellness are terms used often in common parlance, however their meaning and relation to each other is unclear. To probe their meaning, we applied network science methods to estimate and compare the semantic networks associated with beauty and wellness in different age generation cohorts (Generation Z, Millennials, Generation X, and Baby Boomers) and in women and men. These mappings were achieved by estimating group-based semantic networks from free association responses to a list of 47 words, either related to Beauty, Wellness, or Beauty + Wellness. Beauty was consistently related to Elegance, Feminine, Gorgeous, Lovely, Sexy, and Stylish. Wellness was consistently related Aerobics, Fitness, Health, Holistic, Lifestyle, Medical, Nutrition, and Thrive. In addition, older cohorts had semantic networks that were less connected and more segregated from each other. Finally, we found that women compared to men had more segregated and organized concepts of Beauty and Wellness. In contemporary societies that are pre-occupied by the pursuit of beauty and a healthy lifestyle, our findings shed novel light on how people think about beauty and wellness and how they are related across different age generations and by sex.

Unveiling the nature of interaction between semantics and phonology in lexical access based on multilayer networks

An essential aspect of human communication is the ability to access and retrieve information from ones’ ‘mental lexicon’. This lexical access activates phonological and semantic components of concepts, yet the question whether and how these two components relate to each other remains widely debated. We harness tools from network science to construct a large-scale linguistic multilayer network comprising of phonological and semantic layers. We find that the links in the two layers are highly similar to each other and that adding information from one layer to the other increases efficiency by decreasing the network overall distances, but specifically affecting shorter distances. Finally, we show how a multilayer architecture demonstrates the highest efficiency, and how this efficiency relates to weak semantic relations between cue words in the network. Thus, investigating the interaction between the layers and the unique benefit of a linguistic multilayer architecture allows us to quantify theoretical cognitive models of lexical access.

Semantic memory: A review of methods, models, and current challenges

Adult semantic memory has been traditionally conceptualized as a relatively static memory system that consists of knowledge about the world, concepts, and symbols. Considerable work in the past few decades has challenged this static view of semantic memory, and instead proposed a more fluid and flexible system that is sensitive to context, task demands, and perceptual and sensorimotor information from the environment. This paper (1) reviews traditional and modern computational models of semantic memory, within the umbrella of network (free association-based), feature (property generation norms-based), and distributional semantic (natural language corpora-based) models, (2) discusses the contribution of these models to important debates in the literature regarding knowledge representation (localist vs. distributed representations) and learning (error-free/Hebbian learning vs. error-driven/predictive learning), and (3) evaluates how modern computational models (neural network, retrieval-based, and topic models) are revisiting the traditional "static" conceptualization of semantic memory and tackling important challenges in semantic modeling such as addressing temporal, contextual, and attentional influences, as well as incorporating grounding and compositionality into semantic representations. The review also identifies new challenges regarding the abundance and availability of data, the generalization of semantic models to other languages, and the role of social interaction and collaboration in language learning and development. The concluding section advocates the need for integrating representational accounts of semantic memory with process-based accounts of cognitive behavior, as well as the need for explicit comparisons of computational models to human baselines in semantic tasks to adequately assess their psychological plausibility as models of human semantic memory.

Quantifying flexibility in thought: The resiliency of semantic networks differs across the lifespan

Older adults tend to have a broader vocabulary compared to younger adults - indicating a richer storage of semantic knowledge - but their retrieval abilities decline with age. Recent advances in quantitative methods based on network science have investigated the effect of aging on semantic memory structure. However, it is yet to be determined how this aging effect on semantic memory structure relates to its overall flexibility. Percolation analysis provides a quantitative measure of the flexibility of a semantic network, by examining how a semantic memory network is resistant to "attacks" or breaking apart. In this study, we incorporated percolation analyses to examine how semantic networks of younger and older adults break apart to investigate potential age-related differences in language production. We applied the percolation analysis to 3 independent sets of data (total N = 78 younger, 78 older adults) from which we generated semantic networks based on verbal fluency performance. Across all 3 datasets, the percolation integrals of the younger adults were larger than older adults, indicating that older adults' semantic networks were less flexible and broke down faster than the younger adults'. Our findings provide quantitative evidence for diminished flexibility in older adults' semantic networks, despite the stability of semantic knowledge across the lifespan. This may be one contributing factor to age-related differences in language production.

The structure of semantic representation shapes controlled semantic retrieval

An essential aim in the research on semantic cognition is to understand the interplay between the structure of semantic representation and controlled processes that operate on it to generate flexible behaviours. To evaluate the link between semantic network connectivity and semantic control functions (semantic inhibition and switching), we employed a network theory approach and revealed that controlled semantic processing was reliably associated with connectivity of conceptual representation. In particular, our results show that efficient information flow afforded by high connectivity of semantic network is coupled with superior switching but poor inhibition ability. These findings suggest that the network architectures that facilitate efficient semantic activation spreading aid flexible transitions between semantic clusters but impede inhibition employed to suppress inappropriate or interfering semantic representations. Overall, our study provides a novel insight into the mechanisms underlying controlled semantic processing that is recruited to disentangle from habitual structure of semantic representation.

What Can Network Science Tell Us About Phonology and Language Processing?

Contemporary psycholinguistic models place significant emphasis on the cognitive processes involved in the acquisition, recognition, and production of language but neglect many issues related to the representation of language‐related information in the mental lexicon. In contrast, a central tenet of network science is that the structure of a network influences the processes that operate in that system, making process and representation inextricably connected. Here, we consider how the structure found across phonological networks of several languages from different language families may influence language processing as we age and experience diseases that affect cognition during the typical and atypical acquisition of new words, during typical perception and production of speech in adults, and during language change over time. We conclude that the network science approach may not only provide insights into specific language processes but also provide a way to connect the work from these domains, which are becoming increasingly balkanized.

Methodological Considerations for Incorporating Clinical Data Into a Network Model of Retrieval Failures

Difficulty retrieving information (e.g., words) from memory is prevalent in neurogenic communication disorders (e.g., aphasia and dementia). Theoretical modeling of retrieval failures often relies on clinical data, despite methodological limitations (e.g., locus of retrieval failure, heterogeneity of individuals, and progression of disorder/disease). Techniques from network science are naturally capable of handling these limitations. This paper reviews recent work using a multiplex lexical network to account for word retrieval failures and highlights how network science can address the limitations of clinical data. Critically, any model we employ could impact clinical practice and patient lives, harkening the need for theoretically well-informed network models.

The Challenge of Achieving Greater Generalization in Phonological Treatment of Aphasia

BACKGROUND

Stimulus selection is important to anomia treatment because similarity between trained and untrained words in the mental lexicon may influence treatment generalization. We focused on phonological similarity between trained and untrained words from a clinical trial of Phonomotor Treatment (PMT) that showed gains in confrontation naming accuracy of untrained words post-treatment. One way to capture the amount of similarity between the trained and untrained words is to consider the phonological network path distance between words. We posited that the distance between trained and untrained words in a phonological network could account for the improvement in confrontation naming accuracy post-treatment.

AIM

To define the phonological network distance between trained and untrained words that influences change in confrontation naming accuracy post-treatment.

METHODS AND PROCEDURES

We retrospectively analyzed data from 28 people with aphasia who received PMT as part of a clinical trial. Participants completed confrontation naming (baseline, post-treatment, and 3-months post-treatment) of words varying in phonological distance to the treatment stimuli. We used a phonological network to calculate the average shortest path length (ASPL), defined by number of phoneme differences, between an untrained word and all trained words. We used mixed effects regression models to predict change in confrontation naming accuracy of untrained words post-treatment from ASPL. Several post-hoc analyses were also conducted.

OUTCOMES AND RESULTS

We found no effect of ASPL on change in confrontation naming accuracy of untrained words immediately post- and 3-months post-treatment. However, post-hoc analyses indicated significant subject heterogeneity and limitations in observable path distance between trained and untrained words.

CONCLUSION

Despite the clinical trial report that confrontation naming of untrained words improved after PMT, we found no overall effect of ASPL on the amount of improvement. We discuss further investigation of the entire domain of phonological sequence knowledge (the phonological sequence knowledge landscape) and its influence on treatment generalization, and the potential importance of identifying predictors of treatment response to enhance the effects of treatment generalization.

Abstract art paintings, global image properties, and verbal descriptions: An empirical and computational investigation

While global image properties (GIPs) relate to preference ratings in many categories of visual stimuli, this relationship is typically not seen for abstract art paintings. Using computational network science and empirical methods, we further investigated GIPs and subjective preferences. First, we replicated the earlier observation that GIPs do not relate to preferences for abstract art. Next, we estimated the network structure of abstract art paintings using two approaches: the first was based on verbal descriptions and the second on GIPs. We examined the extent to which network measures computed from these two networks (1) related to preference for abstract art paintings and (2) determined affiliation of images to specific art styles. Only semantic-based network predicted the subjective preference ratings and art style. Finally, preference and GIPs differed for sub-groups of abstract art paintings. Our results demonstrate the importance of verbal descriptors in evaluating abstract art, and that it is not useful in empirical aesthetics to treat abstract art paintings as a single category.

Organizational and Formational Structures of Networks in the Mental Lexicon: A State-Of-The-Art through Systematic Review

This state-of-the-art presents a systematic exploration on the use of network patterns in global research efforts to understand, organize and represent the mental lexicon. Results have shown an increase over recent years in the usage of complex, small-world and scale-free network patterns within the literature. With the increasing complexity of network patterns, we see more potential in the inter-disciplinary exploration of the mental lexicon through universal and mathematically-describable, behavioral patterns in small-world and scale-free networks. A systematic review of 36 items of methodologically-selected literature serve as a means to explore how the greater literary body understands network structures within the mental lexicon. Network-based approaches are discriminated between three contrasting varieties. These include: ‘simple networks’, characterized by arbitrarily organized graph patterns of metaphorical importance; ‘connectionist networks’, a broad category of networks which explore the structural features of a system through the analysis of emergent properties; and lastly ‘complex networks’, distinguished as small-world, scale-free networks which follow a strict and mathematically-describable structure in agreement with the Barabási–Albert model. Each network approach is explored in terms of their discernible differences which relate to their parameters and affect their implications. A final evaluation of observed patterns within the selected literature is offered, as well as an elaboration on the sense of trajectory beheld in the research in order to offer insight and orientation for future research.

Phonological network fluency identifies phonological restructuring through mental search

We investigated network principles underlying mental search through a novel phonological verbal fluency task. Post exclusion, 95 native-language Mandarin speakers produced as many items that differed by a single lexical tone as possible within one minute. Their verbal productions were assessed according to several novel graded fluency measures, and network science measures that accounted for the structure, cohesion and interconnectedness of lexical items. A multivariate regression analysis of our participants' language backgrounds included their mono- or multi-lingual status, English proficiency, and fluency in other Chinese languages/dialects. Higher English proficiency predicted lower error rates and greater interconnectedness, while higher fluency in other Chinese languages/dialects revealed lower successive similarity and lower network coherence. This inverse relationship between English and other Chinese languages/dialects provides evidence of the restructuring of the phonological mental lexicon.

Viability in Multiplex Lexical Networks and Machine Learning Characterizes Human Creativity

Previous studies have shown how individual differences in creativity relate to differences in the structure of semantic memory. However, the latter is only one aspect of the whole mental lexicon, a repository of conceptual knowledge that is considered to simultaneously include multiple types of conceptual similarities. In the current study, we apply a multiplex network approach to compute a representation of the mental lexicon combining semantics and phonology and examine how it relates to individual differences in creativity. This multiplex combination of 150,000 phonological and semantic associations identifies a core of words in the mental lexicon known as viable cluster, a kernel containing simpler to parse, more general, concrete words acquired early during language learning. We focus on low (N = 47) and high (N = 47) creative individuals’ performance in generating animal names during a semantic fluency task. We model this performance as the outcome of a mental navigation on the multiplex lexical network, going within, outside, and in-between the viable cluster. We find that low and high creative individuals differ substantially in their access to the viable cluster during the semantic fluency task. Higher creative individuals tend to access the viable cluster less frequently, with a lower uncertainty/entropy, reaching out to more peripheral words and covering longer multiplex network distances between concepts in comparison to lower creative individuals. We use these differences for constructing a machine learning classifier of creativity levels, which leads to an accuracy of 65 . 0 ± 0 . 9 % and an area under the curve of 68 . 0 ± 0 . 8 % , which are both higher than the random expectation of 50%. These results highlight the potential relevance of combining psycholinguistic measures with multiplex network models of the mental lexicon for modelling mental navigation and, consequently, classifying people automatically according to their creativity levels.

On Curiosity: A Fundamental Aspect of Personality, a Practice of Network Growth

Human personality is reflected in patterns-or networks-of behavior, either in thought or action. Curiosity is an oft-treasured component of one's personality, commonly associated with information-seeking proclivities with distinct neurophysiological correlates. The markers of curiosity can differ substantially across people, suggesting the possibility that personality also determines the architectural style of one's curiosity. Yet progress in defining those styles, and marking their neurophysiological basis, has been hampered by fairly fundamental difficulties in defining curiosity itself. Here, we offer and exercise a definition of the practice of curiosity as knowledge network building, one particular pattern of thought behavior. To unpack this definition and motivate its utility, we begin with a short primer on network science and describe how the mathematical object of a network can be used to map items and relations that are characteristic of bodies of knowledge. Next, we turn to a discussion of how networks grow, how their growth can be modeled, and how the practice of curiosity can be formalized as a process of network growth. We pay particular attention to how individuals may differ in how they build their knowledge networks, and discuss how the sort, manner, and action of building can be modulated by experience. We discuss how this definition of the practice of curiosity motivates new experiments and theory development at the interdisciplinary intersection of network science, personality neuroscience, education, and curiosity studies. We close with a note on the potential of network science to inform studies of other domains of personality, and the patterns of thought- or action-behavior characteristic thereof.

The orthographic similarity structure of English words: Insights from network science

Network science has been applied to study the structure of the mental lexicon, the part of long-term memory where all the words a person knows are stored. Here the tools of network science are used to study the organization of orthographic word-forms in the mental lexicon and how that might influence visual word recognition. An orthographic similarity network of the English language was constructed such that each node represented an English word, and undirected, unweighted edges were placed between words that differed by an edit distance of 1, a commonly used operationalization of orthographic similarity in psycholinguistics. The largest connected component of the orthographic language network had a small-world structure and a long-tailed degree distribution. Additional analyses were conducted using behavioral data obtained from a psycholinguistic database to determine if network science measures obtained from the orthographic language network could be used to predict how quickly and accurately people process written words. The present findings show that the structure of the mental lexicon influences lexical access in visual word recognition.

Neuromagnetic correlates of audiovisual word processing in the developing brain

The brain undergoes enormous changes during childhood. Little is known about how the brain develops to serve word processing. The objective of the present study was to investigate the maturational changes of word processing in children and adolescents using magnetoencephalography (MEG). Responses to a word processing task were investigated in sixty healthy participants. Each participant was presented with simultaneous visual and auditory word pairs in "match" and "mismatch" conditions. The patterns of neuromagnetic activation from MEG recordings were analyzed at both sensor and source levels. Topography and source imaging revealed that word processing transitioned from bilateral connections to unilateral connections as age increased from 6 to 17 years old. Correlation analyses of language networks revealed that the path length of word processing networks negatively correlated with age (r = -0.833, p < 0.0001), while the connection strength (r = 0.541, p < 0.01) and the clustering coefficient (r = 0.705, p < 0.001) of word processing networks were positively correlated with age. In addition, males had more visual connections, whereas females had more auditory connections. The correlations between gender and path length, gender and connection strength, and gender and clustering coefficient demonstrated a developmental trend without reaching statistical significance. The results indicate that the developmental trajectory of word processing is gender specific. Since the neuromagnetic signatures of these gender-specific paths to adult word processing were determined using non-invasive, objective, and quantitative methods, the results may play a key role in understanding language impairments in pediatric patients in the future.

Flexibility of thought in high creative individuals represented by percolation analysis

Flexibility of thought is theorized to play a critical role in the ability of high creative individuals to generate novel and innovative ideas. However, this has been examined only through indirect behavioral measures. Here we use network percolation analysis (removal of links in a network whose strength is below an increasing threshold) to computationally examine the robustness of the semantic memory networks of low and high creative individuals. Robustness of a network indicates its flexibility and thus can be used to quantify flexibility of thought as related to creativity. This is based on the assumption that the higher the robustness of the semantic network, the higher its flexibility. Our analysis reveals that the semantic network of high creative individuals is more robust to network percolation compared with the network of low creative individuals and that this higher robustness is related to differences in the structure of the networks. Specifically, we find that this higher robustness is related to stronger links connecting between different components of similar semantic words in the network, which may also help to facilitate spread of activation over their network. Thus, we directly and quantitatively examine the relation between flexibility of thought and creative ability. Our findings support the associative theory of creativity, which posits that high creative ability is related to a flexible structure of semantic memory. Finally, this approach may have further implications, by enabling a quantitative examination of flexibility of thought, in both healthy and clinical populations.

Multiplex lexical networks reveal patterns in early word acquisition in children

Network models of language have provided a way of linking cognitive processes to language structure. However, current approaches focus only on one linguistic relationship at a time, missing the complex multi-relational nature of language. In this work, we overcome this limitation by modelling the mental lexicon of English-speaking toddlers as a multiplex lexical network, i.e. a multi-layered network where N = 529 words/nodes are connected according to four relationship: (i) free association, (ii) feature sharing, (iii) co-occurrence, and (iv) phonological similarity. We investigate the topology of the resulting multiplex and then proceed to evaluate single layers and the full multiplex structure on their ability to predict empirically observed age of acquisition data of English speaking toddlers. We find that the multiplex topology is an important proxy of the cognitive processes of acquisition, capable of capturing emergent lexicon structure. In fact, we show that the multiplex structure is fundamentally more powerful than individual layers in predicting the ordering with which words are acquired. Furthermore, multiplex analysis allows for a quantification of distinct phases of lexical acquisition in early learners: while initially all the multiplex layers contribute to word learning, after about month 23 free associations take the lead in driving word acquisition.

Using the OASES-A to illustrate how network analysis can be applied to understand the experience of stuttering

PURPOSE

Network science uses mathematical and computational techniques to examine how individual entities in a system, represented by nodes, interact, as represented by connections between nodes. This approach has been used by Cramer et al. (2010) to make "symptom networks" to examine various psychological disorders. In the present analysis we examined a network created from the items in the Overall Assessment of the Speaker's Experience of Stuttering-Adult (OASES-A), a commonly used measure for evaluating adverse impact in the lives of people who stutter.

METHOD

The items of the OASES-A were represented as nodes in the network. Connections between nodes were placed if responses to those two items in the OASES-A had a correlation coefficient greater than ±0.5. Several network analyses revealed which nodes were "important" in the network.

RESULTS

Several centrally located nodes and "key players" in the network were identified. A community detection analysis found groupings of nodes that differed slightly from the subheadings of the OASES-A.

CONCLUSIONS

Centrally located nodes and "key players" in the network may help clinicians prioritize treatment. The different community structure found for people who stutter suggests that the way people who stutter view stuttering may differ from the way that scientists and clinicians view stuttering. Finally, the present analyses illustrate how the network approach might be applied to other speech, language, and hearing disorders to better understand how those disorders are experienced and to provide insights for their treatment.

The Hyper-Modular Associative Mind: A Computational Analysis of Associative Responses of Persons with Asperger Syndrome

Rigidity of thought is considered a main characteristic of persons with Asperger syndrome (AS). This rigidity may explain the poor comprehension of unusual semantic relations, frequently exhibited by persons with AS. Research indicates that such deficiency is related to altered mental lexicon organization, but has never been directly examined. The present study used computational network science tools to compare the mental lexicon structure of persons with AS and matched controls. Persons with AS and matched controls generated free associations, and network tools were used to extract and compare the mental lexicon structure of the two groups. The analysis revealed that persons with AS exhibit a hyper-modular semantic organization: their mental lexicon is more compartmentalized compared to matched controls. We argue that this hyper-modularity may be related to the rigidity of thought which characterizes persons with AS and discuss the clinical and more general cognitive implications of our findings.

Local Patterns to Global Architectures: Influences of Network Topology on Human Learning

A core question in cognitive science concerns how humans acquire and represent knowledge about their environments. To this end, quantitative theories of learning processes have been formalized in an attempt to explain and predict changes in brain and behavior. We connect here statistical learning approaches in cognitive science, which are rooted in the sensitivity of learners to local distributional regularities, and network science approaches to characterizing global patterns and their emergent properties. We focus on innovative work that describes how learning is influenced by the topological properties underlying sensory input. The confluence of these theoretical approaches and this recent empirical evidence motivate the importance of scaling-up quantitative approaches to learning at both the behavioral and neural levels.

Speech error and tip of the tongue diary for mobile devices

Collections of various types of speech errors have increased our understanding of the acquisition, production, and perception of language. Although such collections of naturally occurring language errors are invaluable for a number of reasons, the process of collecting various types of speech errors presents many challenges to the researcher interested in building such a collection, among them a significant investment of time and effort to obtain a sufficient number of examples to enable statistical analysis. Here we describe a freely accessible website http://spedi.ku.edu that helps users document slips of the tongue, slips of the ear, and tip of the tongue states that they experience firsthand or observe in others. The documented errors are amassed, and made available for other users to analyze, thereby distributing the time and effort involved in collecting errors across a large number of individuals instead of saddling the lone researcher, and facilitating distribution of the collection to other researchers. This approach also addresses some issues related to data curation that hampered previous error collections, and enables the collection to continue to grow over a longer period of time than previous collections. Finally, this web-based tool creates an opportunity for language scientists to engage in outreach efforts to increase the understanding of language disorders and research in the general public.

A Complex Network Approach to Distributional Semantic Models

A number of studies on network analysis have focused on language networks based on free word association, which reflects human lexical knowledge, and have demonstrated the small-world and scale-free properties in the word association network. Nevertheless, there have been very few attempts at applying network analysis to distributional semantic models, despite the fact that these models have been studied extensively as computational or cognitive models of human lexical knowledge. In this paper, we analyze three network properties, namely, small-world, scale-free, and hierarchical properties, of semantic networks created by distributional semantic models. We demonstrate that the created networks generally exhibit the same properties as word association networks. In particular, we show that the distribution of the number of connections in these networks follows the truncated power law, which is also observed in an association network. This indicates that distributional semantic models can provide a plausible model of lexical knowledge. Additionally, the observed differences in the network properties of various implementations of distributional semantic models are consistently explained or predicted by considering the intrinsic semantic features of a word-context matrix and the functions of matrix weighting and smoothing. Furthermore, to simulate a semantic network with the observed network properties, we propose a new growing network model based on the model of Steyvers and Tenenbaum. The idea underlying the proposed model is that both preferential and random attachments are required to reflect different types of semantic relations in network growth process. We demonstrate that this model provides a better explanation of network behaviors generated by distributional semantic models.

Examining assortativity in the mental lexicon: Evidence from word associations

Words are characterized by a variety of lexical and psychological properties, such as their part of speech, word-frequency, concreteness, or affectivity. In this study, we examine how these properties relate to a word's connectivity in the mental lexicon, the structure containing a person's knowledge of words. In particular, we examine the extent to which these properties display assortative mixing, that is, the extent to which words in the lexicon are more likely to be connected to words that share these properties. We investigated three types of word properties: 1) subjective word covariates: valence, dominance, arousal, and concreteness; 2) lexical information: part of speech; and 3) distributional word properties: age-of-acquisition, word frequency, and contextual diversity. We assessed which of these factors exhibit assortativity using a word association task, where the probability of producing a certain response to a cue is a measure of the associative strength between the cue and response in the mental lexicon. Our results show that the extent to which these aspects exhibit assortativity varies considerably, with a high cue-response correspondence on valence, dominance, arousal, concreteness, and part of speech, indicating that these factors correspond to the words people deem as related. In contrast, we find that cues and responses show only little correspondence on word frequency, contextual diversity, and age-of-acquisition, indicating that, compared to subjective and lexical word covariates, distributional properties exhibit only little assortativity in the mental lexicon. Possible theoretical accounts and implications of these findings are discussed.

Using network science in the language sciences and clinic

A number of variables—word frequency, word length—have long been known to influence language processing. This study briefly reviews the effects in speech perception and production of two more recently examined variables: phonotactic probability and neighbourhood density. It then describes a new approach to study language, network science, which is an interdisciplinary field drawing from mathematics, computer science, physics and other disciplines. In this approach, nodes represent individual entities in a system (i.e. phonological word-forms in the lexicon), links between nodes represent relationships between nodes (i.e. phonological neighbours) and various measures enable researchers to assess the micro-level (i.e. the individual word), the macro-level (i.e. characteristics about the whole system) and the meso-level (i.e. how an individual fits into smaller sub-groups in the larger system). Although research on individual lexical characteristics such as word-frequency has increased understanding of language processing, these measures only assess the "micro-level". Using network science, researchers can examine words at various levels in the system and how each word relates to the many other words stored in the lexicon. Several new findings using the network science approach are summarized to illustrate how this approach can be used to advance basic research as well as clinical practice.

The influence of known-word-frequency on the acquisition of new neighbors in adults: evidence for exemplar representations in word-learning

Previous studies showed that a new word that is similar to many known words will be learned better than a new word that is similar to few known words (Storkel et al., 2006). In the present study we created novel words that were phonological neighbors to lexical hermits-or known words that do not have any phonological neighbors-that varied in frequency of occurrence. After several exposures, participants learned a higher proportion of novel words that were neighbors of high frequency known-words than nonwords that were neighbors of low frequency known-words. The present results have implications for abstractionist versus exemplar models of the mental lexicon and language processing, as well as for accounts of word frequency in models of language processing.

The influence of clustering coefficient on word-learning: how groups of similar sounding words facilitate acquisition

Clustering coefficient, C, measures the extent to which neighbors of a word are also neighbors of each other, and has been shown to influence speech production, speech perception, and several memory-related processes. In this study we examined how C influences word-learning. Participants were trained over three sessions at 1-week intervals, and tested with a picture-naming task on nonword-nonobject pairs. We found an advantage for novel words with high C (the neighbors of this novel word are likely to be neighbors with each other), but only after the 1-week retention period with no additional exposures to the stimuli. The results are consistent with the spreading-activation network-model of the lexicon proposed by Chan and Vitevitch (2009). The influence of C on various language-related processes suggests that characteristics of the individual word are not the only things that influence processing; rather, lexical processing may also be influenced by the relationships that exist among words in the lexicon.

Rigidity, chaos and integration: hemispheric interaction and individual differences in metaphor comprehension

Neurotypical individuals cope flexibly with the full range of semantic relations expressed in human language, including metaphoric relations. This impressive semantic ability may be associated with distinct and flexible patterns of hemispheric interaction, including higher right hemisphere (RH) involvement for processing novel metaphors. However, this ability may be impaired in specific clinical conditions, such as Asperger syndrome (AS) and schizophrenia. The impaired semantic processing is accompanied by different patterns of hemispheric interaction during semantic processing, showing either reduced (in Asperger syndrome) or excessive (in schizophrenia) RH involvement. This paper interprets these individual differences using the terms Rigidity, Chaos and Integration, which describe patterns of semantic memory network states that either lead to semantic well-being or are disruptive of it. We argue that these semantic network states lie on a rigidity-chaos semantic continuum. We define these terms via network science terminology and provide network, cognitive and neural evidence to support our claim. This continuum includes left hemisphere (LH) hyper-rigid semantic memory state on one end (e.g., in persons with AS), and RH chaotic and over-flexible semantic memory state on the other end (e.g., in persons with schizophrenia). In between these two extremes lie different states of semantic memory structure which are related to individual differences in semantic creativity. We suggest that efficient semantic processing is achieved by semantic integration, a balance between semantic rigidity and semantic chaos. Such integration is achieved via intra-hemispheric communication. However, impairments to this well-balanced and integrated pattern of hemispheric interaction, e.g., when one hemisphere dominates the other, may lead to either semantic rigidity or semantic chaos, moving away from semantic integration and thus impairing the processing of metaphoric language.

Investigating the structure of semantic networks in low and high creative persons

According to Mednick's (1962) theory of individual differences in creativity, creative individuals appear to have a richer and more flexible associative network than less creative individuals. Thus, creative individuals are characterized by "flat" (broader associations) instead of "steep" (few, common associations) associational hierarchies. To study these differences, we implement a novel computational approach to the study of semantic networks, through the analysis of free associations. The core notion of our method is that concepts in the network are related to each other by their association correlations-overlap of similar associative responses ("association clouds"). We began by collecting a large sample of participants who underwent several creativity measurements and used a decision tree approach to divide the sample into low and high creative groups. Next, each group underwent a free association generation paradigm which allowed us to construct and analyze the semantic networks of both groups. Comparison of the semantic memory networks of persons with low creative ability and persons with high creative ability revealed differences between the two networks. The semantic memory network of persons with low creative ability seems to be more rigid, compared to the network of persons with high creative ability, in the sense that it is more spread out and breaks apart into more sub-parts. We discuss how our findings are in accord and extend Mednick's (1962) theory and the feasibility of using network science paradigms to investigate high level cognition.

Keywords in the mental lexicon

Network science draws from a number of fields to examine complex systems using nodes to represent individuals and connections to represent relationships between individuals to form a network. This approach has been used in several areas of Psychology to illustrate the influence that the structure of a network has on processing in that system. In the present study the concept of keyplayers in a network (Borgatti, 2006) was examined in the domain of Psycholinguistics. Keyplayers are nodes in a network that, when removed, result in the network fracturing into several smaller components. A set of such nodes was found in a network of phonological word-forms as was another set of foil words, comparable to the "keywords" on a number of lexical and network characteristics. In three conventional psycholinguistic tasks keywords were responded to more quickly and accurately than the foils. A similar trend was observed in an analysis of the keywords and foils (and another set of foils) in the English Lexicon Project. These results open avenues for further exploration of keywords in various areas of language processing, and demonstrate the utility of the network science approach to psycholinguistics and psychology more generally.