The development of taxonomic structure in lexical memory

The Emergence of Richly Organized Semantic Knowledge from Simple Statistics: A Synthetic Review

As adults, we draw upon our ample knowledge about the world to support such vital cognitive feats as using language, reasoning, retrieving knowledge relevant to our current goals, planning for the future, adapting to unexpected events, and navigating through the environment. Our knowledge readily supports these feats because it is not merely a collection of stored facts, but rather functions as an organized, semantic network of concepts connected by meaningful relations. How do the relations that fundamentally organize semantic concepts emerge with development? Here, we cast a spotlight on a potentially powerful but often overlooked driver of semantic organization: Rich statistical regularities that are ubiquitous in both language and visual input. In this synthetic review, we show that a driving role for statistical regularities is convergently supported by evidence from diverse fields, including computational modeling, statistical learning, and semantic development. Finally, we identify a number of key avenues of future research into how statistical regularities may drive the development of semantic organization.

Statistical regularities shape semantic organization throughout development

Our knowledge about the world is represented not merely as a collection of concepts, but as an organized lexico-semantic network in which concepts can be linked by relations, such as "taxonomic" relations between members of the same stable category (e.g., cat and sheep), or association between entities that occur together or in the same context (e.g., sock and foot). To date, accounts of the origins of semantic organization have largely overlooked how sensitivity to statistical regularities ubiquitous in the environment may play a powerful role in shaping semantic development. The goal of the present research was to investigate how associations in the form of statistical regularities with which labels for concepts co-occur in language (e.g., sock and foot) and taxonomic relatedness (e.g., sock and pajamas) shape semantic organization of 4-5-year-olds and adults. To examine these aspects of semantic organization across development, we conducted three experiments examining effects of co-occurrence and taxonomic relatedness on cued recall (Experiment 1), word-picture matching (Experiment 2), and looking dynamics in a Visual World paradigm (Experiment 3). Taken together, the results of the three experiments provide evidence that co-occurrence-based links between concepts manifest in semantic organization from early childhood onward, and are increasingly supplemented by taxonomic links. We discuss these findings in relation to theories of semantic development.

Taxonomic structure in early to middle childhood: A longitudinal study with Zimbabwean schoolchildren

Children’s classification reasoning was examined with longitudinal data for 103 Zimbabwean Black (47) and White (56) children attending a randomly selected sample of public schools. The children varied by gender, social class membership (lower, middle, upper) and race (black, white). The children attempted a set of classification tasks at ages 7, 9, and 11. Responses to the classification tasks were scored in terms of interpretive strategy used to engage the tasks (taxonomic vs. instrumental). Repeated measures MANOVA and post-hoc orthogonal contrasts yielded significant differences in interpretive strategies by age or level of schooling, and social class. Higher social class membership was significantly related to more frequent use of taxonomic rather than functional classification strategies. Results support age/schooling-related effects in the development of taxonomic structure in a non-Western society.

What's in a link: associative and taxonomic priming effects in the infant lexicon

Infants develop a lexical-semantic system of associatively and semantically related words by the end of the second year of life. However, the precise nature of the lexical relationships that underpin the structure-building process remains under-determined. We compare two types of lexical-semantic relationship, associative and taxonomic, using a lexical-priming adaption of the intermodal preferential looking task with 21- and 24-month-olds. Prime-target word pairs were either associatively or taxonomically related or unrelated. A further control condition evaluated the facility of a prime word, in the absence of a target word, to promote target preferences. Twenty-four-month-olds, but not 21-month-old infants, exhibited a priming effect in both associative and taxonomic conditions, pointing to the formation of a lexical-semantic network driven by both associative and taxonomic relatedness late in the second year. The pattern of priming in 24-month-olds indicates the operation of inhibitory processes: unrelated primes interfere with target recognition whereas related primes do not. We argue that taxonomic and associative relationships between words are integral to the emergence of a structured lexicon and discuss the importance of inhibitory mechanisms in shaping early lexical-semantic memory.

The influence of categorical organization on verbal working memory

The aim of this study was to show the positive effects of categorical organization on verbal working memory (WM), in a modified version of a double task, such as the Listening Span Test (LST) (Daneman & Carpenter, 1980). Two experiments were performed comparing sentences with the typical definitional format (i.e., superordinate term, copula, and specification) to sentences simply describing objects or events. The results of the two experiments, with participants from children aged 6 to adults, revealed that word recall was better in Categorical sentences than in Descriptive sentences and are interpreted in terms of retrieval facilitation, due to pre-existing organization in semantic memory, at least from the age of 8 onwards. Recall performance was also better with sentences giving True statements than those giving False statements. Furthermore, Categorical False sentences are more effective in enhancing recall than Descriptive False sentences since they violate well-established semantic expectations. Such variables were also found to act among participants with a lower WM span, by this confirming that pre-existing organized information may compensate for less efficient WM.

Semantic abilities predict expressive lexicon in children with typical and atypical language development

In this study we used a semantic battery assessing the conceptual, lexical, and metacognitive level in semantic relationships to predict expressive lexicon in preschool children with typical and atypical language development. Our regression analyses showed that the tests of our semantic battery altogether accounted for 24% of variance in expressive lexicon after controlling for age and phonological short-term memory. The ability to memorize picture-cue/word pairs that were linked by taxonomic relations made a unique contribution to the expressive lexicon, and was a reliable marker of delayed expressive vocabulary in a group of children with specific language impairment.

Effects of Childhood Hearing Loss on Organization of Semantic Memory: Typicality and Relatedness

OBJECTIVE

The purpose of this research was to study how early childhood hearing loss affects development of concepts and categories, aspects of semantic knowledge that allow us to group and make inferences about objects with common properties, such as dogs versus cats. We assessed category typicality and out-of-category relatedness effects. The typicality effect refers to performance advantage (faster reaction times, fewer errors) for objects with a higher number of a category's characteristic properties; the out-of-category relatedness effect refers to performance disadvantage (slower reaction times and more errors) for out-of-category objects that share some properties with category members.

DESIGN

We applied a new children's speeded category-verification task (vote "yes" if the pictured object is clothing). Stimuli were pictures of typical and atypical category objects (e.g., pants, glove) and related and unrelated out-of-category objects (e.g., necklace, soup). Participants were 30 children with hearing impairment (HI) who were considered successful hearing aid users and who attended regular classes (mainstreamed) with some support services. Ages ranged from 5 to 15 yr (mean = 10 yr 8 mo). Results were related to normative data from .

RESULTS

Typical objects consistently showed preferential processing (faster reaction times, fewer errors), and related out-of-category objects consistently showed the converse. Overall, results between HI and normative groups exhibited striking similarity. Variation in speed of classification was influenced primarily by age and age-related competencies, such as vocabulary skill. Audiological status, however, independently influenced performance to a lesser extent, with positive responses becoming faster as degree of hearing loss decreased and negative responses becoming faster as age of identification/amplification/education decreased. There were few errors overall.

CONCLUSIONS

The presence of a typicality effect indicates that 1) the structure of conceptual representations for at least one category in the HI group was based on characteristic properties with an uneven distribution among members, and 2) typical objects with a higher number of characteristic properties were more easily accessed and/or retrieved. The presence of a relatedness effect indicates that the structure of representational knowledge in the HI group allowed them to appreciate semantic properties and understand that properties may be shared between categories. Speculations linked the association 1) between positive responses and degree of hearing loss to an increase in the quality, accessibility, and retrievability of conceptual representations with better hearing; and 2) between negative responses and age of identification/amplification/education to an improvement in effortful, postretrieval decision-making proficiencies with more schooling and amplified auditory experience. This research establishes the value of our new approach to assessing the organization of semantic memory in children with HI.

Accessing world knowledge: evidence from N400 and reaction time priming

How fast are we in accessing world knowledge? In two experiments, we tested for priming for word triplets that described a conceptual script (e.g., DIRECTOR-BRIBE-DISMISSAL) but were not associatively related and did not share a category relationship. Event-related brain potentials were used to track the time course at which script information becomes available. In Experiment 1, in which participants made lexical decisions, we found a facilitation for script-related relative to unrelated triplets, as indicated by (i) a decrease in both reaction time and errors, and (ii) an N400-like priming effect. In Experiment 2, we further explored the locus of script priming by increasing the contribution of meaning integration processes. The participants' task was to indicate whether the three words presented a plausible scenario. Again, an N400 script priming effect was obtained. Directing attention to script relations was effective in enhancing the N400 effect. The time course of the N400 effect was similar to that of the standard N400 effect to semantic relations. The present results show that script priming can be obtained in the visual modality, and that script information is immediately accessed and integrated with context. This supports the view that script information forms a central aspect of word meaning. The RT and N400 script priming effects reported in this article are problematic for most current semantic priming models, like spreading activation models, expectancy models, and task-specific semantic matching/integration models. They support a view in which there is no clear cutoff point between semantic knowledge and world knowledge.

What’s in a name? Typicality and relatedness effects in children

We studied how category typicality and out-of-category relatedness affect speeded category verification (vote "yes" if pictured object is clothing) in typically developing 4- to 14-year-olds and adults. Stimuli were typical and atypical category objects (e.g., pants, glove) and related and unrelated out-of-category objects (e.g., necklace, soup). Typical and unrelated out-of-category objects exhibited preferential processing (faster reaction times and fewer errors). Variations in typicality and relatedness disproportionately influenced children's performance, with developmental improvement associated with both verbal and nonverbal factors. Underextension versus overextension errors seemed to be associated with independent factors, namely multifaceted maturational factors versus receptive vocabulary skill, respectively. Errors were infrequent, suggesting spontaneous taxonomic classification by all participants. An experiment with printed words in adults replicated results, indicating that typicality and relatedness effects reflected organizational principles of the semantic system, not picture-related processes. This research establishes the viability of an online approach to assessing automatic components of semantic organization in children.

An apple is more than just a fruit: cross-classification in children's concepts

This research explored children's use of multiple forms of conceptual organization. Experiments 1 and 2 examined script (e.g., breakfast foods), taxonomic (e.g., fruits), and evaluative (e.g., junk foods) categories. The results showed that 4- and 7-year-olds categorized foods into all 3 categories, and 3-year-olds used both taxonomic and script categories. Experiment 3 found that 4- and 7-year-olds can cross-classify items, that is, classify a single food into both taxonomic and script categories. Experiments 4 and 5 showed that 7-year-olds and to some degree 4-year-olds can selectively use categories to make inductive inferences about foods. The results reveal that children do not rely solely on one form of categorization but are flexible in the types of categories they form and use.

Developmental differences in sensitivity to semantic relations among good and poor comprehenders: evidence from semantic priming

Semantic priming for category coordinates (e.g. CAT-DOG; AEROPLANE-TRAIN) and for pairs of words related through function (e.g. BROOM-FLOOR; SHAMPOO-HAIR) was assessed in children with good and poor reading comprehension, matched for decoding skill. Lexical association strength was also manipulated by comparing pairs of words that were highly associated with pairs that shared low association strength. Both groups of children showed priming for function-related words, but for the category co-ordinates, poor comprehenders only showed priming if the category pairs also shared high association strength. Good comprehenders showed priming for category-related targets, irrespective of the degree of prime-target association. These findings are related to models of language development in which category knowledge is gradually abstracted and refined from children's event-based knowledge and it is concluded that in the absence of explicit co-occurrence, poor comprehenders are less sensitive to abstract semantic relations than normal readers.