Nouns Mark Category Relations: Toddlers'' and Preschoolers'' Word-Learning Biases

Recent research suggests that preschool children approach the task of word learning equipped with implicit biases that lead them to prefer some possible meanings over others. The noun-category bias proposes that children favor category relations when interpreting the meaning of novel nouns. In the series of experiments reported here, we develop a stringent test of the noun-category bias and reveal that it is present in children as young as 2 years of age. In each experiment, children participated in a 5-item match-to-sample task. Children were presented with a target item (e.g., a cow) and 4 choices, 2 of which belonged to the same superordinate category as the target (e.g., a fox and a zebra) and 2 of which were thematically related to the target (e.g., milk and a barn). In Experiment 1 we demonstrate that novel nouns prompt preschool children to attend to superordinate-level category relations, even in the presence of multiple thematic alternatives. In Experiment 2, we ascertain that the bias is specific to nouns; novel adjectives do not highlight superordinate category relations. In Experiment 3, we demonstrate the noun-category bias in 2-year-olds. The nature and utility of the noun-category bias are discussed.     

Learning proper nouns and common nouns without clues from syntax

Syntax has been noted to play an important role in word learning in English; it distinguishes the fundamental conceptual difference between individuals (coded as proper nouns), nonindividuals (coded as mass nouns), and classes of individuals (coded as count nouns). The Japanese language does not have grammatical markers flagging the distinctions between count nouns and mass nouns, between proper nouns and common nouns, or between singulars and plurals. How Japanese 2- and 4-year-olds assign meaning to novel nouns associated with familiar and unfamiliar animals and inanimate objects was studied in the research reported here. When a novel label was given to an unfamiliar object, children assumed it to be a name for a basic-level object category whether the referent was an animal or an inanimate object. If the named object already had an established name, and if the object was an inanimate object, the children mapped the noun to a subordinate category. When the named object was an animal, however, they tended to interpret the label as a proper name. These results demonstrated that in the absence of useful information from syntax, 2-year-old Japanese children are able to fast map a noun to its meaning by elegantly coordinating word-learning biases and other available sources of information.     

Object Properties and Knowledge in Early Lexical Learning

The ease with which young children learn object nouns suggests that they possess strategies to identify properties critical to lexical category membership. In previous work, young children used a same-shape criterion to extend new count nouns. The present research tested the generality of this shape bias. 2- and 3-year-olds were asked either to extend a novel count noun to new instances, or to choose unnamed objects to go together. The objects varied in shape, size, and texture. For half of the subjects, the objects had eyes--a property strongly associated with certain material kinds. If young children know this association, they should attend to texture as well as shape in classifying objects with eyes. With named objects only, both 2- and 3-year-old children classified eyeless objects by shape and objects with eyes by both shape and texture. The results suggest that very young children possess considerable knowledge about conditional relations between kinds of perceptual properties. Knowledge of such conditional relations may aid children in forming new categories and thus in discovering new word meanings.     

Known and novel noun extensions: attention at two levels of abstraction

Two experiments tested the hypothesis that names direct attention at two levels of abstraction: Known names direct attention to the properties most relevant to the specific category; novel names direct attention to the shape, the property most generally relevant across known object names. English-speaking and Japanese-speaking 3-year-olds were shown a novel object that was named with (a) known nouns referring to things similar in shape or similar in material and color, and (b) novel nouns. Given known nouns, children attended to shape when the name referred to a category organized by shape, but they did not when the name referred to a category organized by other properties. Children generalized novel names by shape. The results are discussed within the debate between shape-based and taxonomic categories.     

Incorporating new words into the lexicon: preliminary evidence for language hierarchies in two-year-old children

In 4 experiments, we examined how young children incorporate new word meanings into their lexicons. 2-year-olds were each taught a new noun for an object that already had a known label (e.g., a "fep" for a dog). Children's interpretations of the new nouns were assessed by asking subjects to select the named toy from an array of 4 toys (e.g., "Point to a fep"). The experiments were designed to determine which of several possible semantic relations between novel and familiar words was most consistent with children's performance. It was found that children often seemed to interpret the new word as referring to a subordinate of the known category. This tendency was reduced when the named object could sensibly receive a proper name (e.g., when the named object was a stuffed animal), particularly when children had to consider both the familiar and the novel label for the object in the same session. Although not all alternative explanations have been ruled out, these results suggest that, from a very young age, children may spontaneously form language hierarchies when they hear a novel work for an object that already has a familiar name.     

How Two- and Four-Year-Old Children Interpret Adjectives and Count Nouns

We examined the role of object kind familiarity (i.e., knowledge of a count noun for an object) on preschoolers' sensitivity to the relation between a novel word's form class (adjective or count noun) and its reference (to a material kind-property or to an object kind). We used a forced-choice match-to-target task, in which children learned a word for one object (e.g., a metal cup), and then chose between 2 other objects. One was from the same object kind but a different material kind (with different related properties, such as color and texture; e.g., a white plastic cup); the other was from a different object kind but the same material kind (with the same related properties; e.g., a metal spoon). In Experiment 1, children learned either a count noun (e.g., "This is a zav") or an adjective (e.g., "This is a zav one"). Within each form class, we crossed the familiarity of the referent object kind (familiar and unfamiliar) with the age of the children (2- and 4-year-olds). The principal finding was that in interpreting an adjective, 4-year-olds were more likely to choose the object sharing material kind with the target if the target was familiar than if it was unfamiliar. No such familiarity effect was evident among 2-year-olds. In Experiment 2, we employed a more unambiguously adjectival frame (e.g., "This is a very zav-ish one"), and replicated the results of Experiment 1. We interpret the results in terms of 2 proposed word learning biases: one that learners initially expect any word applied to an unfamiliar object to refer to a (basic-level) kind of object, and a second that learners prefer words to contrast in meaning. We consider several interpretations of the observed age difference.     

Restricting a Familiar Name in Response to Learning a New One: Evidence for the Mutual Exclusivity Bias in Young Two-Year-Olds

Children under 2¹/₂ years old tend to interpret novel words in accordance with the Mutual Exclusivity Principle, but tend not to reinterpret familiar words this way. Because alternative principle have been proposed that only predict the novel word effects, and because tests of the familiar word effects may have been flawed, a new test was administered. In Experiment 1 ( N = 32), 24- to 25-month-olds heard stories in which a novel noun was used for an atypical exemplar of a familiar noun. When asked to select exemplars of the familiar noun, they showed a small but reliable tendency to avoid the object from the story. In Experiment 2 ( N = 16), the novel nouns in the stories were replaced by pronouns and proper names, and the children did not avoid the story object in the test of the familiar noun. Thus, the aversion to this object that was observed in Experiment I was not due to its greater exposure or its being referenced immediately before testing, but to toddlers' Mutual Exclusivity bias. Their bias is hypothesized to be a form of implicit probabilistic knowledge that derives from the competitive nature of category retrieval.     

Children's Understanding of Knowledge Acquisition: The Tendency for Children to Report That They Have Always Known What They Have Just Learned

Children's attention to knowledge-acquisition events was examined in 4 experiments in which children were taught novel facts and subsequently asked how long they had known the new information. In Experiment 1, 4- and 5-year-olds tended to claim they had known novel animal facts for a long time and also reported that other children would know the novel facts. This finding was replicated in Experiment 2, using facts associated with chemistry demonstrations. In Experiments 3 and 4, children were taught new color words. 5-year-olds, but not 4-year-olds, distinguished between novel and familiar color words, reporting they had not known the novel words before the test session, but they had always known the familiar words. 4-year-olds in Experiment 4 were better able to distinguish novel and familiar color words when the teaching of the novel words was an explicit and salient part of the procedure.     

Children''s Understanding of Action Lines and the Static Representation of Speed of Locomotion

Children's understanding of the static representation of speed of locomotion was explored in 2 experiments. In Experiment 1, 20 7-year-olds and 20 9-year-olds drew pictures of 2 people walking and running at different speeds. Children then made judgments about pairs of unambiguous drawings of a person walking or running, as did a sample of 20 adults. The drawings varied according to whether action lines, background lines, or no lines were present. Children were asked to say which figure appeared to be moving faster. In Experiment 2, 20 7-year-olds, 20 9-year-olds, and 21 adults sorted ambiguous drawings of a person walking and running at different speeds. The pictures again contained action lines, background lines, or no lines. In the drawing task, children more frequently used page position and biomechanical information than action lines to represent fast and slow walking and running. In the judgment task, 7- and 9-year-olds offered equivalent judgments of action lines and background lines, whereas adults distinguished between these pictorial devices. In the sorting task, all subjects distinguished between action lines and background lines and judged that pictures containing action lines looked faster than pictures containing background lines and pictures without lines. Taken together, the results indicate that subjects' judgments were influenced by the form of locomotion and degree of ambiguity in the depicted events they saw. The findings are consistent with the view that different categories of pictorial devices exist, but the effectiveness of each device is contingent upon the perceiver's experience with it and the context in which it appears.     

Predicting the outcomes of physical events: two-year-olds fail to reveal knowledge of solidity and support

Two-year-olds' (N = 153) knowledge of solidity was tested in four search tasks adapted from infant looking-time experiments. In Experiment 1, 2-year-olds failed to search in the correct location for a falling ball after a hidden shelf that blocked its trajectory had been inserted in the apparatus. Experiment 2 extended this finding by showing that 2-year-olds failed to take into account the effects of either removing or inserting a shelf in their search for a toy dropped behind a screen. Experiment 3 examined sensitivity to the constraint provided by a solid barrier on horizontal motion. In all three experiments, 2-year-old children searched initially at the location where they saw the object during familiarization. Experiment 4, using multiple test trials but no familiarization to a pretest location, also showed that 2-year-olds failed to take the presence or absence of a barrier into account when planning where to search for a toy they had seen dropped behind a screen. In all of these studies, 2-year-olds showed no evidence of representing solidity and support constraints on the trajectories of falling objects. Experiments 1 and 3 also included 2 1/2-year-olds (N = 31), who succeeded on these search tasks. The implications of the poor performance of 2-year-olds, in the face of success by very young infants on looking-time measures of sensitivity to similar constraints on object motion, are discussed.     

The Development of a Linkage between Count Nouns and Object Categories: Evidence from Fifteen- to Twenty-One-Month-Old Infants

Recent research suggests that, although young children appreciate many different kinds of conceptual relations among objects, they focus specifically on taxonomic relations in the context of word learning. However, because the evidence for children's appreciation of this linkage between words and object categories has come primarily from children who have made substantial linguistic and conceptual advances, it offers limited information concerning the development of this linkage. In the experiments reported here, we employ a match-to-sample task to focus specifically on the development of an appreciation of the linkage between words (here, count nouns) and object categories in infants in the period just prior to and just subsequent to the naming explosion. The results demonstrate that, for 21-month-old infants, most of whom have recently entered the vocabulary explosion (Experiment 1), and for 16-month-old infants, most of whom have yet to commence the vocabulary explosion (Experiment 2), novel nouns focus attention on taxonomic relations among objects. This is important because it reveals a nascent appreciation of a linkage between words and object categories in infants who are at the very onset of language production. Results are interpreted within a developmental account of infants' emerging appreciation of a specific linkage between count nouns and object categories.     

Acquiring Proper Nouns for Familiar and Unfamiliar Animate Objects: Two-Year-Olds' Word-learning Biases

In 2 studies, 2-year-old children learned a novel word modeled as a proper noun (e.g., "This is Zav") for an animate stuffed toy. Children who learned the word for a familiar object (i.e., one for which they knew a basic-level count noun for the kind) interpreted the word appropriately as a proper noun reliably more often than children who learned the word for an unfamiliar object (i.e., one for which they did not know such a count noun). When the creature was familiar, children typically interpreted the novel word as if it were a proper noun referring uniquely to the labeled individual. When the animal was unfamiliar, children frequently interpreted the word as if it were a count noun referring to a kind of object. Children's spontaneous comments during the tasks provided striking additional evidence that their interpretations of the proper noun varied with the familiarity of the object. The results suggest that young children's sensitivity to the form class of proper nouns is affected by the familiarity of the referent object. The findings are discussed in terms of interpretative biases in word learning.     

Basic level object categories support the acquisition of novel adjectives: evidence from preschool-aged children

Two experiments examined 3- and 4-year-old children's ability to map novel adjectives to object properties. Sixty-four children were introduced to a target (e.g., a bumpy object), and asked to choose between (1) a matching test object (e.g., a different bumpy object), and (2) a contrasting test object (e.g., a smooth object). Four-year-olds successfully extended novel adjectives from the target to the matching test object whether these objects were drawn from the same, or different, basic level categories. In contrast, 3-year-olds' extensions were more restricted. They successfully extended novel adjectives if the target and test objects were drawn from the same basic level category but failed to do so if the objects were drawn from different basic level categories (Experiment 1). However, if 3-year-olds (n = 20) were first permitted to extend a novel adjective to objects within the same basic level category, they were subsequently able to extend that novel adjective broadly to objects from different basic level categories (Experiment 2). Thus, basic level object categories serve as an initial foundation in the process of mapping novel adjectives to object properties.     

Priorities in children's expectations about object label reference: form over color

Young children show considerable interest in color similarity; thus we might expect them to use color as a basis for generalizing object labels. However, natural language observations suggest they do not: children tend to overextend labels based on form similarity and rarely, if ever, overextend labels to objects that share only color. Perhaps, then, children give form priority over color in their expectations about object-label reference. This hypothesis was investigated in 2 studies. In a first study, 40 2- and 3-year-olds sorted 10 sets of 3 pictures that contrasted form with color. Children participated in 1 of 2 conditions: half of the children were shown a target object and asked to "find another one" from among the two choice objects, while the other children heard a novel label for the target (e.g., "See this zom?") and were asked to extend the label (e.g., "Can you find another zom?"). 3-year-olds sorted almost exclusively on the basis of form in both conditions, while 2-year-olds performed differently in the 2 conditions: they frequently chose the color match when labels were absent, but selected by form rather than color when asked to extend a novel label. A second study replicated these findings with novel objects that were real rather than pictured. This study also suggested that 3-year-olds grant color some role in their decisions about object-label reference as long as form differences are reduced to a tolerable level. In sum, young children expect form to be more important than color for guiding object-label reference, even though they may find color interesting when not asked to extend labels. This expectation helps explain the speed and relative ease with which children acquire object labels.     

Young children's knowledge about visual perception: projective size and shape

Although recent research indicates that an increased sensitivity to visual appearances develops around 4 or 5 years of age, evidence from perceptual studies suggests that certain types of appearances, that is, projective size and shape, are not noticed or understood until at least 7. 4 experiments investigated preschool children's knowledge of the projective size--distance and projective shape--orientation relationships. In Experiment 1, 3- and 4-year-olds were asked whether an object should be moved farther or nearer in order to increase or decrease its apparent size. 4-year-olds performed significantly better than chance, but 3-year-olds did not. Experiment 2 showed that 3-year-olds are able to perceive projective size changes, indicating that although they do not fully understand the projective size-distance relationship, the necessary perceptual information is potentially available to them. In Experiment 3, 3- and 4-year-olds were asked to indicate how a circular object should be rotated to make it appear either circular or elliptical. Again, 4-year-olds performed significantly better than chance, but 3-year-olds did not. Again also, the results of Experiment 4 indicate that although 3-year-olds are not aware of the projective shape-orientation relationship, they are capable of attending to changes in projective shape. Thus, the constraints on children's knowledge of the projective size-distance and projective shape-orientation relationships seem to be at least partly cognitive rather than wholly perceptual. These results are interpreted as further evidence for the acquisition of level 2 percept knowledge during early childhood.     

Young children's learning and transfer of biological information from picture books to real animals

Preschool children (N = 104) read a book that described and illustrated color camouflage in animals (frogs and lizards). Children were then asked to indicate and explain which of 2 novel animals would be more likely to fall prey to a predatory bird. In Experiment 1, 3- and 4-year-olds were tested with pictures depicting animals in camouflage and noncamouflage settings; in Experiment 2, 4-year-olds were tested with real animals. The results show that by 4 years of age, children can learn new biological facts from a picture book. Of particular importance, transfer from books to real animals was found. These findings point to the importance that early book exposure can play in framing and increasing children's knowledge about the world.     

Effects of context on judgments concerning the reality status of novel entities

Three studies examined the effects of context on decisions about the reality status of novel entities. In Experiment 1 (144, 3- to 5-year-olds), participants less often claimed that novel entities were real when they were introduced in a fantastical than in a scientific context. Experiment 2 (61, 4- to 5-year-olds) revealed that defining novel entities with reference to scientific entities had a stronger effect on reality status judgments than did hearing scientifically oriented stories before encountering the novel entities. The results from Experiment 3 (192, 3- to 6-year-olds) indicated that definitions that support inferences facilitate reality status judgments more than do definitions that simply associate novel and familiar entities. These findings demonstrate that children share with adults an important means of assessing reality status.     

Young children's understanding of random phenomena

2 experiments on the development of the understanding of random phenomena are reported. Of interest was whether children understand the characteristic uncertainty in the physical nature of random phenomena as well as the unpredictability of outcomes. Children were asked, for both a random and a determined phenomenon, whether they knew what its next outcome would be and why. In Experiment 1, 4-, 5-, and 7-year-olds correctly differentiated their responses to the question of outcome predictability; the 2 older groups also mentioned appropriate characteristics of the random mechanism in explaining why they did not know what its outcome would be. Although 3-year-olds did not differentiate the random and determined phenomena, neither did they treat both phenomena as predictable. This latter result is inconsistent with Piaget and Inhelder's characterization of an early stage of development. Experiment 2 was designed to control for the possibility that children in Experiment 1 learned how to respond on the basis of pretest experience with the 2 different phenomena. 5- and 7-year-olds performed at a comparable level to the same-aged children in Experiment 1. Results suggest an earlier understanding of random phenomena than previously has been reported and support results in the literature indicating an early understanding of causality.     

Preschoolers acquire general knowledge by sharing in pretense

Children acquire general knowledge about many kinds of things, but there are few known means by which this knowledge is acquired. In this article, it is proposed that children acquire generic knowledge by sharing in pretend play. In Experiment 1, twenty-two 3- to 4-year-olds watched pretense in which a puppet represented a "nerp" (an unfamiliar kind of animal). For instance, in one scenario, the nerp ate and disliked a carrot. When subsequently asked generic questions about real nerps, children's responses suggested that they had learned general facts (e.g., nerps dislike carrots). In Experiment 2, thirty-two 4- to 5-year-olds learned from scenarios lacking pretend speech or sound effects. The findings reveal a long overlooked means by which children can acquire generic knowledge.