Elizabeth F. Shipley

The emergence of the child as grammarian

Abstract Demonstrations of some young childrens awareness of syntactic and semantic properties of language are presented. Rudiments of such ‘meta-linguistic’ functioning are shown in two-year olds, who give judgments of grammaticalness in a role-modelling situation. The growth of these abilities is documented for a group of five to eight-year old children, who are asked explicitly to give judgments of deviant sentences. Adult-like behavior, in these talented subjects, is found to emerge in the period from five to eight years. Possible relations of meta-linguistic functioning to other ‘meta-cognitive’ processes are suggested.

The class-inclusion task: Question form and distributive comparisons

The class-inclusion task is regarded by Piaget as a measure of the childs mastery of the structure of hierarchical classification. Class-inclusion was improved by changing the wording of the question to conform to standard English usage. A theoretical argument is offered that the childs difficulties with this task derive from confusion of collective comparisons, in which properties of classes are compared, and distributive comparisons, in which properties of elements are compared. A grammatical constraint on expression of distributive comparisons—an element of a class cannot be compared to an element of an included subclass—is hypothesized to be overgeneralized to expressions referring to collective comparisons such as the class-inclusion task. This hypothesis accounts for the improvement in class-inclusion performance with changes in wording of the question and for the finding that young childrens responses to class-inclusion questions and to ungrammatical requests for comparison of an element of a class and an element of an included subclass are similar: the children respond readily but understand wrongly that the comparison involves coordinate classes.

Induction from a Single Instance: Formation of a Novel Category.

This study examines whether preschoolers can use information from a known category to induce a characteristic attribute of a novel, contrasting category based on a single instance. We showed 32 four-year-olds three instances of a Given Category and one instance of a Target Category. These objects could vary along two attribute dimensions, such as color and shape. All instances of the Given Category shared identical values of one attribute (e.g., all were blue), but could have different values of the other attribute (e.g., a circle, a square, and a triangle). The single instance of the Target Category was different from the Given on both attribute dimensions (e.g., a red diamond). Children gave yes/no judgements as to whether additional objects were instances of the Target Category. There were two possible sources of information about the relevance of an attribute to classification: explicit (labeling) and implicit (variation in the Given Category). There were four conditions such that each source of information was either available or not. Both types of information were effective in eliciting inductions of the relevant kind of attribute and the characteristic value of this attribute in the novel category (explicit: p = .0004; implicit: p = .031). This suggests that children use an inductive bias that the instances of two related but distinct categories tend to be alike in the same way.

Detection and recognition: Experiments and choice models ☆

Abstract An experiment is described in which each of 3 observers participated in 12 different detection and recognition tasks; relevant choice models based on Luces work are examined. The stimulus parameter from a two-alternative forced-choice task is shown to be related to the stimulus parameter from the corresponding yes-no detection task by a distance representation. An assumption is made to relate stimulus parameters from recognition tasks to stimulus parameters from detection tasks; again a distance representation is used and the predicted relation is supported by the data. Several extensions of the choice models to composite tasks that require both recognition and detection are examined. Multistage choice models in which recognition occurs first and is followed by detection are judged most adequate. Data from tasks with uncertainty in one of two aspects of the stimulus are compared with data from composite tasks to evaluate the assumption that covert responses actually occur in the former tasks and influence the overt responses. With one exception, response proportions summed over the irrelevant response from the composite tasks are similar to response proportions from the tasks with uncertainty. The exception, recognition when no signal is presented on one-half the trials, yields response proportions similar to those found in the simple recognition task. In general, the covert response assumption appears to be supported whenever both stages of choice are required.

A constraint on comparisons: Equally detailed alternatives

Abstract Childrens errors in collective comparison tasks are atributed to the comparison of the wrong classes. It is specifically hypothesized that alternative classes are erroneously formed subject to the constrain that the same kinds of properties are criterial for each alternative. This hypothesis, called the equally detailed alternatives hypothesis, is tested in three experiments with hierarchically organized stimuli and requests to compare a superordinate class and a nonincluded subclass, for example, dogs and yellow cats. In all three experiments 4-year-old childrens comparisons were found to be in accord with the hypothesis. In the first experiment both perceptual and linguistic factors determine which classes are compared. In the second experiment, as predicted, erroneous subclass comparisons were more common when all subclasses were distinguished by the same kinds of properties. In the third experiment, the children were asked to partition the stimulus objects into the classes to be compared. The vast majority of partitions were erroneous and in accord with the hypothesis. Potential benefits of the constrain are considered.

Labelling patterns and object naming

We examined the effects of different labelling patterns on the generalization of object names. Two-year-olds, three-year-olds and adults were shown two ‘standard’ objects, which were named with the same label, or with two different labels, or with no label at all. Participants were then asked whether objects morphed to be intermediate to the standards belonged to one of the labelled categories or, in the No Label condition, were ‘like’ one of the standards. The Same Label condition showed generalization to all intermediates, whereas the Different Label and No Label conditions showed division of the intermediates into two separate categories, with somewhat sharper division under Different Label. These results suggest two possible mechanisms of lexical learning: ‘boosting’ the equivalence of different exemplars through label identity, and ‘differentiating’ the exemplars through differences in labelling. The studies provided strong evidence for boosting. Learners are sensitive to the distribution of labels across exemplars, and they hold powerful assumptions about the relationship between these distributions and the underlying naming space. These findings have implications for the early emergence of cross-linguistic differences in lexical learning.

A signal detection theory analysis of a category judgment experiment

A category-judgment experiment is the reported. Rating-scale judgments of loudness were made with five or eight intensity levels of a tone and with different frequencies of presentation of the stimuli. Obtained ROC curves for pairs of stimuli were judged to support application of signal-detection theory. However, certain common assumptions of the theory are called into question by the data. Discrimination was poorer when more categories were available. This is attributed to an increase in criterion variance with an increase in the number of criteria. The slope of the ROC curves for different stimulus pairs suggested that stimulus variance is neither constant nor does it necessarily increase with stimulus intensity It is argued that stimulus variance may depend upon the distribution of internal observations over all trials.

The what-if of counting

In comments on our counting paptir (Shipley & Shepperson, 1990) Gallistel and Gelman (1990) express their reservations about the possibility of gene:ral dispositions accounting for the child’s mastery of counting. We share their doubts about & power of general dispositions alone, but feel a reasonable case cam; be made for general dispositions plus a competency specific to the domain of numerosity. What if, in addition to general dispositions, babies and young children have the ability to directly perceive numerosity, to subitize, before they learn to count? Several types of data are consistent with the existence of subitizis,J in precounters. Habituation studies indicate that intints can detect differences between two numerosities in the subitizing range of one to four entities, but not larger numerosities (Starkey & Cooper, 3980; Starkey, Spelke, & Gelman, 1988; Strauss & Curtis, 1984). In addition, some young children apparently recognize the cardinality of small sets before they can count the members of the sets, that is, they use a count term correctly to describe the set before they can rote count to that term. For instance, one of us observed that two of her own children used two correctly for sets of two objects, such as cookies, before they could count to two. Another young child we know used both two and three correctly before he could count at all. Wagner and Walters (1982) report similar instances. owever, the status of subitizing is highly controversial. Recently, the existence of subitizing has been eloquentiy denied (Gallistel, 1988) and supported (Davis & Perusse, 1988: 603-604). We find the two positions equally compelling and hence think it wcrth while to explore the consequences of as’. Jming that precounters can subitize. ;n our earlier paper we proposed that general dispositions lead the child to imitate relevant aspects of another person’s counting performance. They