Using Speakers' Referential Intentions to Model Early Cross-Situational Word Learning

ABSTRACT— Word learning is a "chicken and egg" problem. If a child could understand speakers' utterances, it would be easy to learn the meanings of individual words, and once a child knows what many words mean, it is easy to infer speakers' intended meanings. To the beginning learner, however, both individual word meanings and speakers' intentions are unknown. We describe a computational model of word learning that solves these two inference problems in parallel, rather than relying exclusively on either the inferred meanings of utterances or cross-situational word-meaning associations. We tested our model using annotated corpus data and found that it inferred pairings between words and object concepts with higher precision than comparison models. Moreover, as the result of making probabilistic inferences about speakers' intentions, our model explains a variety of behavioral phenomena described in the word-learning literature. These phenomena include mutual exclusivity, one-trial learning, cross-situational learning, the role of words in object individuation, and the use of inferred intentions to disambiguate reference.     

On some worldly worries

Israel Scheffler and others have had trouble accepting such drastic theses in my work as that worlds, even old ones, are made by right versions, even new ones, and that two conflicting versions may both be right. But further explication shows how such theses have advantages over the more usual common-sense alternatives.Published by the author at Emerson Hall, Harvard University, Cambridge, Massachusetts, 02138, U.S.A., on 1 September 1988.NOTICE: This paper may be copied and circulated with due credit but without charge. (It was accepted by theJournal of Philosophy, but withdrawn by that journal on the eve of publication because of a dispute over rights.)     

On scaling models applied to data from several groups

Statistical methods are presented to facilitate a more complete analysis of results obtained when a scaling model is applied to data from two or more groups. These methods can be used to (a) compare the corresponding estimated latent distributions obtained using the scaling model applied to the different groups, (b) compare the corresponding estimated item reliabilities (or item response error rates) for the different groups, and (c) test whether the scaling model applied to the several groups can be replaced by a more parsimonious scaling model that includes various homogeneity constraints (i.e., constraints that describe which parameters in the model are the same for the several groups). Various kinds of scaling models are considered here in the multiple-group context.Support for this research was provided in part by the National Science Foundation, to Clogg by Grant No. SES-7823759 and to Goodman by Grant No. SES-8303838. Clogg and Goodman were Fellows at the Center for Advanced Study in the Behavioral Sciences when part of the research was done, with financial support provided in part by National Science Foundation grant BNS-8011494 to the Center. The authors are indebted to Mark P. Becker and James W. Shockey for helpful comments.     

Cognitive processes in verbal-number production: inferences from the performance of brain-damaged subjects

This article presents a model of the cognitive processes involved in the spoken production of verbal numbers (e.g., thirteen thousand four hundred two). On the basis of single-case studies of two brain-damaged subjects with number production deficits, we argue that verbal-number production involves the generation of a syntactic frame that constitutes a plan for the production of the appropriate sequence of words. The syntactic frame specifies each to-be-retrieved word in terms of a number-lexical class (i.e., ones, teens, or tens) and a position within that class. These class/position-within-class specifications guide the retrieval of lexical representations from a production lexicon that is partitioned into functionally distinct ones, teens, and tens classes. We conclude with a brief discussion of the rationale for, and advantages of, using patterns of impaired performance as a basis for drawing inferences about normal cognition.     

DEVELOPMENT AND EVALUATION OF AN INFANT-CARE TRAINING PROGRAM WITH FIRST-TIME FATHERS

We evaluated the effectiveness of a multicomponent package in training infant-care skills to first-time fathers. After developing and socially validating a set of infant-care skills, we assessed the effects of training in a hospital-based program with expectant fathers (Experiment 1) and in a home-based program with fathers having varied degrees of experience with their infants (Experiment 2). In both experiments, a multiple probe design demonstrated that the training package was responsible for producing criterion performance by the expectant and first-time fathers. A 1-month generalization probe in Experiment 1 showed that the effects transferred across training conditions (training doll to human infant) and settings (hospital to home). An increase in the number of infant-stimulation activities performed by fathers was also observed in both experiments.     

Social framework testimony in employment discrimination cases

The use of social framework testimony by social psychologists in employment discrimination cases is expanding. After reviewing the legal background of this type of testimony, we describe its content and discuss an important Title VII case of sex discrimination where the testimony played a critical role. We also distinguish social framework testimony from other types of expert testimony by psychologists.     

A framework for assessing program institutionalization

This article presents a framework for assessing the institutionalization of programmatic innovations in nonprofit community agencies and in schools. Institutionalization is the attainment of long-term viability and integration of innovations in organizations. Institutionalization is often characterized as the final stage in an innovation diffusion process. The framework was derived from a qualitative study of ten health promotion programs that were innovations in their host organizations. The framework is a two dimensional matrix: one dimension consists of organizational subsystems; the other consists of levels of institutionalization termed passages, routines, and niche saturation. The cells of the matrix are the basis for assessing program institutionalization. Robert M. Goodman is a research assistant professor in the Department of Health Behavior and Health Education at the University of North Carolina's School of Public Health. Currently, Dr. Goodman directs a four-year intervention study on the dissemination of tobacco prevention curricular funded by the National Cancer Institute. Allan Steckler is an associate professor in the Department of Health Behavior and Health Education at the University of North Carolina's School of Public Health. Dr. Steckler is also the principal investigator of the National Cancer Institute tobacco prevention study.     

Perinatal complications in births to low socioeconomic status schizophrenic and depressed women.

Pregnancy and birth complications in births to 57 schizophrenic, 28 depressed, and 31 well women were studied. The sample was of low socioeconomic status and predominantly African-American. The study extended earlier work on the perinatal status of infants born to schizophrenic women by including measures of severity of maternal disturbance, mother's age, IQ, and premorbid social competence, and family composition. The results show that maternal competence and the mother's diagnosis of schizophrenia were significant variables in determining the likelihood of less adequate prenatal care and more complicated births. The results indicate the importance of an assessment not only of a disturbed woman's diagnosis but also of her personal background and social competence in determining the likelihood of obstetrical complications.     

Exposure to community violence and social support as predictors of anxiety and social and emotional behavior among African American children

We examined the effects of exposure to violence and social support on self-reported state and trait anxiety and parental rated problem behaviors among school-aged African-American children in low and high violence areas. Ninety-seven (97) fourth through sixth graders and their parents were interviewed about exposure to community violence, social support, and state and trait anxiety. Results indicate that trait anxiety was correlated positively with exposure to violence and negatively with social support. State anxiety was correlated negatively with family social support. Problem behaviors were correlated negatively with family support and income. Hierarchial multiple regression analyses indicated that children's reports of their exposure to community violence continued to explain significant variability in trait anxiety, and problem behavior after controlling for both income and social support measures. Social support from peers, family and teachers played differential roles in predicting problem behaviors among children from high and lower violence areas. These findings suggest that in planning intervention programs for children exposed to violence, greater attention to empowering parents to support their children, to fostering peer group support, and to bolstering teacher support may be useful, but attention to the underlying socio-political causes of violence exposure is essential.