The structure of infant intelligence: An historical overview

An historical overview of investigations that were concerned with delineating the structure and composition of infant intellectual activity is presented. The events which both preceded and culminated in the conceptualization of infant intelligence as a unitary trait, or factor, are traced. Thereafter, recent studies are reviewed which have guided contemporary investigators to contend that infant intellectual activity is comprised of varied sets of relatively independent abilities and factors. Several conclusions are made concerning the points of view which various investigators have held regarding the structure of infant intelligence.     

Isolating the components of intelligence

In a recent volume (Sternberg, 1977b), I proposed a method called componential analysis that provides a means to isolate the components of intelligent performance. The method was described in detail, and then tested in a series of experiments. But an important question was left unanswered by this early work: Is componential analysis generalizable to tasks other than analogies? More recently, I have been investigating the generalization of the methodology to other tasks, and have also been engaged in extending the methodology in order to increase its flexibility of application. The present article (a) briefly summarizes the structure of a componential analysis, (b) describes the method of precuing, the method originally used for isolating the components of intelligence, and (c) describes new methods that have also been successful in isolating these components in a variety of reasoning tasks.     

Computing with connections in visual recognition of origami objects

This paper summarizes our initial foray in tackling Artificial Intelligence problems using a connectionist approach. The particular task chosen was the visual recognition of objects in the Origami world as defined by. The two major questions answered were how to construct a connectionist network to represent and recognize projected line drawings of Origami objects and what advantages such an approach would have. The structure of the resulting connectionist network can be described as a hierarchy of parameter or feature spaces with each node in each of the feature spaces representing a hypothesis about the possible existence of a specific geometric feature of an Origami object. The dynamic behavior of the network is a form of iterative refinement or relaxation whose major characteristic is to prefer more globally interesting interpretations of the input over locally pleasing ones. Examples from the implementation illustrate the system's ability to deal with forms of noise, occlusion and missing information. Other benefits are an inherently parallel approach to vision, limitation of explicit ordering of the search involved in matching model to instance and the elimination of backtracking due to the sharing of partial results as the search progresses. Extensions and problems are also discussed.     

Cerebral asymmetry in the perception of American sign language.

American Sign Language (ASL) offers a valuable opportunity for the study of cerebral asymmetries, since it incorporates both language structure and complex spatial relations: processing the former has generally been considered a left-hemisphere function, the latter, a right-hemisphere one. To study such asymmetries, congenitally deaf, native ASL users and normally-hearing English speakers unfamiliar with ASL were asked to identify four kinds of stimuli: signs from ASL, handshapes never used in ASL, Arabic digits, and random geometric forms. Stimuli were presented tachistoscopically to a visual hemifield and subjects manually responded as rapidly as possible to specified targets. Both deaf and hearing subjects showed left-visual-field (hence, presumably right-hemisphere) advantages to the signs and to the non-ASL hands. The hearing subjects, further, showed a left-hemisphere advantage to the Arabic numbers, while the deaf subjects showed no reliable visual-field differences to this material. We infer that the spatial processing required of the signs predominated over their language processing in determining the cerebral asymmetry of the deaf for these stimuli.     

Self- and other regulation during the infancy period

The “other regulation” views of Lev Vygotsky and of Jerome Bruner are contrasted with Jean Piaget's self-regulatory model of infant development. Examples of adult behavior in Piaget's writings and in mother-infant interactions suggest that adults perform behaviors which are analogous to Piaget's type a and b self-regulation. Data from infant games are provided to illustrate the processes by which mothers regulate the environments for their infants and the consequences of such maternal regulation for infant development.     

Error detection processes in statistical problem solving

Problem solvers' error detection processes were studied by instructing 16 subjects to think aloud when solving two statistical problems. The evaluative episodes occurring in subjects' protocols were analyzed into Affirmative evaluation, Direct error-hypotheses, Error suspicion, and Standard check episodes, the last three of which are assumed to cover all main types of error detection processes. Most errors (78%) were found to have contributed to a solution part that triggered some evaluative episode. However, only one-third of the undetected errors had contributed to such a solution part. The Standard check episodes, seen as centrally-invoked, only led to the detection of few errors in proportion to the number of times they were performed. Evidence was found for two types of spontaneous error detections, one occurring abruptly and the other as a result of a more elaborated error detection process, initiated by the solver perceiving the solution as dissatisfying or strange. The perception of a symptom was a fairly reliable source of information about errors. However, subjects often did not manage to detect the error after having noticed a symptom. The closer a relevant Error suspicion episode followed an error, the greater was the probability of detecting the error. Good problem solvers detected a higher proportion of their errors compared to poor problem solvers, probably due to differences in the processes leading up to the triggering of an evaluative episode rather than to differences after the episode had been triggered.