Douglas J. Herrmann

Only Connections: A Critique of Semantic Networks

This article examines theories that assume that semantic networks account for the mental representation of meaning. It assesses their similarities and divergencies, and argues that as a class of theories they remain too powerful to be refuted by empirical evidence. The theories are also confronted by a number of problematical semantic phenomena that arise because networks deal with the connections between concepts rather than with their connections to the world. The solution to these problems could be embodied in a new network system, but such a system would differ in both structure and function from current network theories.

The similarity and diversity of semantic relations

There is a rich variety of semantic relations in natural languages. Subjects’ perceptions of similarities among relations were studied for a wider variety of relations than had been used in previous studies. Forty subjects sorted 31 cards bearing five example pairs of each of 31 semantic relations. Subjects were able both to distinguish the relations and to perceive their similarities. A hierarchical clustering analysis of the sorting data indicated that the subjects perceived five families of semantic relations (contrasts, class inclusion, similars, case relations, and part-wholes). The five families were distinguished in terms of three properties of semantic relations: contrasting/noncontrasting, logical/pragmatic, and inclusion/noninclusion. Within each family, relations also were sorted in ways consistent with their defining properties. Relations were therefore viewed not as unanalyzable primitives, but in terms of the relational properties that distinguished them.

An empirical taxonomy of part-whole relations: Effects of part-whole relation type on relation identification

Abstract A taxonomy of part-whole (meronym) relations was developed (Experiment 1). Subjects sorted examples of relations and named each relation with a part-term, e.g. component, member, portion The resulting empirical taxonomy distinguished three major types of meronymy: part-whole (cup-handle), stuff (cup-china), and phase (growing up-adolescence). The part-whole relations were further subdivided into eight types: integral object-component (car-wheel), event-feature (circus-trapeze act), topological part-area (room-corner), collection-member (forest-tree), area-place (desert-oasis), time-occasion (February-Valentines Day), measure-unit (mile-yard) and mass-portion (pie-slice). Relations adjacent in the taxonomy tended to be named with the same part-term. In Experiment 2 subjects made yes/no decisions about word pairs in answer to the question, “Is A part of B?” Types of meronym pairs were presented in blocks Responses were slower at the start of a new block. This result indicated that the type of mero...

Similarities and Differences in Meaning in Six Cultures

The present study explores use of multidimensional scaling to cross-cultural investigation involving six cultures and two categories varying in abstractness: animals and emotions. Cross-cultural agreement in the scaling solutions was quite high but highest for the Emotions category. Additionally, the solutions of both categories across cultures revealed details that reflected culture-specific meaning. Thus, the results demonstrate the utility of multidimensional scaling analysis for the cross-cultural investigation of meaning.

The proper role of clusters in mathematical models of continuous recall

Abstract Previous research has shown that the number of words cumulatively recalled ( N ) at time ( t ) is a negatively accelerated function that reaches an asymptote as t → ∞. Research has also shown that the increase in N with t occurs in bursts or clusters. Several models purport to account for this cumulative recall curve in terms of cluster characteristics. The present research shows that previous models have not in fact successfully linked continuous recall to cluster characteristics. This research demonstrates that cluster models need to employ three empirical characteristics of clusters: T b , the time between clusters; T w , the average time between words within a cluster; and W c , the number of words within a cluster. It is shown that these three quantities determine the cumulative recall curve, and these three quantities may in turn be characterized by four parameters. Of these four parameters, only three actually characterize the cumulative recall curve. Two parameters determine the initial slope and final asymptote of the curve, while a third parameter, which we introduce for the first time, characterizes the curves shape. This latter parameter may be interpreted as the ratio ofthe time spent in retrieving and discarding a cluster that has been previously recalled to the amount of time spent in retrieving and outputting a newly encountered cluster. It is pointed out that previous success in fitting the cumulative recall data with a two-parameter function may be explained by the fact that this parameter lies in a restricted range about unity. Further experimental work is suggested to elucidate the behavior of this new parameter. Two models are then proposed to account for these characteristics of clusters and the shape of the recall curve.

Differences in the Search Rate of Children and Adults in Short-Term Memory.

Abstract When people recognize a test item as belonging or not belonging to a previously presented set, recognition latency increases with the number of items in the set. Although some evidence suggests otherwise, it is currently held that the rate of this increase is the same for children and adults. In contrast, the present experiments indicated a much slower search rate for second graders (mean age = 7.33 years) than for seventh (mean age = 12.50 years) and twelfth graders (mean age = 17.25 years). Moreover, search rate for second graders was invariant under instructions which did or did not emphasize speed and with presentation of the memory set in either the auditory or visual modality. Experiment II showed that the slow search rate for children was not due to differences in encoding between children and adults.

The Role of Category Size in Continuous Recall from Semantic Memory

Summary The time to recognize a word as an exemplar of a category has been found to be little influenced by the size of the category. In contrast, three recall studies have indicated that the rate of continuous recall of exemplars of a category is a function of the categorys size. However, these recall studies used large categories which dealt with topics different from the small categories, confounding size effects with differences in stimulus meaning. The present study investigated the hypothesis that category size affects the rate of continuous recall while controlling the semantic characteristics of large and small categories. Thirty-six college students generated examples for one or the other member of 14 nested category pairs: e.g., Plants, Trees. The results showed first that the instantaneous rate of recall (words/sec) at any time was greater for large categories than for small categories except when t = o. Second, the functions time constant (defined as the time to approach producing approximat...

TASK APPROPRIATENESS OF MNEMONIC TECHNIQUES

The present article reviewed research in which two or more mnemonic techniques in the paradigms of either free recall, paired-associate, or serial learning have been compared. The review showed that the most effective mnemonic differed across the three paradigms: imagery mediation for paired-associate learning, the story mnemonic for free recall learning, and the method of loci for serial learning. Each mnemonic varies in its suitability for different paradigms, and each paradigm is facilitated more by the use of certain mnemonics than by others. These conclusions support the view that memory processing differs across memory tasks. They also indicate that differences in memory performance across tasks are due not only to differences in basic processes (such as recognition and perception) elicited by tasks but also to the task appropriateness of control processes applied to a task.

Number of Available Associations and Rate of Association For Categories in Semantic Memory

Summary Bousfield and Sedgewick showed that the cumulative number of associations produced to a category as a function of time could be described by an exponential function having the parameters m (rate of association) and C (total number of associates produced). In two experiments using male and female college students (N = 16 and N = 14, respectively), the present study investigated the hypothesis suggested by previous research, that the values of m and C for different categories are negatively correlated. S s produced examples of different categories in four-minute periods per category on mimeographed sheets. The predicted negative correlation across categories was found (Experiment I, r = .54, p < .05, one-tailed test; Experiment 2, r = .75, p < .01). Categories with a low m value had a high C value and vice versa. The effect can be explained by a model which assumes exemplars are stored more or less densely in semantic memory.

Memory Contrivances in Everyday Life

Phenomena of human memory traditionally have been explained in terms of memory processes. However, memory performance and the claims that people make about memory performance may also reflect social processes such as contriving to make a memory act appear different from what actually occurred The present research investigated the extent of the use of memory contrivances as reported on a questionnaire, Memory Use in Social Environments (MUSE). Factor analysis of the responses of the 92 subjects to the MUSE items identified the following eight forms of contrivances (listed in order from most to least used): ignoring anothers remembering, excuses, self-deprecation self-enhancement, cover-ups for self-forgetfulness, politeness, insults and misevaluation. In a second study, 43 subjects identified the social goals underlying the eight contrivances. These results indicate a role for social processes in memory performance not recognized previously; the results also suggest specific ways in which cognitive performances can be represented in social interaction. Thus, the phenomena of memory contrivances have implications for the confluence of memory theory and social psychological theory in cognition.