Filling gaps: Decision principles and structure in sentence comprehension

The correct grammatical characterization of sentences containing filler-gap dependencies is a topic of considerable theoretical interest in linguistics. In some grammatical frameworks, these dependencies are representef in terms of conditions on the permissible indexing of structures (or alternative structure evaluation conditions) which a representation must adhere to in order to be well-formed. In other frameworks, constraints on permissible filler-gap dependencies are simply inherent in the set of phrase structure rules contained in the grammar of a language.The processing of sentences with multiple (potential) filler-gap dependencies was investigated in two experiments. The first experiment provided evidence for three claims. First, the human sentence processor abides by a strategy of assigning the most recent potential filler to a gap. Hence, ‘recent filler’ sentences where this assignment proves to be correct takes less time to comprehend than ‘distant filler’ sentences where this decision turns out to be incorrect. Second, the recent filler strategy is itself just a special case of a more general strategy of assigning the most salient potential filler to a gap. Third, unambiguous sentences in which a filler-gap assignment is disambiguated by ‘control’ information specified by individual verbs gives rise to the same recent filler errors as ambiguous sentences. This suggests that tentative filler-gap assignments are made by the processor before all of the relevant constraints on permissible filler-gap dependencies are consulted by the processor.The second experiment tested an alternative hypothesis that the more complex ‘distant filler’ sentences took longer to comprehend in the first experiment only because these sentences often contained verbs which license two adjacent gaps. The experiment showed that there was a significant recent filler effect in sentences that did not contain adjacent gaps and that this effect did not interact with verb class.The finding that the processor delays use of verb-control information is extremely surprising. It may be explained by the fact that this information is only relevant to one type of gap (‘equi-gaps’) and what type of gap the processor is dealing with often can not be determined unambiguously at the time when it initially encounters a gap in its left-to-right processing of a sentence.If our interpretation of these findings is correct, they argue for a considerable amount of structure in the sentence comprehension system. Further, they favor a view of sentence processing in which processing operations involving constraints on the permissible indexing (or evaluation) of structures lag behind the processor's structure building operations. Hence, the results favor those grammatical theories which preserve this distinction over grammatical theories which provide a uniform characterization of all syntactic well-formedness conditions.     

Integrating information across eye movements

A series of experiments was conducted in which a word initially appeared in parafoveal vision, followed by the subject's eye movement to the stimulus. During the eye movement, the initially displayed word was replaced by a word which the subject read. Under certain conditions, the prior parafoveal word facilitated naming the foveal word. Three alternative hypotheses were explored concerning the nature of the facilitation. The verbalization hypothesis suggests that information acquired from the parafoveal word permits the subject to begin to form the speech musculature properly for saying the word. The visual features integration hypothesis suggests that visual information obtained from the parafoveal word is integrated with foveal information after the saccade. The preliminary letter identification hypothesis suggests that some abstract code about the letters of the parafoveal word is stored and integrated with information available in the fovea after the saccade. The results of the experiments supported the latter hypothesis in that information about the beginning letters of words was facilitatory in the task. The other two hypotheses were disconfirmed by the results of the experiments.     

Asymmetries in the perceptual span for Israeli readers

Native Israeli readers read Hebrew and English text as their eye movements were monitored. A window of text moved in synchrony with their eye movements and the window was either symmetrical about the fixation point or offset to the left or right. When subjects were reading Hebrew, the perceptual span was asymmetric to the left and when they were reading English it was asymmetric to the right. The results point out the importance of attentional factors in reading.     

Long-term memory and retrieval at ages 2, 3, 4

Two experiments are reported examining the contents and accessibility of a subset of the knowledge represented in long-term memory by very young children. In Experiment 1, children aged 2 to 4 were asked in a production task to name items commonly found in particular rooms in a house, and then to verify the appropriate locations of a set of miniature replicas of household items in a dollhouse. In Experiment 2, object context was manipulated to observe if further external context improves or changes production. Children 3 and 4 years old were asked to produce appropriate room items under one of three conditons: either presented no room objects as in the first experiment, or given room objects highly- or less-frequently associated in the first study.Very young children revealed considerable knowledge in this domain, and even at the youngest ages, core defining information was retrieved. It was equally clear, however, that young children know much more than they can produce, that there is noticeable improvement in ability to retrieve information over this age range, and that additional external object context does not necessarily facilitate retrieval from long-term memory.     

The effects of sound duration on newborns' head orientation

Two experiments assessed the importance of sound duration for eliciting head orientation responses from newborn infants. In Experiment 1, thirty infants turned with equal frequency toward 20-s continuous rattle sounds and 20-s trains of rattle segments. The duration of the rattle segments--14 and 100 ms (2/s), or 500 ms (1/s)--did not influence the likelihood of turning. Response latencies and durations proved quite similar for all stimuli. In Experiment 2, twenty-four infants heard continuous rattle sounds of four different durations: 1, 5, 10, and 20 s. They turned reliably to all stimulus durations; furthermore, the magnitude and temporal characteristics of head orientation responses did not differ for the four stimulus durations. These results suggest that the newborn's head orientation response may reflect a motor program that is initiated by auditory input and then executed in a similar fashion regardless of further stimulation.     

Effects of eye contact and social status on the perception of a job applicant in an employment interviewing situation

The present study examines the independent and joint effects of variations in eye contact and social status on the way in which job applicants tend to be perceived in employment interviewing situations. The key dependent variable in the analysis is the interviewer's perception of how well the applicant is suited for a job requiring or involving self-confidence. The results, based on a laboratory analogue of an employment interview, show statistically significant main effects for eye contact and social status, as well as a significant two-way interaction effect. Effects of the manipulations on job-specific ratings are also reported, and implications of the research for biases in employment interviewing are discussed.     

Syntactic presupposition in sentence comprehension

Two experiments investigated the role of syntactic presupposition in sentence comprehension. In Experiment I subjects verified cleft, pseudocleft and factive complement sentences with respect to preceding context paragraphs, which contradicted either the assertion or the presupposition of the target sentence. Subjects took significantly longer to verify sentences with false presuppositions than sentences with false assertions. In Experiment II subjects verified cleft and pseudocleft sentences with respect to subsequently presented pictures. Once again, verification times for sentences with false presuppositions were significantly longer than verification times for sentences with false assertions. It was argued that these findings are more adequately explained by a “structural” hypothesis, than in terms of strategies designed to locate given and new information.     

Recognition memory for objects in very young children: The effect of shape and label similarity on preference for novel stimuli

In two studies, 88 children 12 to 38 months of age were briefly shown 24 common objects on familiarization trials. On subsequent test trials, these objects reappeared paired with novel objects. On one-third of the test trials pairs were similar in shape, on another third of the test trials pairs shared a common label, and on the remaining test trials pairs were unrelated. Looking times for each of the two objects on test trials were measured. All but the youngest age group showed significantly longer looking times for novel than for familiar objects for all three kinds of test pairs. However, for 19- to 31-month-olds preference for novel stimuli in the common Label Condition, although significant, was less than preference for novel stimuli when shape was similar or when unrelated stimuli were paired. Recognition in the oldest group was unaffected by condition. The results suggest a developmental period during which label or conceptual similarity has a special significance for recognition, and are discussed within theoretical frameworks concerning the very young child's developing semantic and categorization skills.     

The influence of attention upon anticipatory arousal,habituation, and reactivity to a noxious stimulus

Forty subjects were assigned to four groups, an External-attending Group, an Internal-attending Group, a Control Group, and a Distracted Group. All groups were presented with six trials of a 0.5-sec 110-db white noise. A measure of heart rate deceleration verified the experimental manipulation of attention. During the anticipatory period, the Internal-attending Group exhibited the greatest and the Distracted Group exhibited the least reactivity among the four groups. The two specially attending groups exhibited greater galvanic skin-response conditioning during the anticipatory period than the other groups. There was no support for the hypothesis that heightened attention facilitates habituation. Two measures were differentially sensitive to direction of attention during the anticipatory period. The External-attending Group exhibited reliably greater anticipatory deceleration of heart rate than the Internal-attending Group, while the Internal-attending Group exhibited reliably more nonspecific electrodermal responses than the External-attending Group. This latter finding, along with other evidence, suggests that an inward direction of attention tends to exacerbate anticipatory anxiety as well as reactivity to the impact of a noxious stimulus.     

Size-distance perception in preschool children.

Children between three and six years of age matched the “apparent” and “real” size of familiar and unfamiliar objects 3, 6, or 9 feet away. Prior to the experimental sessions, the children were divided into two groups: (a) those who could distinguish the phenomenal from the real sizes of the arcs in the Jastrow illusion (the “Realists”) and (b) those who could not (the “Phenomenalists”). The results suggest that all children perceived size constancy up to distances of 9 feet solely on the basis of visual information.     

Interactive processes in sentence comprehension and recognition

Two experiments examined the roles played by semantic and surface information in reading and recognizing sentences. Subjects read sentences in normal and inverted typography. Their recognition of meaning and other sentence features was tested using sentences whose typography, wording, and/or meaning were either the same as or different from that in the first set of sentences. In Experiment 1, subjects either read aloud or performed a sentence continuation task. For originally inverted sentences, recognition of meaning was high, irrespective of task demands. For originally normal sentences, recognition was low for Read Aloud subjects and high for Sentence Continuation subjects. Sentence recognition was affected by repetition of wording and typography. Experiment 2 replicated the results with the read aloud task and showed the second reading of originally inverted sentences to be equally swift for paraphrase and verbatim test forms. It was concluded that reading and recognition are interactive processes, involving conceptually driven and data driven operations. The interaction of operations may be either automatic or controlled. While processing of normal typography is automatic, inverted typography induces controlled processing, resulting in better retention. Furthermore, semantic and surface information are conceptualized as interacting components of comprehension and memory processes.     

The Ringelmann effect: Studies of group size and group performance

Ringelmann's classic finding—that the addition of co-workers in a rope-pulling task leads to a linear decrement in the individual group member's average performance—was reexamined experimentally. Study I attempted to replicate the effect, using groups of subjects ranging in size from 1 to 6. Performance dropped significantly as group size was increased from one individual to two or to three, but the addition of a fourth, fifth, or sixth member produced insignificant additional decrements; thus, the effect was not linear but curvilinear. Study II was designed to examine possible sources of performance loss, separating the factors of “coordination” and “motivation” loss (Steiner, 1972). The possibility of intermember incoordination was eliminated, while motivation loss remained free to vary: Each experimental subject pulled alone, and in “groups” where he believed there were from one to five other members. Once again, individual performance declined significantly with the addition of the first and second perceived co-worker, but then leveled off for perceived group sizes three to six. Some implications are discussed.     

Stimulus-reinforcer and response-reinforcer relations in the control of conditioned appetitive headpoking (“goal tracking”) in rats

Experiment 1 studied the effect of several Pavlovian appetitive conditioning procedures on rats' headpoking into a food tray (goal tracking). The procedures included forward delay conditioning, CS-alone extinction, differential conditioning, and simultaneous compound conditioned inhibition training. In general, the headpoke behaved in all of these treatments much like a Pavlovian CR; however, one could also say that the headpoke behaved like an adventitiously reinforced operant for which the CS was an S^D. Experiment 2, therefore, used the differential-nondifferential technique (E. Gamzu & D. R. Williams, Science, 1971, 171, 923–925), and Experiment 3 used an omission technique (F. D. Sheffield, in W. F. Prokasy, Ed., Classical conditioning, New York: Appleton-Century-Crofts, 1965; D. R. Williams & H. Williams, Journal of the Experimental Analysis of Behavior, 1969, 12, 511–520), to try to separate the role of stimulus-reinforcer and response-reinforcer relations in controlling the headpoke. These techniques proved inadequate. The results of Experiment 2 could be given either operant or Pavlovian interpretations. Those of Experiment 3 showed that headpoking is dominated by response-reinforcer, rather than by stimulus-reinforcer, relations when the two compete but forced no conclusion about which dominates when the two act together as in acquisition. Implications for pigeon autoshaping are discussed.     

Toward the analysis of close relationships

Two central questions are addressed: (1) How does one distinguish among relationships differing in closeness at any single point in time? (2) How do relationships change over either a short or a long time span? The first question suggests looking inside the Person-Other “intersection.” Findings from several empirical studies are described. The second question leads to discussions of short-term and long-term temporal processes. Regarding long-term pair processes, it is proposed that research can be stimulated by recognizing a five-phase sequence which extends from (a) initial attraction, to (b) building a relationship, (c) continuation, (d) deterioration, and (e) ending; transitions between adjacent phases are considered. The importance of theoretical models is emphasized throughout.     

Role of referent concreteness in the acquisition of passive sentence comprehension through abstract modeling

The present experiment tested the hypothesis that development of syntactic comprehension through verbal modeling is enhanced by referent concreteness as a contextual influence. Young children heard a model narrate a series of events in passive form while the model either performed the corresponding activities, showed pictures portraying the same activities, or displayed no referential aids. In accord with prediction, verbal modeling with enactive referents produced higher levels of comprehension of passives than modeling with pictorial referents or modeling without referential aids. Modeling with pictorial referents and modeling without referents did not differ in overall efficacy. However, modeling alone produced results that were less consistent across different measures of comprehension. Children who lacked understanding of passives were more dependent on concrete referents than those who had some initial comprehension of the linguistic form. The results suggest that verbal modeling with pictorial referents and verbal modeling alone facilitate comprehension of passives, whereas verbal modeling with enactive referents promotes learning. Findings of a supplemental experiment reveal that the effects of verbal modeling on comprehension are enhanced when syntactic forms occur in a meaningful verbal context.     

Eye movements of children, adults, and elderly persons during inspection of dot patterns

The effectiveness of three fading and two nonfading procedures were compared in training kindergarten children on an oddity problem in which shape was the relevant dimension. The fading procedures employed supplementary cues which were gradually eliminated during training. The cues included either saturating the odd stimulus with a red color, saturating the nonodd stimuli with a red color, or varying the degree of illumination of the nonodd stimuli. One nonfading group received training on a structurally simpler oddity problem prior to transfer to a more complex oddity problem, and a second nonfading group received the standard oddity training procedure. While significant differences in acquisition were not obtained between training groups, the oddity task performance and the results from a test of dimensional observation revealed that the fading procedures were differentially effective in transferring Ss′ observation from the dimension of the fading cue to that relevant for solution of the oddity problem.     

Metaphoric comprehension: Studies in reminding and resembling

The theoretical problems posed by metaphoric comprehension are discussed in the context of experiments on prompted recall. Listeners heard sentences of the form “Topic is (like) Vehicle.” In most cases, a statement of the implicit resemblance (the “ground”) was very effective in prompting recall of its related metaphor. This result could not be attributed to the activation, transfer, or additive combination of pre-existing properties of the topic and vehicle terms or to pre-existing associations between grounds and sentence terms. It is argued that the vehicle domain guides a novel schematization of the topic domain, that the perceived resemblance is a higher-order relation among entities (both explicit and implicit) in each domain, and that this abstract relation constitutes the “functional memory unit.” Prompted recall may begin with recognition of this previously experienced relation.     

Children's comprehension of complex sentences

Four tasks were given to children from 4–12 to test their comprehension of complex sentences containing main verbs taking underlying sentences as their complements (Sally knew that she was early). In an imperatives task, very young children interpreted only the complement verb and ignored the complex verb. In a short-term memory task, sentences with two negations usually lost the second not in recall. In direct questioning and anomaly-detection tasks, children tended to make pragmatic inferences and excessively depend on knowledge about the world, as opposed to linguistic information. Overall results showed that even sixth graders had not yet attained adult-level comprehension of complex sentences.     

Central capacity limits in consistent mapping,visual search tasks: Four channels or more?

The issue of limits in central processing capacity is of long-standing interest to psychologists. Some behaviors appear to draw on limited resources, others appear to have no resource limits. The existence of behaviors which do not compete for limited resources is suggested by the results from visual search experiments which keep constant the mapping of stimuli to responses. Behavior in such experiments is referred to as automatic detection. Results are reported which suggest, unlike the above, that resources are really quite limited in consistent mapping, visual search tasks. The results from this study (and other studies) are in agreement with a model of visual information processing behavior which assumes that the maximum number of cognitive comparison operations which can be executed simultaneously is restricted to somewhere in the neighborhood of four (the exact number depending on the subject). The model makes a strong prediction about the relation between the rate of stimulus presentation and accuracy, a prediction which is supported by the results. It is tentatively proposed that the steady-state, limited-channel model developed here may provide the beginnings of an integrated treatment of certain key results from subitizing and iconic memory experiments as well as visual search experiments.