La riduzione della motilità oculare in seguito a quesiti di natura visuo-spaziale: valutazione delle ipotesi dell'arousal e dell'interferenza visiva.

INTRODUCTION. In studies in which lateral eye movements were used as indicators of left versus right hemispheric activation during cognitive processing several authors (e.g. Ehrlichman, Weiner & Baker, 1974; De Gennaro & Violani, 1988) reported that there was a greater proportion of "stares" (absence of eye movements) following spatial questions than following verbal questions. This effect may be explained by two hypothesis: 1) the interference hypothesis; 2) the arousal hypothesis. The interference hypothesis (Singer, Greenberg & Antrobus, 1971) assumes that, since solving spatial questions involves visual imagery, visual stimuli in the testing environment may compete and interfere with internally generated visual images. In order to avoid overloading of visual processors, eye movements causing processing of new environmental visual input are suppressed or reduced. In fact experimental studies do not confirm this hypothesis: Meskin & Singer (1974), Ehrlichman (1980) and Ehrlichman & Barrett (1983) found no increase of the ocular motility in different conditions of environmental "enrichment-impoverishment" (when the experiment was run in a lit environment as compared to a dark one, or when the subject was facing an interviewer as compared when facing a blank screen). The arousal hypothesis posits that, since increase in oculomotor activity is associated with cognitive arousal (e.g. Lorens & Darrow, 1962; Amadeo & Shagass, 1963; Klinger, Gregoire & Barta, 1973), the increase in the number of stares following spatial questions reflects the fact that these are less difficult and arousing. Goals of the present study are to replicate the finding of a greater proportion of stares following spatial questions, within a study on the direction of reflective eye movements (De Gennaro, Devoto & Violani, in this Volume), and to assess the arousal hypothesis by evaluating whether longer latencies of reflection for verbal questions as compared to spatial questions confirmed that the former were indeed more difficult than the latter. METHODS. 64 female subjects were tested in a "lateral eye movements" experiment. For the procedure, see De Gennaro et al. (in this Volume); relevant for the present study is the definition of questions' difficulty which was posited a priori and validated by asking a group of 20 subjects to sort sets of verbal and spatial questions, homogeneous with respect to the implied cognitive processes, in each of two levels of difficulty (De Gennaro & Violani, 1988). From the video-recordings of each individual session a scorer, "blind" with respect to the content of each question, counted stares, defined as instances in which there was no detectable eye movement in the lapse between the end of the question and the beginning of the verbal answer. Latencies of reflection were defined by the time elapsed between end of each question and beginning of the ensuing answer. In order to ensure blindness of the scoring, the audio output of the video-recorder was off, and end of each question and beginning of answer were detected by the turning on and off of a LED. RESULTS AND CONCLUSIONS. Five-way mixed ANOVA (Distance x Fixation instruction x Sequence x Difficulty x Question type) on the number of stares showed the following main effects: "Fixation Instruction" (F=25.35; p<.00001) indicating more stares when subjects were requested to fixate a central point during the presentation of the questions (X=2.70) than when they were not (X=.95), "Difficulty" (F=25.38; p<.00001) indicating more stares for "easy" (X=2.05) than for "difficult" questions (X=1.61) and "Question Type" (F=76.70; p<.00001) indicating more stares for spatial (X=2.22) than for verbal questions (X=1.44). No other main effect nor interaction was significant. The same Anova on mean latencies of reflection showed a significant main effect for "Difficulty" (F=159.84; p<.00001), indicating shorter latencies for easy (X=8.88 sec.) than for difficult questions (X=14.88 sec.), and for "Question Type" (F=30.05; p<.00001), indicating shorter latencies for spatial (X=10.48 sec.) than for verbal questions (X=13.28 sec.). Furthermore a significant Difficulty x Question Type interaction (F=27.91; p<.00001) shows that, only for difficult questions, latencies of reflection were shorter for spatial (X=13.73 sec.) than for verbal questions (X=17.33 sec.; Duncan post-hoc test, p<.01), while for easy questions this difference was not significant (Spatial questions, X= 8.52 sec.; Verbal questions, X=9.23). The effect of more stares for spatial questions was replicated once more. The results on the mean latencies of reflection partially support the arousal hypothesis. In fact only for the questions defined as difficult the verbal ones were indeed found to have longer latencies of reflection than spatial ones. The lack of a significant interaction between difficulty and type of question on the proportion of stares leaves the problem of the causes of the greater proportion of stares following the "easy" spatial questions unanswered.