Neuropsychologia.

1968,

Vol. 6, pp. 105 10I13. Pergamon

Press. Printed in England

FORM

PERCEPTION

ACROSS SENSORY
GEORGE KRAUTHAMER

MODALITIES*

Parkinson Research and Information Center, College of Physicians and Surgeons, Columbia University, 630 West 168th Street, New York 10032 (Received 16 November
1967)

Abstract-Forty-eight adult students were tested for their ability to recognize the tactile equivalents of visual patterns and the visual equivalents of tactile patterns. A modification of the paired comparisons method was used. The patterns consisted of nonsense forms presented either as stationary or traced contours. Prior to the crossmodal task, all Ss were trained to criterion on a purely visual and a purely tactile version of the task. Results on the crossmodal task indicated the ability of all groups of Ss to recognize patterns across sensory modalities. Crossmodal pattern perception was never better than intramodal perception; the level of performance was largely dependent upon the method of stimulus presentation (stationary or traced contours). No conclusive evidence for asymmetry of crossmodal perception was obtained.

to speak of visual and tactile form perception as entirely independent phenomena. This, however, is not necessarily the case; tactile and visual form perception may be associated with effects not restricted to the stimulated modality. Information conveyed by the somatosensory system would then become available, in part, at least, to the visual system and, likewise, visually given information could be utilized by the somatosensory system. Such phenomena would be crossmodal or intermodal as distinct from modality-specific or intramodal events. Attempts to explore crossmodal transfer in monkeys have yielded conflicting results. Both negative [2, 31 and positive [4-61 findings have been obtained. Other reports implicate the nature of the task and suggest that crossmodal transfer may or may not occur, depending on what is supposed to be transferred [7-l I]. In part, the problem is similar for man. No crossmodal transfer of a visual-auditory rhythm discrimination was found by COLE etal.[ 121 but various tests for the crossmodal transfer of spatial forms have yielded positive results [13-H]. More than thirty years ago R&&z [16] found that common illusions, as caused by Miiller-Lyer, Z611ner, and similar figures, could be experienced tactually if the subjects were presented with equivalent tactile stimulus figures. More recently, SEMMES et al.[ 171 observed positive transfer from vision to touch, when brain-injured subjects performed complex tactile tasks, which were logically analogous to previously given complex visual tasks. Positive transfer from vision to touch and touch to vision, was also found by GAYDOS [18] on a task combining verbal association with pattern recognition. Neither experiment,

IT IS customary

* Except for some minor changes in the Introduction, in order to take intc account the more recent experimental results, the article is published as written in 1959. It is based on a thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at New York University. An abstract has been published [I]. 105

106

GEOROEKRALFTHAMER

however, represents a direct test for the crossmodal transfer of form or pattern perception. Both designs were of the classical transfer of training type i.e. subjects were first trained in one modality and then in the other, so that relearning scores provided the only evidence for crossmodal transfer. Furthermore, both studies relied greatly on explicit verbal mediation. Indeed, the correct association of test patterns and words was the evidence for transfer in the experiment of GAYDOS. Since language is supra-modal and not linked exclusively to any one modality, it constitutes an added dimension which is best reduced to a minimum in crossmodal transfer tests. The present experiment explores transfer of pattern perception between vision and touch in a pattern discrimination and recognition test. Nonsense patterns are used to reduce the mediating role of language and the test design makes crossmodal performance above chance levels extremely difficult, if not impossible, in the absence of transfer mechanisms. To determine the effect of methods of pattern exposure on crossmodal transfer, the experiment is carried out under two different conditions of stimulation. A subsidiary question concerns the possible asymmetry of transfer across modalities (better in one direction than the other). METHOD
Subjects. Forty-eight subjects participated in the experiment. All were college students between the ages of 18 and 26 yr. None had a known history or apparent symptoms of nervous system pathology. Procedure. The test consisted of 32 nonsense patterns, approximately 3 x 3 cm in size and arranged in eight sets of four similar patterns (Fig I). Subjects sat facing a wooden screen which contained a window at eye-level for visual pattern exposure. Directly below it was an opening for insertion of the hand during tactile pattern presentation. Visual presentation was always binocular. Tactile presentation was alternately to the right and left palm held in a stationary position. The orientation of the patterns was such that the upper part of the visual patterns corresponded, tactually, to the part distal on the palm (nearest fingertips). Right-left position remained the same in both modalities. Crossmodal ros.k. On this task Ss Here required to identify the tactile equivalent of a visual standard pattern or, conversely, the visual equivalent of a tactile pattern. A nonsense pattern was first shown to Sin one modality e.g. vision, and designated by E as the Standard Pattern. Following this initial exposure the same pattern was again presented, but now tactually, as one of a set of four similar patterns. The order of exposure of these four Comparison Patterns was randomized. S was required to indicate whether the given Comparison Pattern was identical or different from the Standard Pattern just shown. The heteromodal Standard Pattern was shown to S before each comparison, and the complete presentation of one set, consisting of four crossmodal comparisons, was designated as one trial. Criterion for success consisted of two errorless consecutive trials (eight paired comparisons) in which Shad to correctly identify the Standard Pattern as identical to one of the four Comparison Patterns and as different from the remaining three Comparison Patterns. Ss were not informed of the results of their performance. Modality-specific tusk. The method of pattern presentation was the same as on the crossmodal task but the entire task was given in one modality. Initial presentation of the Standard Pattern was in the same sense modality as the subsequent exposure of the Comparison Pattern (i.e. all visual or all tactile). Order of testing. All Ss were first trained to criterion on the modality-specific forms of the test. Twentyfour Ss received tactile training before visual training, for the other 24 Ss the order of training was reversed. The order of testing for the .Ss was as follovvs: Modality-specific tasks 1st 2nd Visual Visual Tactile Tactile tactile tactile visual visual Crossmodal task 3rd Tactile-Visual Visual-Tactile Visual-Tactile Tactile-Visual N

Task sequence Group Group Group Group I 2 3 4

I2 I2 12 I2 48

FORM PERCEPTION

ACROSS SENSORY MODALITIES

107

This modality-specific training fulfilled a double purpose: Poor performance on a crossmodal task may reflect a difficulty in visual or tactile form perception itself, rather than a specifically crossmodal difficulty [19]. Modality-specific performance provided the baseline necessary for a comparison Modaiity-specific performance also served as a test of crossmodal transfer of the learn-relearn type employed in earlier investtgations [18, 171. Initial tactile training may have a practice effect on subsequent visual performance, and vice versa, prior visual training may affect modality-specific tactile performance. Following the two modality-specific tasks, .Ss were presented with the crossmodal task which was designated as Visual-Tactile if the Standard Pattern was visual and the Comparison Pattern tactile, and
STANDARD A. P B.

ti17

C.

D.

Lh

E.

I3

F. ccl G.

H.

FIG. 1. The eight sets of stimulus patterns. Tactile-Visual if the Standard Pattern was exposed tactually and the Comparison Pattern visually. One-half of the 48 Ss were given the Tactile-Visual task and one-half the Visual-Tactile one, Stimulus conditions. Alternate Ss were assigned to the successive (traced) and simultaneous (stationary) methods of pattern presentation, in no instance were mixed methods employed.* (1) Successive stimulation. The patterns were prepared as plywood stencils which guided a plastic stylus when the patterns were traced on the ss palm. For visual presentation the stylus was replaced by a modified flashlight projecting a pinpoint of light on a translucent glass window. This was called the successive (traced) method of presentation since both visual and tactile patterns actually consisted of a point describing a path in space and time. (2) Simultaneous stimulation. Each pattern was exposed at once, in its totality, and this was called the simultaneous (stationary) method. In this case the patterns were exposed visually as drawn in black on small white cards. The patterns for tactile presentation resembled cookie-cutters. They consisted of thin copper strips partially embedded in blocks of wood which E gently pressed into ss palm. Treatment of results. Analyses of variance were performed for the statistical treatment of the results. When warranted by the outcome of the analyses of variance, the differences between individual means were further evaluated by means of t-tests.
l As used here, successive and simultaneous refer only to the manner of pattern exposure. They are not to be confused with the exposure of several patterns at once (concurrent) or one after the other (consecutive). On all tests only one pattern at a time was shown.

108

GEORGE KRAUTHAMER

RESULTS The results demonstrated that form perception across two sensory modalities can occur in either direction, but that the extent, or efficiency of crossmodal pattern perception depends on the conditions of stimulation and order of testing (Fig. 2).
120 I10 too 5 : c a 0 0 i* 2 F 90 80 70 6O 50 40 30 20 IO 0
VISUAL I MODALITY -SPECIFIC TACTILE TACTILE-VISUAL I CROSSMODAL PERFORMANCE VISUAL-TACTILE

0 0

Smwltaneaus Successwe

stimulation sl~mulat~an

Fig. 2.

The number of trials to criterion on the modality-specific and crossmodal tasks for two different methods of stimulation.

All Ss learned the preliminary modality-specific Visual and Tactile tasks to criterion (Table 1). As expected, visual form perception was significantly superior to tactile form perception irrespective of the method of stimulation (p<[Link]). Furthermore, visual pattern perception was significantly better with the method of simultaneous stimulation than the method of successive stimulation. In contrast, tactile pattern perception required the same number of trials with either method of stimulation. Since it is possible that learning to recognize the patterns in one modality may have had a beneficial effect on subsequent relearning in another modality (transfer of training across modalities) performance on the visual task before tactile training was compared with visual Likewise, tactile performance was analysed to determine performance after tactile training.
Table 1. Mean number of trials to criterion on modality-specific tasks Significance of difference -COW1 <O+Ol

Method of stimulation Simultaneous

Visual 18 28 <0001

Tactile 44 46 NS

Significance of difference

FORM PERCEFTION

ACROSS SENSORY [Link]

109

possible practice effects accruing from prior visual training under two different methods of stimulation. It could not be demonstrated that prior tactile training leads to better visual performance. Neither for simultaneous nor for successive methods of stimulation was the visual performance better when it followed tactile training. Likewise, prior visual training did not lead to better tactile performance when the patterns were traced (successive stimulation) but with the simultaneous method of stimulation (exposure of complete pattern) tactile pattern recognition was significantly better (~~0.01) if preceded by the visual form of the test (Table 2).
Table 2. Mean number of trials to criterion on a modality-specific training in another modality Tactile Successive -45 47 NS 51 37 < 0.01 27 30 NS 17 18 NS Simultaneous Successive task before and after

Task order _____ 2nd 1st -.-Tactile Visual Visual Tactile

-_

Visual Simultaneous

Significance of difference

The results of modality-specific tasks demonstrate that Ss are able to correctly perceive and discriminate the test patterns visually or tactually. Any performance decrement on the subsequent crossmodal task reflects, therefore, a loss of information due to the crossmodal nature of the event (Table 3). No such loss was in evidence when the method of successive stimulation was used. The identification of forms across two different modalities required a mean of 48 trials and thus was no more difficult than the purely tactile task which was the more difficult of the two modality-specific tasks, and required a mean of 46 trials. With simultaneous stimulation, however, performance on the crossmodal task was markedly inferior to either the purely visual or purely tactile performance. Transfer in a visual-tactile direction now required an average of 111 trials which is twice the number of trials it took the same twelve Ss to learn to criterion the modality-specific tactile task (Mean=41). The situation is more complex on the tactile-visual task due to a significant interaction with the prior modality-specific task order. Of the twelve Ss those (N = 6), who first trained on a tactile and then on a visual task during their preliminary modality-specific training, did as poorly as those who had been given the crossmodal visual-tactile task. On the other hand, the six Ss who were given the preliminary modality-specific tasks in the reverse order, i.e. visual before tactile, did no worse on the crossmodal tactile-visual task (M=48) than on the modality-specific tactile one (M = 40). Their performance thus resembles that of the successive stimulation group. An analysis of covariance ruled out the possibility that their superior performance on the tactile-visual task was due to an initial superiority in tactile pattern perception. There was no evidence for a general asymmetry of transfer. For the successive stimulation groups visual-tactile transfer was equal to tactile-visual transfer. Though transfer in the tactile-visual direction tended to be superior to visual-tactile transfer with simultaneous

110

GWRGI KRAUTHAMER

Table 3. Mean number of trials to criterion for all groups on crossmodal tasks Modality-specific task order
1st Visual 2nd Tactile Tactile-visual Visual-tactile Tactile Visual Visual-tactile Tactile-visual Crossmodal task Method of stimulation Successive ..-__ 49 51 52 41 Simultaneous Signif. of diff. NS <[Link] c 0.001 <[Link]

--___-____48 113 109 95

stimulation, the difference is significant for only one of the two groups. Similarly, transfer of training on the preliminary modality-specific task provided no evidence for asymmetrical transfer across modalities. An attempt was also made to rank the eight sets of patterns according to order of difficulty. This varied from task to task and S to S. When all Ss and all tasks were pooled, patterns A and F ranked as the easiest, and patterns D, G, and H as the most difficult
for both methods of stimulation.

DISCUSSION The intersensory effects of simple stimuli have been the subject of many psychophysical experiments [20-221. These showed that alterations of sensory thresholds and stimulus intensity in one modality occurred as a function of concurrent stimulation in another modality. The present experiment investigated sensory interaction not at a threshold level but in terms of form or pattern perception. This particular aspect of intersensory relationships between touch and sight is an old issue. Though it was already raised by DIDEROT in 1773 in his discussion of the Molineux problem [23, 241, it is now generally associated with VON SENDENS study of visual perception following surgical extraction of congenital cataracts [25]. VON SENDENS conclusion, seemingly supported by animal studies [26, 271, is that transfer of perception from touch to vision does not occur. A somewhat different conclusion, however, emerged in a more recent study of a case of early blindness due to cornea1 opacity [28]. When given his sight in late middle age,
this patient was able to recognize block letters by sight which he had previously to recognize by palpation during his long period of blindness. learned

It should be pointed out that VONSENDEN based his conclusion upon a bibliographic analysis only. An experimental study of early blind, late blind, and normal blindfolded

children on a tactile task, led DREVER[29] to conclude that heteromodal learning is of importance. A conclusion different from VON SENDENS may also be drawn from Ivo KOHLERS
work [30]. When normal visual-tactile and other intersensory relationships are experimentally disarranged through optical techniques, complete perceptual and behavioral readaptation takes place. These results support the existence of crossmodal transfer mechanismsat least in man. In reptiles and amphibians visual-tactile relationships can also be disarranged by surgical rather than optical means. However, when this is done, the result is a total lack of readaptation [3 I].

FORM PERCEPTION ACROSS SENSORY MODALlTlES

111

The basis for crossmodal transfer remains unknown. RJ%SZ [161postulated a common substrate for all perception (hence supposedly identical visual and tactile illusions). Such a concept is not far removed from older views about sensory unity which were, occasionally, tinged with a touch of mysticism [32-341. Equally speculative, though more physiologically oriented, are some ideas expressed by LASHLEY[35, 361. In his discussion of sequentially organized behavior patterns he postulated a cerebral space-coordinate system which might permit the required functional interchangeability of spatial position and temporal order across sensory modalities. In the present experiment the method of simultaneous stimulation can be viewed as stimulation with patterns having only spatial attributes, whereas the method of successive stimulation provides the further attribute of temporal sequence. These two methods led to different results. Spatio-temporal pattern exposure (successive stimulation) led to uniformly good crossmodal transfer whether it be from vision to touch or touch to vision. On the other hand, exposure of patterns with spatial properties only (simultaneous stimulation) produced considerable interference with crossmodal transfer. The one simultaneous stimulation group which showed no decrement on the crossmodal task had been given tactile training immediately preceding the tactile-visual transfer task. Consequently the more difficult tactile patterns were repeatedly presented without intervening exclusively visual pattern exposures. This may have provided the kind of perceptual training (overlearning?) necessary to overcome the adverse effects of this particular method of stimulation. The possibility must also be considered that crossmodal transfer may be highly effective if visual and tactile perception are of the same order of difficulty, as was the case with the successive method of stimulation, but that highly disparate levels of tactile and visual pattern perception tend to interfere with crossmodal transfer. Generation of nonsense patterns based on principles of information theory may represent one way of developing visual and tactile patterns of equivalent difficulty, regardless of the method of stimulation [37,38]. In the present experiment an interesting difference was observed between visual and tactile pattern perception seemingly related to stimulus conditions. As expected, when the visual pqtterns consisted of a pinpoint of light describing a path in space (successive stimulation) Ss required considerably more trials to learn the task than when the same patterns were presented as stationary line drawings (simultaneous stimulation). Quite unexpected, however, were the results of the tactile task. It made no difference whether the patterns were traced out on the palm or whether they were pressed into the palm in the form of cookiecutters; the number of trials to criterion was the same for both methods of stimulation. This is contrary to the widely-held notion that movement is essential to perception in the tactile modality [39]. The results are supported, however, by %vBszs [16] observation that haptic illusions are equally present under conditions of movement and rest, and by CRITCHLEYS claim [40] that blind Ss are able to read Braille under a great variety of conditions. When transfer across modalities was looked for in the classical transfer design the results were negative. Training first in one modality and then in another does not seem to create the condition best suited for crossmodal transfer, regardless of the method of stimulation. If sensory segregation were complete, visual perception could of course derive no benefit from tactile perception and vice versa. On the other hand, the absence of sensory

112

GEORGEKRAUTHAMER

segregation does not by itself insure the effective crossmodal transfer of perceptual information-it merely makes it possible under the appropriate stimulus conditions. SUMMARY AND CONCLUSION

Pattern perception across sensory modalities was explored in 48 normal Ss by means of a pattern identification and discrimination test. The design of the task made successful performance contingent upon direct crossmodal comparisons of tactile and visual nonsense patterns. The test patterns were either traced or presented at once in their entity. While form perception across modalities can, presumably, not be better than either purely visual or purely tactile pattern perception it need, likewise, not be worse. This suggests that there is no necessary loss of information in transmitting a neural message about a pattern from one sense modality to another. Under appropriate conditions such information is equally available both to touch and vision. REFERENCES I. KRAUTHAMER, G. Form perception across sensory modalities. Amer. Psycho/., 14, 396, 1959. 2. ETTLINGER, G. Cross-modal transfer of training in monkeys. Behaviour 16, X-65, 1960. 3. BURTON,D. and ETTLINGER, G. Cross-modal transfer of training in monkeys. Nature 186, 1071-1072, 1960. 4. STEPIEN,L. S. and CORDEAU,J. P. Memory in monkeys for compound stimuli. Am. J. Psycho/. 73, 388-395, 1960. 5. WILSON, M. and WILSON, W. A., JR. Intersensory facilitation of learning sets in normal and brain operated monkeys. J. Comp. Physiol. Psychol. 55,931-934, 1962. 6. WUON, W. A., JR. and SHAFER, 0. C. Intermodality transfer of specific discriminations in the monkey. Nature 197, 107, 1963. 7. WEGENER, J. G. Cross-modal transfer in monkeys. J. Comp. Physiof. Psychot. 59,450-452, 1965. 8. WUON, W. A., JR. Intersensory transfer in normal and brain operated monkeys. Neuropsychologia 3, 363-370, 1965. 9. TE~BER,H. L. Postscript: Some needed revisions of the crossmodal views of agnosia. Neuropsychologia 3, 371-378, 1965. 10. E~LINGER, G. and BLAKEMORE, C. E. Cross-modal transfer of conditional discrimination training in monkeys. Narrtre 210, 117-I 18, 1966. 11. BLAKESLEE, P. and GUNTER, R. Cross-modal transfer of discrimination learning in Cebus monkeys. Behaviour 26,76-90, 1966. 12. COLE, M., CHOROVER, S. L. and ETTLINGER, G. Cross-modal transfer in man. Nature 191, 1225-1226, 1961. 13. BIRCH,H. G. and LEPFORD, A. Intersensory development in children. Monogr. Sot. Res. Child Develop. Zs, l-48, 1963. 14. RUDEL,R. G. and TEUBFR,H. L. Decrement of visual and haptic Muller-Lyer illusion on repeated trials: A study of crossmodal transfer. Quart. J. fip. Psycho/. 15, 125-131, 1963. 15. RUDEL, R. G. and TEUBER,H. L. Crossmodal transfer of shape discrimination by children. heuropsychologia 2, 1-8, 1964. 16. RBvhz, G. System der optischen und haptischen Raumtauschung. 2. Psychol. 131,296-375, 1934. 17. SEMMES, J., WEINSTEIN, S., GHENT, L. and TEUBER,H. L. Performance on complex tactual tasks after brain injury in man: Analysis by locus of lesion. Am.J. Psychol. 67, 220-240, 1954. 18. GAYDOS,H. F. Intersensory transfer in the discrimination of form. Am. J. Psycho/. 69, 107-l 10, 1956. 19. WEINSTEIN, S., SEMMES, J., GHENT, L. and TEUBER,H. L. Spatial orientation in man after cerebral injury: II. Analysis according to concomitant defects. J. Psychol. 42, 249-263, 1956. 20. RYAN, T. A. Interrelations of the sensory system in perception. Psychol. Bull. 37, 659698, 1940. 21. GILBERT,G. M. Inter-sensory facilitation and inhibition. J. gen. PsychoI. 24, 381-407, 1941. 22. LONDON,I. D. Research on sensory interaction in the Soviet Union. Psycho1 BuII 51, 531-568, 1954. 23. DIDEROT,D. Lettre sur les Aveugles. In Oeuvres Completes, Vol. 1, pp. 279-330. Gamier, Paris, 1875. 24. JAMES,W. Principles ofPsychology, Vol. 2. Dover, New York, 1950. 25. SENDEN,M. VON. Raum und GestaItauff?tssungbei operierten Blindgeborenen vor und nach der Operation. Barth, Leipzig, 1932. A. H. The development of visual perception in man and chimpanzee. Science 106,107-108, 1947. 26. RIESEN,

FORMPERCEPTION ACROSS SENSORY M0DALITlE.S

113

27. NISSEN,R. W., CHOW, K. L. and SW, J. Effects of restricted opportunity for tactual kinesthetic, and manipulative experience on the behavior of a chimpanzee. Am. J. Psychol. 64,485-507, 1951. 28. GREOORY, R. L. and WALLACE,J. G. Recovery from early blindness. A case study. Exp. Psychol. Monogr. 2, l-46, 1963. 29. DREVER,J. Early learning and the perception of space. Am. J. Psychol. 68, 605-614, 1955. 30. KOHLER, I. Ueber Aufbau und Wandlung der Wahmehmungswelt. Insbesondere ueber bedingte Empfindungen. Oesterr. Akad. Uissensch., 227/I, 32-84, 1951. 31. SPERRY, R. W. Mechanism of neural maturation. In Handhook of Experimental Psychology. S. S. STEVENS (Editor). Wiley, New York, 1951. 32. BECHER,E. Cehirn und Seele. Heidelberg, 1911. 33. WERNER,H. Comparative Psychology and Mental Development. Harper, New York, 1940. 34. HORNBOSTEL, E. M. EON. The unity of the senses. In A Source Book of Gestalt Psychology, W. P. ELLIS (Editor). Harcourt-Brace, New York, 1938. 35. LASHLEY, K. S. The problem of serial order in behavior. In Cerebral Mechanisms in Behavior, The Hixon Symposium, L. A. JEFFR~ES (Editor). John Wiley, New York, 1951. 36. LASHLEY, K. S. Functional interpretation of anatomic patterns. In Proc. Assoc. Res. Nerv. Ment. Dis. Vol. 30, P. BARD (Editor). Williams & Wilkin, Baltimore, 1952. 37. ANDERSON, N. S. and LEONARD, J. A. The recognition, naming, and reconstruction of visual figures as a function of contour redundancy. J. exp. Psychol. 56,262-270, 1958. 38. F~rrs, P. M., WEINSTEIN, M., RAFQPORT,M., ANDERSON, N. and LMNARD, J. A. Stimulus correlates of visual pattern recognition: A probability approach. J. exp. Psychol. 51, l-11, 1956. 39. KATZ, D. Der Aufbau der Tastwelt. Z. Psychol. Erg. Bd. II, 1925. 40. CRITCHLEY, M. Tactile thought, with special reference to the blind. Brain 76, 19-35, 1953. R&sun&-Le degre de transfert dinformation spatiale entre les modalitbs visuelles et tactiles on a tent& de determiner sur un groupe de 48 etudiants. Des configurations spatiales (nonsense patterns) ont Btt presentQs aux sujets; il sagissait de discerner ces couples de figures par perception tactile ou visuelle. Des configurations mobiles et stationnaires ont ete utilisees dans differentes conditions experimentales. Lexp&ience intermodale etait preceded par des experiences purement visuelles et purement tactiles pour entrainer les sujets jusqua un certain niveau de reconnaissance intramodale de ces configurations. Les resultats ont demontrd lexistence dun te1 transfer? intermodal; son niveau nest jamais superieur a la reconnaissance intramodale et il depend en outre largement du mode de presentation (configurations mobiles ou stationnaires). 11 nttait pas possible de demontrer avec certitude une asymmetric de transfert dam le sens visuel-tactile ou dans le sens inverse. Zusannnenfassung-Achtundvierzig Studenten wurden auf ihre Flhigkeit geprilft equivalente Gestalten wiederzuerkennen die entweder optisch oder taktil vorgelegt wurden. Eine Art der gepaarten Vergleichsmethode wurde angewandt. Die Gestalten bestanden aus stationaren oder gezogenen nonsense Kontouren. Vor dem intermodalen Versuch wurden die Versuchspersonen in einer rein optischen und einer rem taktilen Art des Versuches bis zu einem gewissen Erkennungskriterium der Gestalten trainiert. Die Ergebnisse des intermodalen Hauptversuches z&ten das es allen Versuchsgruppen miiglich war diese Art von Kontourformen aufgekreuztem Sinnesweg wiederzuerkennen. Das intermodale Erkennungsvermiigen war niemals besser als das rein opt&he oder taktile. Die Hbhe des intermodalen Erkennungsvermiigen war stark von der Reizmethode abhangig (stationare oder gezogene Gestalten). Eine miigliche Asymmetrie des mtermodalen Erkennungsvermiigen konnte nicht bestimmt nachgewiesen werden.