- Email: [email protected]
The role of the right hemisphere in processing negative sentences in context
Neuropsychologio. Printed in Great
Vol. 24, No. 2, pp. 289-292. Brikun.
@E&3932/86 $3.00+0 00 c, 1986 Pergamon Press Ltd.
1986
NOTE THE ROLE OF THE RIGHT HEMISPHERE IN PROCESSING SENTENCES IN CONTEXT
NEGATIVE
VALENT~NAD’URSO,* GANFRANCO DENES,~$ STEFANOTEsrAt and CARLO SEMENZA* *Istituto
di Psicologia
and tClinica
Neurologica
(Accepted 2 September
dell’Universit8
di Padova
1985)
Abstract--Given a pattern consisting of x 1, x 2, , x,,similar elements and ~,, which is perceived as different from the former, it is more plausible to assert that y is not x rather than to assert that x is not y (WASON. J. verb. Learn verb. Behav. 4,7-l 1, 1965). In order to appreciate such a difference, the entire set has to be considered. Right-hemisphere brain-damaged patients were submitted to a series of visually presented patterns, each pattern consisting of seven similar items and one dissimilar item. Their task was to complete a statement referring to a single element ofeach pattern. Statements were either simple affirmative or negative sentences. Errors and reaction times were recorded. Patients with a right-hemisphere injury were found to be insensitive to the plausibleeimplausible dimension in completing negative sentences. It is hypothesized that right-hemisphere brain-damaged patients are less adequate in this task because they are less capable in putting each element into the visual context.
INTRODUCTION SOMELINGUISTICtasks and computations may require as a prerequisite a normal functioning of the right hemisphere. Thus CARAMAZZA et al. [Z] reported that right-hemisphere brain-damaged patients (RHBDP) showed a selective deficit in solving two term series problems in which comparative adjectives in the premise and question are antonymous. In other words, when presented with problems such as “A is taller than B, who is taller?” (congruent problems), RHBDP performed as well as the control group, while they showed a selective impairment in processing sentences such as “A is taller than B, who is shorter?” (incongruent problems). Caramazza et al. proposed that this deficit in verbal reasoning was due to an impairment of a non-linguistic, imaginal process subserved by the right hemisphere. Actually they argue that, in the case of incongruent problems, subjects must search the spatial representation in order to respond correctly. However this issue has not been fully investigated. It is our opinion that similar shortcomings in RHBDP performance may occur not only with comparatives but also in processing other syntactic constructions, for instance some negative sentences. Full appreciation of the plausibility of a negative statement may. in fact, depend upon contextual clues which may be of a non-verbal nature. In normal subjects, responses to negative statements are generally longer than the affirmative ones, even when the amount of information conveyed by the statements is equal [7]. But, in 1965, WASON [8] found that reaction times (RT) to what are called plausible negatives are sigmlicantly shorter than that of implausible ones. In Wason’s experiment Ss were shown display cards with eight circles, seven of one colour and one of another colour. After seeing and describing the whole scene Ss were required to fill in sentences like “Circlr No. 3 is. .” or”Circlr n;o. 3 is nof .“. Affirmative sentences elicited faster RT. Among negative sentences those whose referent was the exceptional colour circle (plausible negatioes) produced significantly faster RT than negatives that stated that one of the seven similar circles was not of the same colour as the exceptional one (implausible negatives). WASON [S] interpreted his results on the basis ofan exceptionality hypothesis: given a set ofsimilar stimuli Y ,, Y2. and a stimulus J which is perceived as different from those, it is more plausible to assert that J’ is not Y rather than to assert that x is not y, In summary, normal subjects show different RT in completing negative sentences according to the relational status of the single element to be described. tTo whom reprint requests should Giustiniani 5, 35100. Padova, Italy.
be addressed
at: Clinica
289
Neurologica
dell’Universiti
di Padova,
Via
290
NOTE
Negative sentences are pragmatically justified by a context [h]. For instance, if we say “Today we did not play tennis”, this statement makes sense only if we play tennis quite often, and this inference is justified in our interlocutors. Wason’s experiment clearly shows a case of a single context that affects the adequacy of negative sentences, In fact. in order to distinguish between exceptional and normal elements the whole configuration has to be taken into account. It is therefore hypothesized, according also to the findings of Caramazza rr ul. that RHBDP would not show the plausibility effect in a case of a context constituted of a visual pattern.
METHODS The experimental condition, derived largely from WASO~U [SJ, consisted offour types of statements, each matched with a stimulus composed by seven similar items (black and yellow squares) and one dissimilar item (black square in the case that similar items were yellow or vice versa): one affirmative statement about the colour of the dissimilar item (DA); one negative statement about the colour of the dissimilar item (DN); one affirmative statement about the colour of a similar item (SA); one negative statement about the colour of a similar item (SN). Items were presented in two rows offour squares each, numbered consecutively from 1 to 8. Each of the four types of statement, referring to a specific item in the stimulus. was presented with the predicate omitted. for example “rhe square No. 7 is .“; “the squire No. 3 is nor .” Stimuli and statements were presented simultaneously on a TV screen controlled by an Apple II computer. Sixty-four statements were presented, counterbalanced for the following conditions: (1) affirmative vs negative; (2) referring to the dissimilar vs similar item; (3) seven black, one yellow square vs seven yellow, one black square; (4) correct answer key on the right vs correct answer key on the left. Order of presentation of the stimuli was randomized. According to our hypothesis (and to Wason’s results). the difference between the time taken to respond to negative and affirmative statements about the colour of the dissimilar items would be less than the corresponding difference for statements about the similar items, i.e. (DN DA)c(SN-SA), for normal subjects. In the RHBDP’s group such differences would be of equal sire.
Twelve RHBDP (seven males and five females), admitted to the Clinica Neurologica dell’Universita di Padova were tested. Their mean age was 62 yr. Educational level was 5 yr. The etiology of the lesion was neoplastic in 10 cases and vascular in two cases. Laterality of lesion was assessed by clinical examination, supplemented by CT scan: five of the patients (four males) had an anterior lesion (frontal or fronto-parietal), while seven (three males) had a posterior lesion, The patients were tested when the lesion was stable and all of them were free from any language comprehension disturbance as tested on clinical basis. As controls we tested 12 patients, admitted to our Department for pathology not involving the central nervous system (NBDP) and 18 left-hemisphere brain-damaged patients (LHBDP). mostly aphasics. NBDP and RHBDP were matched for age, sex and educational level.
The subjects were told that their task consisted in pressing with their right index finger. as quickly as possible, one of the two adjacent keys. showing a black or yellow square in order to denote the subject of the incomplete statement that appeared together with the stimulus. The response keys were selected from the right part of the key board, covering the remaining keys. The experimental task was preceeded by a training session conststing of repeated presentation of sets of stimult and incomplete sentences which ran until the S was performmg at a level of 75% correct. Once completed the training, the experimental session started, consisting of 64 sets stimulussstatement, each preceded by an acoustic warning signal, presented in blocks of 16. Errors and choice RT were recorded.
RESULTS All RHBDP and NBDP but only eight out of 18LHBDP were able to meet the criterion in the training session. A group of LHBDP was then left, hardly to be considered as representative of the left-hemisphere patients (five of them being aphasic-mild to moderate-and three non-aphasic). In the count of errors for each condition of the test. this latter group displayed far more errors than the other groups (see Table 1). Furthermore, as far as RT are concerned, the data from the left-hemisphere group showed a much larger dispersion than from the other groups. Finally they were significantly younger. For all these reasons, and also for the fact that our prediction was concerned with RHBDP only, we treated the data from LHBDP in a separate analysis. RHBDP were first divided in two groups according to the anterior or posterior site of the lesion; since calculation of the means brought almost equal figures in each condition RHBDP were collapsed in a single group for further analyses.
291
NOTE Table 1. Mean percentage of errors of the three groups for the four kinds of statement
NBDP RHBDP LHBDP
DA
DN
SA
SN
6.25 5.73 8.59
8.85 16.66 23.44
3.64 5.21 11.12
14.58 19.79 24.22
NBDP and RHBDP did not differ as far as the total amount of errors was concerned. A two-way ANOVA with repeated measures showed in fact no reliable group et&t nor interaction. Only type of sentences was significant (F= 1.35; df= 1, 22; ns.): all groups made less errors in the affirmative sentences. In order to test the prediction (DN-DA) < (SN-SA), median RT on correct responses for each kind of statement (see Table 2) and subsequently the differences (DN-DA) and (SN-SA) were computed for each subject (see Table 3). The study of RT is justified here because there is no problem of speed/accuracy trade off--errors and RT being in the same direction. We also judged RT on correct responses as the best measurement for patients’ performance in regard to the particular task in which we were interested. In fact, errors in the case of brain-damaged patients may come from sources of different nature, for instance perseveration and other attentional disorders, that may prove very difficult to control. Differences in RT, (DN-DA) and (SN-SA) served as input for a two-way ANOVA, in which group (NBDP and RHBDP) was the between factor and type of sentence (dissimilar or similar) was the within factor. Only the interaction was significant (F= 6.3 1; df= 1, 22; Pi 0.025). Post hoc analysis via the Newman-Keuls tests revealed the basis for such interaction: for NBDP (DN-DA) was significantly smaller than (SN-SA), (P
NBDP RHBDP LHBDP
(mean of medians, in msec) for the four in parentheses are S.D.s
DA
DN
SA
SN
3350 (1097) 5260 (998) 3949 (794)
3726 (1031) 6234 (1142) 4659 (1484)
3195 (580) 5337 (978) 4021 (1098)
4160 (1155) 6163 (1092) 5080 (1504)
Table 3. Mean values of (DN -DA) and (SN-SA) for the three groups (msec). (DN-DA) and (SN-SA) are differences between medians obtained from each subject in the ‘dissimilar’ and in the ‘similar’ conditions (DNPDA) NBDP RHBDP LHBDP
316 974 710
(SN-SA) 965 826 1059
292
NOTE
CONCLUSIONS Results of this study show that RHBDP take, as normals, more time in processing negative sentences than affirmative ones. On the other hand, unlike normals, they are insensitive to the context in processing negative sentences. Several hypotheses could be forwarded on an examination of this latter result. An explanation hased on an elementary spatial deficit can be easily ruled out: the performance in detecting the exceptional item was well above chance and there was no position effect. No specific claim about the nature ofthe processing of the stimulus context can be forwarded, whether it might be simultaneous or rapidly sequential. Again. the lack of any position effect of the critical element rules out an interpretation in terms of kind of process. It has been demonstrated [I, 51 that RHBDP tend, as a group, to show longer RT than patients with lesions of comparable size to the contralateral hemisphere. Since our patients were no exception to this, showing abnormally long RT, it could be hypothesized that the lack ofdifference in processing the two types of negative sentences. could be due to a floor effect. In order to rule out this hypothesis we looked at the performance of LHBDP with comparable RT: m this latter group, however. there was a clear difference between (DN- DA) and (SN SA). A linguistically-based hypothesis can be easily ruled out since syntactically the two negative sentences are identical. Pragmatic aspects are, instead, central IO the interpretation of these results. since the nature of the context is the only parameter in which the two types of negative sentence differ among each other. Indeed RHBDP have been demonstrated IO be insensitive to several aspects of pragmatics by GAKDNEK er a[. 133. However their study dealt with the quite complex issue generally referred to as “knowledge of the world”. Such knowledge is not, by defimtion. clearly specified and encompasses innumerable components. On the contrary, a few well-specified stimuli constitute all knowledge the task of our experiment requires to show the plausibility effect in normals. the only difficulty in organizing and memorizing such knowledge may come by the fact that the whole visual configuration should he taken into account. This difficulty is proved here to be insuperable for RHBDP, with negative consequences on their ability to face a linguistic task. n~kno~vled~ements-This work was partly supported by a C.N.R. grant to G.D. We would like to thank Mr Sandro Bettella for his help in building the stimuli. A preliminary version of this paper was presented at the Seventh European Meeting of INS (Aachen, June 12 15, 1984).
REFERENCES 1. BENSON,D. F. and BARTON,M. I. Disturbances in constructional abilities. Cortr.u 6, 19~46, 1970. 2. CAKAMAZZA, A., GORDON. J., ZUKIF, E. B. and DE LUCA. D. Right-hemispheric damage and verbal problem solving behavior. Bruin Lung. 3, 41-46, 1976. 3. GARDNER, H.. BKOWNELL, H. H.. WAPNER, A. and MITHELOW. D. Missing the point: the role of the right hemisphere in the processing ofcomplex linguistic materials. In Cognirire Processing in the Right Hrmispherr. E. PIXECMAN (Editor). pp. 169 191. Academic Press, New York, 1983. 4. H~~AEx. H. and ALBERT, M. L. Human Neuroprychology. Wiley, New York, 1978. 5. HOWES, D. and BOLLER. F. Simple reaction times: evidence for focal impairment from Ic.~on\ cjf the right hemisphere. Brain 98, 317. 322, 1975. 6. WASON, P. C. The processing of positive and negative information. Q. JI exp. J’s)?rhol. 11, 92- 107, 1959. 7. WASON, P. C. Response to affirmative and negative binary statements. Er. J. Ps)~chol. 52, 133 142. 1961. 8. WASON, P. C. The contexts of plausible denial. J. trerh. Learn. cerh. Behac. 4, 7 I I. 1965.
Vol. 24, No. 2, pp. 289-292. Brikun.
@E&3932/86 $3.00+0 00 c, 1986 Pergamon Press Ltd.
1986
NOTE THE ROLE OF THE RIGHT HEMISPHERE IN PROCESSING SENTENCES IN CONTEXT
NEGATIVE
VALENT~NAD’URSO,* GANFRANCO DENES,~$ STEFANOTEsrAt and CARLO SEMENZA* *Istituto
di Psicologia
and tClinica
Neurologica
(Accepted 2 September
dell’Universit8
di Padova
1985)
Abstract--Given a pattern consisting of x 1, x 2, , x,,similar elements and ~,, which is perceived as different from the former, it is more plausible to assert that y is not x rather than to assert that x is not y (WASON. J. verb. Learn verb. Behav. 4,7-l 1, 1965). In order to appreciate such a difference, the entire set has to be considered. Right-hemisphere brain-damaged patients were submitted to a series of visually presented patterns, each pattern consisting of seven similar items and one dissimilar item. Their task was to complete a statement referring to a single element ofeach pattern. Statements were either simple affirmative or negative sentences. Errors and reaction times were recorded. Patients with a right-hemisphere injury were found to be insensitive to the plausibleeimplausible dimension in completing negative sentences. It is hypothesized that right-hemisphere brain-damaged patients are less adequate in this task because they are less capable in putting each element into the visual context.
INTRODUCTION SOMELINGUISTICtasks and computations may require as a prerequisite a normal functioning of the right hemisphere. Thus CARAMAZZA et al. [Z] reported that right-hemisphere brain-damaged patients (RHBDP) showed a selective deficit in solving two term series problems in which comparative adjectives in the premise and question are antonymous. In other words, when presented with problems such as “A is taller than B, who is taller?” (congruent problems), RHBDP performed as well as the control group, while they showed a selective impairment in processing sentences such as “A is taller than B, who is shorter?” (incongruent problems). Caramazza et al. proposed that this deficit in verbal reasoning was due to an impairment of a non-linguistic, imaginal process subserved by the right hemisphere. Actually they argue that, in the case of incongruent problems, subjects must search the spatial representation in order to respond correctly. However this issue has not been fully investigated. It is our opinion that similar shortcomings in RHBDP performance may occur not only with comparatives but also in processing other syntactic constructions, for instance some negative sentences. Full appreciation of the plausibility of a negative statement may. in fact, depend upon contextual clues which may be of a non-verbal nature. In normal subjects, responses to negative statements are generally longer than the affirmative ones, even when the amount of information conveyed by the statements is equal [7]. But, in 1965, WASON [8] found that reaction times (RT) to what are called plausible negatives are sigmlicantly shorter than that of implausible ones. In Wason’s experiment Ss were shown display cards with eight circles, seven of one colour and one of another colour. After seeing and describing the whole scene Ss were required to fill in sentences like “Circlr No. 3 is. .” or”Circlr n;o. 3 is nof .“. Affirmative sentences elicited faster RT. Among negative sentences those whose referent was the exceptional colour circle (plausible negatioes) produced significantly faster RT than negatives that stated that one of the seven similar circles was not of the same colour as the exceptional one (implausible negatives). WASON [S] interpreted his results on the basis ofan exceptionality hypothesis: given a set ofsimilar stimuli Y ,, Y2. and a stimulus J which is perceived as different from those, it is more plausible to assert that J’ is not Y rather than to assert that x is not y, In summary, normal subjects show different RT in completing negative sentences according to the relational status of the single element to be described. tTo whom reprint requests should Giustiniani 5, 35100. Padova, Italy.
be addressed
at: Clinica
289
Neurologica
dell’Universiti
di Padova,
Via
290
NOTE
Negative sentences are pragmatically justified by a context [h]. For instance, if we say “Today we did not play tennis”, this statement makes sense only if we play tennis quite often, and this inference is justified in our interlocutors. Wason’s experiment clearly shows a case of a single context that affects the adequacy of negative sentences, In fact. in order to distinguish between exceptional and normal elements the whole configuration has to be taken into account. It is therefore hypothesized, according also to the findings of Caramazza rr ul. that RHBDP would not show the plausibility effect in a case of a context constituted of a visual pattern.
METHODS The experimental condition, derived largely from WASO~U [SJ, consisted offour types of statements, each matched with a stimulus composed by seven similar items (black and yellow squares) and one dissimilar item (black square in the case that similar items were yellow or vice versa): one affirmative statement about the colour of the dissimilar item (DA); one negative statement about the colour of the dissimilar item (DN); one affirmative statement about the colour of a similar item (SA); one negative statement about the colour of a similar item (SN). Items were presented in two rows offour squares each, numbered consecutively from 1 to 8. Each of the four types of statement, referring to a specific item in the stimulus. was presented with the predicate omitted. for example “rhe square No. 7 is .“; “the squire No. 3 is nor .” Stimuli and statements were presented simultaneously on a TV screen controlled by an Apple II computer. Sixty-four statements were presented, counterbalanced for the following conditions: (1) affirmative vs negative; (2) referring to the dissimilar vs similar item; (3) seven black, one yellow square vs seven yellow, one black square; (4) correct answer key on the right vs correct answer key on the left. Order of presentation of the stimuli was randomized. According to our hypothesis (and to Wason’s results). the difference between the time taken to respond to negative and affirmative statements about the colour of the dissimilar items would be less than the corresponding difference for statements about the similar items, i.e. (DN DA)c(SN-SA), for normal subjects. In the RHBDP’s group such differences would be of equal sire.
Twelve RHBDP (seven males and five females), admitted to the Clinica Neurologica dell’Universita di Padova were tested. Their mean age was 62 yr. Educational level was 5 yr. The etiology of the lesion was neoplastic in 10 cases and vascular in two cases. Laterality of lesion was assessed by clinical examination, supplemented by CT scan: five of the patients (four males) had an anterior lesion (frontal or fronto-parietal), while seven (three males) had a posterior lesion, The patients were tested when the lesion was stable and all of them were free from any language comprehension disturbance as tested on clinical basis. As controls we tested 12 patients, admitted to our Department for pathology not involving the central nervous system (NBDP) and 18 left-hemisphere brain-damaged patients (LHBDP). mostly aphasics. NBDP and RHBDP were matched for age, sex and educational level.
The subjects were told that their task consisted in pressing with their right index finger. as quickly as possible, one of the two adjacent keys. showing a black or yellow square in order to denote the subject of the incomplete statement that appeared together with the stimulus. The response keys were selected from the right part of the key board, covering the remaining keys. The experimental task was preceeded by a training session conststing of repeated presentation of sets of stimult and incomplete sentences which ran until the S was performmg at a level of 75% correct. Once completed the training, the experimental session started, consisting of 64 sets stimulussstatement, each preceded by an acoustic warning signal, presented in blocks of 16. Errors and choice RT were recorded.
RESULTS All RHBDP and NBDP but only eight out of 18LHBDP were able to meet the criterion in the training session. A group of LHBDP was then left, hardly to be considered as representative of the left-hemisphere patients (five of them being aphasic-mild to moderate-and three non-aphasic). In the count of errors for each condition of the test. this latter group displayed far more errors than the other groups (see Table 1). Furthermore, as far as RT are concerned, the data from the left-hemisphere group showed a much larger dispersion than from the other groups. Finally they were significantly younger. For all these reasons, and also for the fact that our prediction was concerned with RHBDP only, we treated the data from LHBDP in a separate analysis. RHBDP were first divided in two groups according to the anterior or posterior site of the lesion; since calculation of the means brought almost equal figures in each condition RHBDP were collapsed in a single group for further analyses.
291
NOTE Table 1. Mean percentage of errors of the three groups for the four kinds of statement
NBDP RHBDP LHBDP
DA
DN
SA
SN
6.25 5.73 8.59
8.85 16.66 23.44
3.64 5.21 11.12
14.58 19.79 24.22
NBDP and RHBDP did not differ as far as the total amount of errors was concerned. A two-way ANOVA with repeated measures showed in fact no reliable group et&t nor interaction. Only type of sentences was significant (F= 1.35; df= 1, 22; ns.): all groups made less errors in the affirmative sentences. In order to test the prediction (DN-DA) < (SN-SA), median RT on correct responses for each kind of statement (see Table 2) and subsequently the differences (DN-DA) and (SN-SA) were computed for each subject (see Table 3). The study of RT is justified here because there is no problem of speed/accuracy trade off--errors and RT being in the same direction. We also judged RT on correct responses as the best measurement for patients’ performance in regard to the particular task in which we were interested. In fact, errors in the case of brain-damaged patients may come from sources of different nature, for instance perseveration and other attentional disorders, that may prove very difficult to control. Differences in RT, (DN-DA) and (SN-SA) served as input for a two-way ANOVA, in which group (NBDP and RHBDP) was the between factor and type of sentence (dissimilar or similar) was the within factor. Only the interaction was significant (F= 6.3 1; df= 1, 22; Pi 0.025). Post hoc analysis via the Newman-Keuls tests revealed the basis for such interaction: for NBDP (DN-DA) was significantly smaller than (SN-SA), (P
NBDP RHBDP LHBDP
(mean of medians, in msec) for the four in parentheses are S.D.s
DA
DN
SA
SN
3350 (1097) 5260 (998) 3949 (794)
3726 (1031) 6234 (1142) 4659 (1484)
3195 (580) 5337 (978) 4021 (1098)
4160 (1155) 6163 (1092) 5080 (1504)
Table 3. Mean values of (DN -DA) and (SN-SA) for the three groups (msec). (DN-DA) and (SN-SA) are differences between medians obtained from each subject in the ‘dissimilar’ and in the ‘similar’ conditions (DNPDA) NBDP RHBDP LHBDP
316 974 710
(SN-SA) 965 826 1059
292
NOTE
CONCLUSIONS Results of this study show that RHBDP take, as normals, more time in processing negative sentences than affirmative ones. On the other hand, unlike normals, they are insensitive to the context in processing negative sentences. Several hypotheses could be forwarded on an examination of this latter result. An explanation hased on an elementary spatial deficit can be easily ruled out: the performance in detecting the exceptional item was well above chance and there was no position effect. No specific claim about the nature ofthe processing of the stimulus context can be forwarded, whether it might be simultaneous or rapidly sequential. Again. the lack of any position effect of the critical element rules out an interpretation in terms of kind of process. It has been demonstrated [I, 51 that RHBDP tend, as a group, to show longer RT than patients with lesions of comparable size to the contralateral hemisphere. Since our patients were no exception to this, showing abnormally long RT, it could be hypothesized that the lack ofdifference in processing the two types of negative sentences. could be due to a floor effect. In order to rule out this hypothesis we looked at the performance of LHBDP with comparable RT: m this latter group, however. there was a clear difference between (DN- DA) and (SN SA). A linguistically-based hypothesis can be easily ruled out since syntactically the two negative sentences are identical. Pragmatic aspects are, instead, central IO the interpretation of these results. since the nature of the context is the only parameter in which the two types of negative sentence differ among each other. Indeed RHBDP have been demonstrated IO be insensitive to several aspects of pragmatics by GAKDNEK er a[. 133. However their study dealt with the quite complex issue generally referred to as “knowledge of the world”. Such knowledge is not, by defimtion. clearly specified and encompasses innumerable components. On the contrary, a few well-specified stimuli constitute all knowledge the task of our experiment requires to show the plausibility effect in normals. the only difficulty in organizing and memorizing such knowledge may come by the fact that the whole visual configuration should he taken into account. This difficulty is proved here to be insuperable for RHBDP, with negative consequences on their ability to face a linguistic task. n~kno~vled~ements-This work was partly supported by a C.N.R. grant to G.D. We would like to thank Mr Sandro Bettella for his help in building the stimuli. A preliminary version of this paper was presented at the Seventh European Meeting of INS (Aachen, June 12 15, 1984).
REFERENCES 1. BENSON,D. F. and BARTON,M. I. Disturbances in constructional abilities. Cortr.u 6, 19~46, 1970. 2. CAKAMAZZA, A., GORDON. J., ZUKIF, E. B. and DE LUCA. D. Right-hemispheric damage and verbal problem solving behavior. Bruin Lung. 3, 41-46, 1976. 3. GARDNER, H.. BKOWNELL, H. H.. WAPNER, A. and MITHELOW. D. Missing the point: the role of the right hemisphere in the processing ofcomplex linguistic materials. In Cognirire Processing in the Right Hrmispherr. E. PIXECMAN (Editor). pp. 169 191. Academic Press, New York, 1983. 4. H~~AEx. H. and ALBERT, M. L. Human Neuroprychology. Wiley, New York, 1978. 5. HOWES, D. and BOLLER. F. Simple reaction times: evidence for focal impairment from Ic.~on\ cjf the right hemisphere. Brain 98, 317. 322, 1975. 6. WASON, P. C. The processing of positive and negative information. Q. JI exp. J’s)?rhol. 11, 92- 107, 1959. 7. WASON, P. C. Response to affirmative and negative binary statements. Er. J. Ps)~chol. 52, 133 142. 1961. 8. WASON, P. C. The contexts of plausible denial. J. trerh. Learn. cerh. Behac. 4, 7 I I. 1965.