NOTE ABNORMAL
EYE MOVEMENT BEHAVIOUR DURING TEXT READING NEGLECT SYNDROME: A CASE STUDY HAM-OTTO KARUTH*
Department
of Neurology.
IN
and WALT~K HLHIX
Rheinisch-Westfilische
Technische
Hochschule
Aachcn.
Germany
(Rec~riwd19 June I99 I : c~c.cy~trd 2XFdmur~ 1992) Abstract-The eye movement hehavtour of a patient suffering from a right basal ganglia infarction with a left-sided hemineglect hut without any visual field defects was investigated during reading. The eye movements were registered by means ofan i.r. light technique (pupil corncal reflection method). The main findings were abnormal return sweeps. Whereas in normal readers the end ofonc line of text is linked to the beginning of the new line by a long leftward saccadc, the return sweeps of the hemineglect patient stereotyptcally ended in the middle of the next hne. They were followed by sequences of short saccades indicating silent backward reading until a linguistically plausible continuation of sentences from the previous line was found. irrespective of the actual beginning of text. The shortened return sweeps could not be attributed to a general oculomotor disturbance. The spatial border for the occurrence of the patient’s abnormal scanning pattern (left half of texts) clearly did not depend on a retinal coordinate frame of reference but rather has to he attributed to a different body-centred reference system.
INTRODUCTION FKOM C'LINK'AL descriptions of hemineglect, dcviatmns in reading are well known although there are only relatively omissions of few detailed experimental studies. Several authors (e.g. Refs [IO]. [ 121 and 1141) reported contralateral letters in single word reading, an issue whtch recently received new interest hecausc of its implications for cognitive models of reading [ 131. In contrast, whole text reading was rarely studied probably due to limitattons in registratton techniques. When normal subjects read a text, their eye movements exhibit the following phases scanning.At the beginning. readers show one long orienting saccade to the upper left corner a text succeded by a few short correcttvc saccadea in order to locate the gaze on the initial text constituent. Then. the first line ts scanned by a sequence ofsaccades and fixations. The left-to-right scanning may he interupted by regressive eye movements indicating dilhculties of visual and’or linguistic information processing. When the end of a lint is reached. normal readers show the follovving routine hehaviour: The last fixation is slightly extended and a return sweep is initiated, i.e. a long leftward saccade to the beginning ofthe next line. Usually a few corrective saccadcs are necessary before the scanntng ofthe new line can be started. It is obvious that left visual hemineglect after right brain damage would lead to difficulties especially tn two phases of the reading process, namely. searching for the beginning of the text and performing the return sweep from the end ofone hne to the beginning of the next one. In the present study. we report on the text reading bchaviour in ;t GLX of pure left visual hemincglect without vtsual field defects.
of
of
METHODS H.S.. a 56-year-old police officer. sustained a thromhotic occlusion at the bifurcation artery several hours after EC IC bypass surgery. A CT scan made 2 weeks po\t-onset *Address for correspondence: D-5100 Aachen. Germany.
Dr H. 0. Karnath.
Department
593
of Neurology.
of the right internal carotid revealed an extcndcd right
RWTH
Aachen.
Pauwelsstrassc,
594
NOTI
basiul ganglia lesion Icaving the primary \~\ual pathway intact. GOI ~xa\ perlmetry yelded no signs ofwaual lield dcfccts L7]. Nelthcr paresis ofgaze nor any other oculomotor abnormahties were observed. H.S. shoucd clear signs of a left-sided \~sual hemincglcct on the following tabkx: lint bisection. copying of a Ilwcr. completion of a clockfxc. dctcction of visualstimuli (clinical confrontation method) and naming oft;lchistoscopicall~ prrsented picturcr under bilateral stimulation (as opposed to unilateral stimulation). Ncglcct symptoms &crt‘ also ohserved in ocryday behabiour on the ward. A more detailed description of H.S. wzs gi\cn in Ref. 171. The eye mobcmcnt stud) w\ carried out 5 weeks pot-CVA.
Stimuli consisted ofn~nc written stories, which conceycd humorous everyday lift cpibodes. Each star) was mxlc up of six scntcnces. resulting in a total of Y x 6 = 54 sentences. Each scntencc started at the beginning of a new line and continued unto the succeeding one. Thus. each text extended over 12 lines. We >!stematicallq varied the cxtcnsion oftt’xt into each second lint per sentence. Some xntenccs ended far left of the midpoint, others in the right hnlfofthc line. Thih particular format oftcxts gave us the opportunitv to study the relationship between performIng roturn sweeps and finding the linguistic continuation of the individual sentences. Furthermore. wc baried the linguistic content at the beginning of each second lmc per sentences. In 30 instances a linguistically plausible continuation M#ould rc:hult even ifthc first constituent(~) on the xcond line waqwerel left out during wading. In the other 24 imianco. any omision of words hould hc linguistically unacccptahle. H.S. ~+a atkcd to read all nine text\. As a control group ue studied nine non-brain-damaged bubjcct, (ficc female, four malcj) ranging in age hctween 25 and 5X years. Each of the control subject\ &as presented with one ol the nine texts. Thu.4 we obtained 54 scntcnccs being read hq the control group as uell as by H.S.
The ICXIS were projected from slides fully colering the projection screen (75 cm high x IO0cm wide, 24 x 32 dc~l of visual angle). Each line contained X0 character spaces with ;I G/e of0.7 x 0.4 Eye movements sere reglstercd Kith the pup11 cornea1 rcflcction method 1161. The apparatw consisted of a microprocessor system (DEBIC 84). \+ hich \+a\ used with ;I conccntional cidro system (c)c- and scent-camera). The system allows 25 SO measurement\ per second. The mean Euclidian accuracy ib 0.7 with a standard dcliation (SD) of I .h [S]. The sub~ccts WI with their heads liked at a dlstancc of 175 cm from the projectwn wccn. Calibration took place immediately bcforc each registr-ntion. The subjca had to tinate 20 light cmitring diodes which here evenly distributed over tht- screen. Data reduction was achlcvcd b> a modified version of the program EMAN [I I]. Succesivc measurement point\ \\erc combined into “liuution” a long 3s they fell into a gliding window of0.5 The minimllm duration of fixation\ \\;I\ Xl ill 10 mscc.
RESULTS In Figs I and 2 ;Lcharacteristic scanpnth is shown for a normal subject and for the hemtneglect patlent H.S. Each of the two text lines is plotted twice. m order to illustrate the four phases of scnnn~np: acarching for the beginning. reading the first line. return weep. wading the second line.
As expected. the control suh,jects always reached the besinning ufscntcnccs in one long saccade foIlwed by a few corrective saccades (Fig. I ). H.S. had ditfculties which were rellected in both lrrcgular scanning and linguistic crrorb. He left out initial v,.ords when rending aloud in 31 out of 54 sentences (57 “:I). However. hi5 scanning demonstrated in the majority ofsentencez. (70”,,,) that hc. ncccrtheless. directed hi5 gare lefthard\. ix. towards the beginning of tat. During reading aloud. H.S. sho\\cd normal scanmng parameters for some IIIKS. but for other\ wbnormal lxha+iour a illustrated in l.‘ip. 2. (For ;I more dctailcd descrlptlon of these pha\ch of the reading prows.\ see Ref. [.I]. J
H S. shv~\sd ;I .rtnklng dilTcrcncc to ihc long saccadcq normal rcadcrs made Icftwnrds ac‘ro\s the H hole lint (cf. FIFE I and 2). Whcreab the mean (dlsgonal) length of the return aweps in the control group was 24.4 (SD =4.X I. IH.S. only performed an averaged return suecp of I I .5 (SD = 3.5 I. The return s!%eepsofthe controls Icd them close to the bcglnninp of the text in the second lone ofthc vzntcncca (distance to the hegInning: mean =4.X (SD=?.0 )I. \vherca H.S. landed with ;I distance of 17.1 (SD -3 7 ). On the a\cragc. the return s\he~‘pbof H.S. were halfas long ;I\ the ona of normal controls. and their landing positions fell into the middle of the next lint. The distance between the starting point of the return awecps of H.S. and the end oftext in line I could not explain the ~Iivrtcncd rclurn s\lccps (controls: mean 7 I .? (ST)-j.4 ) H.S.: 1.3 (SD1.6 )I. Another possible cuplatwtion for this beha\iour could bc that the return sweeps of H.S. hcrc aimed at the end-points of the written test in c~ch wcond lint (the lcngh of \\hlch \\as \~.,ternat~call> varied XI-OS\ the 54 wntencca). but the Spearman
NOTE
ByI
AT
ElNEn
OACKEL FIEL
595
EINEM
ALTEREN
HE4RN Pi_OTZLlCcl
vor*r
BE
\I
260
BLUMENTOPF AUF DEN KOPF.
Fig. I. Example for reading phases in a normal control subject (male. S9 qears): (I ) searching for the beginning of the sentence: (2~ reading aloud of the first line of the sentence: (3) return sweep and corrective saccades: (4) reading aloud of the second line of the sentence. (Fixations with a duration of 160 msec and more are plotted as numbers.)
rank-order correlation between the length ofthe written text and the length of the return sweeps was low (i/=0.33). In H.S.. we observed overshooting return sweeps when the text ended in the right half of the projection screen. i.e. landing of gaze within the text. In contrast, undershoots were found with texts ending in the left halfof the screen. In contrast to the previously reported correlation. the length ofover- ‘undershoots was highly correlated (ii = 0.77 I Mith the length of the written text in the line indicating that the return sweeps in their majority landed closer to the midpotnt of the line than to the end-points of the written text (f’eLerlty of hemineglect on the rending behakiour. W&I.KYK rt trl. (personal communication) found no return sweeps during text reading in their left-sided hemineglect patient B.Q. [ 151.The patient was unable to find a linguistically plauaiblc continuation oftext from one line to the other. She complained about it. but. nevertheless. did not show any systematic backward reading a5 found in our patient H.S. Different components of the neglect sqndromc L3. h] are prominent in the two patients. which ma) reflect different degrees of severity. B.Q. showed an attentional orientation bias to the ipsilateral side. Ia\ w H.S. The cognitive demands of text reading apparently have oberrldden an Ipsilateral orientation txas m H.S. Thus. his di?turhance of the internal egocentric frame for rcprcscnting spatial orlcntation hccamc more apparent.
.~~~,lo~~/t,dyr,tlr,lr.,~Thls study was supported by a grant from the Deutsche For\chungsgemelnschaft. Dr K. Willmes for helpful comments on the manuscript.
We thank
NOII
REFERENCES I. 2. 3. 4. 5. 6.
BK.zI>. W. R. Vtsual diaoricntat~on \*ith special rcfcrcncc to lesion& of the rlghr ccrcbral hemisphere. Brtrirl 64, 244 272. 1941. Dl RI NII. E.. GI.~III.IYI. M.. FA(;I.IO\I, P. and BAKHII:KI. C Attcntional shift tov,nrds the rightmost btunull 111 patients Wllh icrt visual neglect. C0rrr.x 25, 23 I 737, 1989. G~INOTT~. G.. D'FKMI, P. and DI. Bows. C. Aspects cliniqucs et mtcanlames de la n@ligence visuo-\patialc. Krr. ,Yrwo/. ( t’wi.5) 145, 626 634. 19X9. HUHI K. W.. Gr ILLOI. Ci. and KAKNATH, H. 0. Blickbcwe~ungsstrateFicn hcim L,esen mit linkascltiger Hemianopsie und Hcmincglect. %. L/in. /‘\yckol. 17, 244 259. 198X. H~I+I.K. W., Li.11K. G. and L.ASS.C. Eye movement hehavlor III aphasia. In \r,lrrop.\!,c,/~ol(~~/~ of f$, :Mor C~IC~I\. c‘. W. JOH1.SrON and F. J. PIKO~~OLO (Editors), pp. 201 233, Lawrencc Erlbaum. Hillsdale. NJ. 1988. KAKNATH. H. 0. Dcficitb of attention in acute and recovered visual hcmi-neglect. .Yc,urop\!‘c,ho/oUirl 26, 27 43.
IW8
7. KAKNATH. H. 0. and HAKT.II W. Residual information processing in the neglected visual half-field. J. .l’r~rrcd 234. I X0 I X4, 1987. X KAKNA TH. H 0.. SC.HI.~~;KI.I P and FIX.HL.K, B. Trunk orientation as the determining factor of the “contr~~later;il“ deficit in the neglect syndrome and as the physical anchor of the internal representation of bodb orientation in space. Broir~ 114, 1997 2014. 1991. ‘9 KI%.woIxr+..M. Mechanisms of unilateral neglect. In :Vcur~,plf~.si~,/~~~/i~,[~/ tr~l .\‘c,~rrop’!,“/t,~/~~~/i~,~~/ :l.~prc I \ r>/ Spc~ricd h’cylr~r.M. JI I\UULKoII (Editor), pp. 69 X6. North Holland. New York, 1987. IO. KIUSISOI.KNI.. M. and WAKKINGION, E. A variety of reading disability associated with right hemisphere leGon\. J. :Ilrw~~/. !‘l’euro,\urq. P\j,c+ti
EYE MOVEMENT BEHAVIOUR DURING TEXT READING NEGLECT SYNDROME: A CASE STUDY HAM-OTTO KARUTH*
Department
of Neurology.
IN
and WALT~K HLHIX
Rheinisch-Westfilische
Technische
Hochschule
Aachcn.
Germany
(Rec~riwd19 June I99 I : c~c.cy~trd 2XFdmur~ 1992) Abstract-The eye movement hehavtour of a patient suffering from a right basal ganglia infarction with a left-sided hemineglect hut without any visual field defects was investigated during reading. The eye movements were registered by means ofan i.r. light technique (pupil corncal reflection method). The main findings were abnormal return sweeps. Whereas in normal readers the end ofonc line of text is linked to the beginning of the new line by a long leftward saccadc, the return sweeps of the hemineglect patient stereotyptcally ended in the middle of the next hne. They were followed by sequences of short saccades indicating silent backward reading until a linguistically plausible continuation of sentences from the previous line was found. irrespective of the actual beginning of text. The shortened return sweeps could not be attributed to a general oculomotor disturbance. The spatial border for the occurrence of the patient’s abnormal scanning pattern (left half of texts) clearly did not depend on a retinal coordinate frame of reference but rather has to he attributed to a different body-centred reference system.
INTRODUCTION FKOM C'LINK'AL descriptions of hemineglect, dcviatmns in reading are well known although there are only relatively omissions of few detailed experimental studies. Several authors (e.g. Refs [IO]. [ 121 and 1141) reported contralateral letters in single word reading, an issue whtch recently received new interest hecausc of its implications for cognitive models of reading [ 131. In contrast, whole text reading was rarely studied probably due to limitattons in registratton techniques. When normal subjects read a text, their eye movements exhibit the following phases scanning.At the beginning. readers show one long orienting saccade to the upper left corner a text succeded by a few short correcttvc saccadea in order to locate the gaze on the initial text constituent. Then. the first line ts scanned by a sequence ofsaccades and fixations. The left-to-right scanning may he interupted by regressive eye movements indicating dilhculties of visual and’or linguistic information processing. When the end of a lint is reached. normal readers show the follovving routine hehaviour: The last fixation is slightly extended and a return sweep is initiated, i.e. a long leftward saccade to the beginning ofthe next line. Usually a few corrective saccadcs are necessary before the scanntng ofthe new line can be started. It is obvious that left visual hemineglect after right brain damage would lead to difficulties especially tn two phases of the reading process, namely. searching for the beginning of the text and performing the return sweep from the end ofone hne to the beginning of the next one. In the present study. we report on the text reading bchaviour in ;t GLX of pure left visual hemincglect without vtsual field defects.
of
of
METHODS H.S.. a 56-year-old police officer. sustained a thromhotic occlusion at the bifurcation artery several hours after EC IC bypass surgery. A CT scan made 2 weeks po\t-onset *Address for correspondence: D-5100 Aachen. Germany.
Dr H. 0. Karnath.
Department
593
of Neurology.
of the right internal carotid revealed an extcndcd right
RWTH
Aachen.
Pauwelsstrassc,
594
NOTI
basiul ganglia lesion Icaving the primary \~\ual pathway intact. GOI ~xa\ perlmetry yelded no signs ofwaual lield dcfccts L7]. Nelthcr paresis ofgaze nor any other oculomotor abnormahties were observed. H.S. shoucd clear signs of a left-sided \~sual hemincglcct on the following tabkx: lint bisection. copying of a Ilwcr. completion of a clockfxc. dctcction of visualstimuli (clinical confrontation method) and naming oft;lchistoscopicall~ prrsented picturcr under bilateral stimulation (as opposed to unilateral stimulation). Ncglcct symptoms &crt‘ also ohserved in ocryday behabiour on the ward. A more detailed description of H.S. wzs gi\cn in Ref. 171. The eye mobcmcnt stud) w\ carried out 5 weeks pot-CVA.
Stimuli consisted ofn~nc written stories, which conceycd humorous everyday lift cpibodes. Each star) was mxlc up of six scntcnces. resulting in a total of Y x 6 = 54 sentences. Each scntencc started at the beginning of a new line and continued unto the succeeding one. Thus. each text extended over 12 lines. We >!stematicallq varied the cxtcnsion oftt’xt into each second lint per sentence. Some xntenccs ended far left of the midpoint, others in the right hnlfofthc line. Thih particular format oftcxts gave us the opportunitv to study the relationship between performIng roturn sweeps and finding the linguistic continuation of the individual sentences. Furthermore. wc baried the linguistic content at the beginning of each second lmc per sentences. In 30 instances a linguistically plausible continuation M#ould rc:hult even ifthc first constituent(~) on the xcond line waqwerel left out during wading. In the other 24 imianco. any omision of words hould hc linguistically unacccptahle. H.S. ~+a atkcd to read all nine text\. As a control group ue studied nine non-brain-damaged bubjcct, (ficc female, four malcj) ranging in age hctween 25 and 5X years. Each of the control subject\ &as presented with one ol the nine texts. Thu.4 we obtained 54 scntcnccs being read hq the control group as uell as by H.S.
The ICXIS were projected from slides fully colering the projection screen (75 cm high x IO0cm wide, 24 x 32 dc~l of visual angle). Each line contained X0 character spaces with ;I G/e of0.7 x 0.4 Eye movements sere reglstercd Kith the pup11 cornea1 rcflcction method 1161. The apparatw consisted of a microprocessor system (DEBIC 84). \+ hich \+a\ used with ;I conccntional cidro system (c)c- and scent-camera). The system allows 25 SO measurement\ per second. The mean Euclidian accuracy ib 0.7 with a standard dcliation (SD) of I .h [S]. The sub~ccts WI with their heads liked at a dlstancc of 175 cm from the projectwn wccn. Calibration took place immediately bcforc each registr-ntion. The subjca had to tinate 20 light cmitring diodes which here evenly distributed over tht- screen. Data reduction was achlcvcd b> a modified version of the program EMAN [I I]. Succesivc measurement point\ \\erc combined into “liuution” a long 3s they fell into a gliding window of0.5 The minimllm duration of fixation\ \\;I\ Xl ill 10 mscc.
RESULTS In Figs I and 2 ;Lcharacteristic scanpnth is shown for a normal subject and for the hemtneglect patlent H.S. Each of the two text lines is plotted twice. m order to illustrate the four phases of scnnn~np: acarching for the beginning. reading the first line. return weep. wading the second line.
As expected. the control suh,jects always reached the besinning ufscntcnccs in one long saccade foIlwed by a few corrective saccades (Fig. I ). H.S. had ditfculties which were rellected in both lrrcgular scanning and linguistic crrorb. He left out initial v,.ords when rending aloud in 31 out of 54 sentences (57 “:I). However. hi5 scanning demonstrated in the majority ofsentencez. (70”,,,) that hc. ncccrtheless. directed hi5 gare lefthard\. ix. towards the beginning of tat. During reading aloud. H.S. sho\\cd normal scanmng parameters for some IIIKS. but for other\ wbnormal lxha+iour a illustrated in l.‘ip. 2. (For ;I more dctailcd descrlptlon of these pha\ch of the reading prows.\ see Ref. [.I]. J
H S. shv~\sd ;I .rtnklng dilTcrcncc to ihc long saccadcq normal rcadcrs made Icftwnrds ac‘ro\s the H hole lint (cf. FIFE I and 2). Whcreab the mean (dlsgonal) length of the return aweps in the control group was 24.4 (SD =4.X I. IH.S. only performed an averaged return suecp of I I .5 (SD = 3.5 I. The return s!%eepsofthe controls Icd them close to the bcglnninp of the text in the second lone ofthc vzntcncca (distance to the hegInning: mean =4.X (SD=?.0 )I. \vherca H.S. landed with ;I distance of 17.1 (SD -3 7 ). On the a\cragc. the return s\he~‘pbof H.S. were halfas long ;I\ the ona of normal controls. and their landing positions fell into the middle of the next lint. The distance between the starting point of the return awecps of H.S. and the end oftext in line I could not explain the ~Iivrtcncd rclurn s\lccps (controls: mean 7 I .? (ST)-j.4 ) H.S.: 1.3 (SD1.6 )I. Another possible cuplatwtion for this beha\iour could bc that the return sweeps of H.S. hcrc aimed at the end-points of the written test in c~ch wcond lint (the lcngh of \\hlch \\as \~.,ternat~call> varied XI-OS\ the 54 wntencca). but the Spearman
NOTE
ByI
AT
ElNEn
OACKEL FIEL
595
EINEM
ALTEREN
HE4RN Pi_OTZLlCcl
vor*r
BE
\I
260
BLUMENTOPF AUF DEN KOPF.
Fig. I. Example for reading phases in a normal control subject (male. S9 qears): (I ) searching for the beginning of the sentence: (2~ reading aloud of the first line of the sentence: (3) return sweep and corrective saccades: (4) reading aloud of the second line of the sentence. (Fixations with a duration of 160 msec and more are plotted as numbers.)
rank-order correlation between the length ofthe written text and the length of the return sweeps was low (i/=0.33). In H.S.. we observed overshooting return sweeps when the text ended in the right half of the projection screen. i.e. landing of gaze within the text. In contrast, undershoots were found with texts ending in the left halfof the screen. In contrast to the previously reported correlation. the length ofover- ‘undershoots was highly correlated (ii = 0.77 I Mith the length of the written text in the line indicating that the return sweeps in their majority landed closer to the midpotnt of the line than to the end-points of the written text (f’eLerlty of hemineglect on the rending behakiour. W&I.KYK rt trl. (personal communication) found no return sweeps during text reading in their left-sided hemineglect patient B.Q. [ 151.The patient was unable to find a linguistically plauaiblc continuation oftext from one line to the other. She complained about it. but. nevertheless. did not show any systematic backward reading a5 found in our patient H.S. Different components of the neglect sqndromc L3. h] are prominent in the two patients. which ma) reflect different degrees of severity. B.Q. showed an attentional orientation bias to the ipsilateral side. Ia\ w H.S. The cognitive demands of text reading apparently have oberrldden an Ipsilateral orientation txas m H.S. Thus. his di?turhance of the internal egocentric frame for rcprcscnting spatial orlcntation hccamc more apparent.
.~~~,lo~~/t,dyr,tlr,lr.,~Thls study was supported by a grant from the Deutsche For\chungsgemelnschaft. Dr K. Willmes for helpful comments on the manuscript.
We thank
NOII
REFERENCES I. 2. 3. 4. 5. 6.
BK.zI>. W. R. Vtsual diaoricntat~on \*ith special rcfcrcncc to lesion& of the rlghr ccrcbral hemisphere. Brtrirl 64, 244 272. 1941. Dl RI NII. E.. GI.~III.IYI. M.. FA(;I.IO\I, P. and BAKHII:KI. C Attcntional shift tov,nrds the rightmost btunull 111 patients Wllh icrt visual neglect. C0rrr.x 25, 23 I 737, 1989. G~INOTT~. G.. D'FKMI, P. and DI. Bows. C. Aspects cliniqucs et mtcanlames de la n@ligence visuo-\patialc. Krr. ,Yrwo/. ( t’wi.5) 145, 626 634. 19X9. HUHI K. W.. Gr ILLOI. Ci. and KAKNATH, H. 0. Blickbcwe~ungsstrateFicn hcim L,esen mit linkascltiger Hemianopsie und Hcmincglect. %. L/in. /‘\yckol. 17, 244 259. 198X. H~I+I.K. W., Li.11K. G. and L.ASS.C. Eye movement hehavlor III aphasia. In \r,lrrop.\!,c,/~ol(~~/~ of f$, :Mor C~IC~I\. c‘. W. JOH1.SrON and F. J. PIKO~~OLO (Editors), pp. 201 233, Lawrencc Erlbaum. Hillsdale. NJ. 1988. KAKNATH. H. 0. Dcficitb of attention in acute and recovered visual hcmi-neglect. .Yc,urop\!‘c,ho/oUirl 26, 27 43.
IW8
7. KAKNATH. H. 0. and HAKT.II W. Residual information processing in the neglected visual half-field. J. .l’r~rrcd 234. I X0 I X4, 1987. X KAKNA TH. H 0.. SC.HI.~~;KI.I P and FIX.HL.K, B. Trunk orientation as the determining factor of the “contr~~later;il“ deficit in the neglect syndrome and as the physical anchor of the internal representation of bodb orientation in space. Broir~ 114, 1997 2014. 1991. ‘9 KI%.woIxr+..M. Mechanisms of unilateral neglect. In :Vcur~,plf~.si~,/~~~/i~,[~/ tr~l .\‘c,~rrop’!,“/t,~/~~~/i~,~~/ :l.~prc I \ r>/ Spc~ricd h’cylr~r.M. JI I\UULKoII (Editor), pp. 69 X6. North Holland. New York, 1987. IO. KIUSISOI.KNI.. M. and WAKKINGION, E. A variety of reading disability associated with right hemisphere leGon\. J. :Ilrw~~/. !‘l’euro,\urq. P\j,c+ti
