- Email: [email protected]
Graphic Neglect—More Than the Sum of the Parts
NeuroImage 14, S91–S97 (2001) doi:10.1006/nimg.2001.0821, available online at http://www.idealibrary.com on
Graphic Neglect—More Than the Sum of the Parts Peter W. Halligan* ,1 and John C. Marshall† *School of Psychology, Cardiff University, Cardiff CF10 3YG, United Kingdom; and †University Department of Clinical Neurology, Radcliffe Infirmary, Oxford OX2 6HE, United Kingdom Received March 15, 2001
Patients who seem to “ignore” objects or people on one side of space have been described in the medical literature for well over a century. The term “visuospatial neglect” is now used to describe the cluster of behaviors whereby patients after unilateral cerebral lesions (most frequently of right parietal cortex) fail to attend or explore (predominantly) the side of space contralateral to the lesion. Although the condition comprises a complex disruption of space-related behaviors, the prevailing view was that the different symptoms could be accommodated in terms of damage to one of three different cognitive mechanisms mediating attention (e.g., K. M. Heilman and E. Valenstein, Ann. Neurol. 5: 166 –170, 1979), intention (R. T. Watson, E. Valenstein, and K. Heilman, Ann. Neurol. 3: 505–508, 1978), and/or representation (E. Bisiach, Q. J. Exp. Psychol. 46: 435– 461, 1993). The general consensus favors an attentional deficit but the notion of attention has always proved conceptually slippery and difficult to operationalize (P. W. Halligan and J. C. Marshall, Cogn. Neuropsychol. 11: 167–206, 1994a). In this paper, we consider how drawing performance after right brain damage in patients with “visual neglect” reveals the involvement and interplay of several cognitive deficits, including aspects of mental representation and spatial awareness. © 2001 Academic Press Key Words: spatial cognition; unilateral visual neglect; parietal cortex; visual arts.
INTRODUCTION One reason that so many competing accounts have been put forward to explain neglect stems from the relatively poor understanding of what constitutes the functional architecture of normal spatial cognition. Current notions of psychological “space” suggest a complex interplay between distinct (and often hemispheri1 To whom correspondence and reprint requests should be addressed at School of Psychology, Cardiff University, P.O. Box 901, Cardiff CF10 3YG, UK. Fax: 029 2087 4858. E-mail: [email protected].
cally lateralized) systems that deal with what is where, how to act accordingly, and what we are conscious of at any one time. Kinsbourne (1974) stressed that there are many lateralized systems in the normal brain that interact in the detection and exploration of “ambient changes in the environment.” The time-honored clinical syndromes into which impairments of visual and spatial functions are classified were not preestablished in any theoretically principled fashion. Given the association of severe and persistent neglect with large right hemisphere lesions, it is likely that the observed symptoms are due to interactions between “neglect(s)” and other damaged right hemisphere functions (De Renzi, 1982; Grusser and Landis, 1991). Yet relatively early in its description as a neuropsychological entity, unilateral neglect “became conceptually more and more segregated within a syndrome of which it constituted the most frequent but by no means the sole or even necessary feature” (Bisiach and Berti, 1986). On the assumption that the three main theoretical accounts of visuospatial neglect are not mutually exclusive, selective damage to attentional, intentional, and representational processes may provide the basis for a coarse-grained taxonomy of relevant behavioral symptoms. By adopting finer classifications, “it should be possible to identify subsystems of spatial cognition which are equal in complexity to those elucidated for language and memory processes ” (Mattingley et al., 1992). However, since large naturally occurring lesions do not respect post hoc selective functional explanations, differential damage to a range of specific cognitive processes will interact to bring about the characteristic behaviors traditionally associated with visuospatial neglect. It may accordingly be difficult to explain neglect in a particular patient by any single one of the three accounts referenced earlier (Caramazza and Hillis, 1990; Halligan and Marshall, 1994a). The problem, then, remains how to extrapolate meaningful subtypes of neglect in the absence of an explicit unifying model of normal visuospatial processing that can constrain the different interpretations of pathological performance.
S91
1053-8119/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.
S92
HALLIGAN AND MARSHALL
BACK TO BASICS One way of addressing this problem is to carry out a close(r) examination of some task-demands involved in the clinical tests used to diagnose and define neglect in the first place (Halligan and Marshall, 1992). Although it is often forgotten, once the diagnosis has been made, these tests continue to provide crucial operational definitions and entry criteria for theoretical speculation and further empirical study of neglect. Many of these tests are decidedly pre-theoretical and originate from clinical bedside procedures developed in the early parts of the 20th century: Examples include cancellation, line bisection, copying and spontaneous drawing. Group studies of visual neglect sometimes set the inclusion criteria for diagnosing neglect as impaired performance on any one of these tests (Ogden, 1985), although other investigations emphasize an aggregate score based on several tests (Halligan et al., 1991). Until the late 1980s most of these tests were considered measures of the same underlying disturbance, albeit with differential sensitivity (Halligan et al., 1989). Subsequent single case and group reports revealed a different picture (Barberi and De Renzi, 1989; Marshall and Halligan, 1995; Cubelli et al., 1991). Halligan and Marshall (1992) demonstrated a reliable classic double dissociation between line bisection and cancellation (two prototypical neglect tasks) in two patients with unilateral right hemisphere stroke. There must accordingly be a functional (and presumably a neurophysiological and anatomical) distinction between the processes involved in successful execution of these two tasks. Supporting evidence, with larger groups of patients, has been reported by Binder et al. (1992) and McGlinchey-Berroth et al. (1996). Reports of dissociations between the standard tasks used to define visual neglect should come as no surprise if one considers the different complex cognitive and spatial demands tapped by these individual tasks (Pizzamiglio et al., 1992; Halligan and Marshall, 1992; Kinsella et al., 1993). Although these tasks may share a common “face validity” with regard to visual processing in peripersonal space, simple task analysis illustrates how these conventional tests selectively sample different forms of spatially specific impairment. Factor analytic studies of large groups of right brain damaged patients support this claim. Using a battery of six traditional clinical tests of neglect, Kinsella et al. (1993) showed that the pattern of results could be accounted for by two main factors: attentional scanning and internal representation. An examination of the relationship between a similar group of neglect tests using a cluster analysis based on the average linkage method (Aldefender and Blashfield, 1984) demonstrated significant intratest differences (Pizzamiglio et al., 1992). In particular, this study concluded that drawing tasks, which depend heavily on mental spatial
FIG. 1. Sketch of a parrot by a right-brain-damaged patient with visuospatial neglect.
representations“ must be considered to be relatively independent of the other tests.” Thus far, most task analysis studies have focused on the differences between cancellation and bisection performance (Halligan and Marshall, 1992, 1998; Marshall and Halligan, 1996; Agrell et al., 1997). However, florid “visual neglect” on copying and drawing may provide more insight into the condition (see Fig. 1). The skill shown by this amateur sketcher in drawing what he does include could almost convince one that the omissions are a deliberate artistic strategy . . . except, for the fact he always neglects the left, including food located on the left side of his plate! For many textbooks of neuropsychology and behavioral neurology, such illustrations provide indisputable evidence of a deficit that “amputates or distorts” the mental representation of the visual world (Bisiach, 1993; Bisiach et al., 1996). Such graphic neglect is equally evident in the work of professional painters and sculptors who have suffered right parietal stroke (Halligan and Marshall, 1997). NEGLECT IN THE VISUAL ARTS Artistic expression requires the integration of many basic cognitive processes concerned with perceptual, motor, and representational systems. But practice and skill in these domains has little prophylactic value: The work of several professional artists has been significantly affected by their visual neglect following right hemisphere lesions (see Cantagallo and Della Sala,
S93
GRAPHIC NEGLECT
1998, and Mazzucchi et al., 1994 for review). In his book “Art, Mind and Brain: A Cognitive Approach to Creativity” (1982), Howard Gardner illustrates how the work of Lovis Corinth (1858 –1925), the German painter on the boundary between impressionism and expressionism, was influenced by neglect following his stroke in 1911. Other painters whose work was affected by visual neglect include Otto Dix (1891–1969) and Anton Raderscheidt (1892–1970). In 1990, we had an opportunity to study the effects of visual neglect on the work of the distinguished English artist Tom Greenshields. Tom expressed himself visually using his left hand, having lost the use of his right hand in a farming accident. In 1989 he suffered a stroke, which included a mild sensory–motor weakness of his left arm and leg. Despite making a good physical recovery, Tom (in addition to other visuoconstructive problems) was unable to attend appropriately to the left part of what he was attempting to draw and sculpt (Halligan and Marshall, 1997). Figure 2a shows the high degree of artistic skill he had acquired using his left hand prior to his stroke. By contrast, Fig. 2b shows his performance several months after his right hemisphere stroke; the figure is poorly proportioned and parts of the left side are evidently distorted or omitted. The lack of correct proportion shows an evident impairment in the ability to relate parts to the whole. This impairment is also seen in the addition of a supernumerary toe. His neglect performance did not result from a subclinical motor deficit or hypometria; when drawing Tom clearly demonstrated that he was able to reach and make use of the left side of the large page despite omitting salient features of objects and figures contained therein. Figure 3 provides a good illustration of “object-based” neglect. Although both candlestick holder and face clearly show neglect for their respective left sides (as viewed by the artist), the candlestick is not totally ignored despite being located to the left of the face. CLINICAL TESTS AND REPRESENTATIONAL NEGLECT Tests such as line bisection and cancellation provide quantifiable and reliable indications of deficits in spatial search and/or spatial judgement. The extent to which either depends on mental imagery or the internal generation of mental representations is, by comparison with drawing at least, minimal. Although drawing can establish the integrity of the patient’s mental representations, adequate performance also depends on relatively intact manual motor execution. The first persuasive clinical evidence that avoided the necessity of manual response arose from the pioneering study by Bisiach and Luzzatti in 1978: They found that right brain damaged patients with visual neglect
tended to omit details from the left side of their memorial recollections of familiar scenes. When asked to describe from memory these known views from a given vantage point, patients would fail to include what was on the left from that viewpoint. If the viewpoint was reversed, items on the left from this new vantage point were now neglected while details from the previously neglected “left side,” were now reported. This finding of “representational neglect” has been confirmed many times. Although lateral cueing in these imagery tasks (Bisiach et al., 1981) could improve or worsen performance, the data were interpreted as providing the first unequivocal evidence that internal analogue space representation had been compromised in patients with visuospatial neglect. Subsequent reports of “pure” cases of representational neglect without visual neglect on other traditional tasks (Guariglia et al., 1993; Beschin et al., 1997; Peru and Zapparoli, 1999) provide further support for the argument that representational and perceptual spatial domains are subserved by (at least partially) independent neural mechanisms. GRAPHIC TASK ANALYSIS Unlike more easily quantifiable tasks such as bisection and cancellation, how one performs drawing/copying tasks is not fully constrained by the stimulus presented. Graphic tasks typically require some attempt to produce relevant local parts (in sequence) that are appropriately related to the overall global pattern. Moreover, graphic tests permit online perceptual feedback which in turn can allow for cross-cueing and selfcorrection. When drawing from memory, the patient’s initial response comes before any stimulus is perceived. Unlike bisection, copying and cancellation, there is nothing on the page prior to the patient commencing the task. Finally, perceptual analysis of one’s ongoing drawing involves a direct comparison of the mental representation of what one wants to draw and what has been drawn at any one time. In some cases, omissions in drawing decrease dramatically when such perceptual feedback is reduced or removed (Halligan and Marshall, 1994a; Mesulam, 1985): The patients make more accurate drawings with their eyes closed than open! Impairments of graphic skill that might be attributed to neglect can therefore result from at least three distinct types of deficit: lateralized impairment to (1) manual motor performance (including hypometria); (2) memorial recall and/or representational image generation; and (3) online visual perceptual feedback from the drawing itself (attentional). Drawing tasks thus sample a wider set of potential cognitive impairments than do other traditional neglect tests and may accordingly help explain the reported differences in sensitivity and dissociations between drawing/copying and line bisection and
S94
HALLIGAN AND MARSHALL
FIG. 2. Drawings by a professional artist, Tom Greenshields, before (a) and after (b) the artist’s stroke.
S95
GRAPHIC NEGLECT
FIG. 3.
Drawing by Tom Greenshields showing “object-based neglect” for two objects (candlestick and face).
cancellation (Pizzamigia et al., 1992; Kinsella et al., 1993). Figures 4a and 4b show something of the range of neglect impairments revealed by drawing. The patient fails to complete local features on the left side [the left eye in (a) and the left upper torso, left eye and left contour of the left leg in (b)]. This pattern of performance is difficult to reconcile with a general impairment of attention or with hypometria since many leftsided features are included and the overall composition is relatively well preserved. Although the patient knows full well that bodies and faces have a left side (and can verbally express this fact), she repeatedly neglects features on the left when drawing or copying. Furthermore, when looking at her supposedly “finished” drawing, she fails to notice that she has not overtly completed the left side. The left side of her spatial (but not verbal) mental representation may have failed to be activated and hence is not expressed in her drawing. Such failure might result from a local fault in
the structure that implements internal space representation. Alternatively, the deficit could stem from failure to attend to parts of the analogical mental image and hence not transcribe those parts in the overt drawing. PERCEPTUAL AWARENESS Both the above accounts, however, beg the question as to why patients fail to notice what they have done (or rather not done). Current theoretical approaches to neglect cannot readily reconcile why many patients unimpaired on one aspect of a task are grossly impaired on another aspect involving the very same stimulus. To merely say that the patients have left neglect is of course a description and not an explanation of their problems. Although not always explicitly stated, the common assumption is that such patients are not “aware” of their neglect, but this is often not true: many patients with intractable chronic neglect have
S96
HALLIGAN AND MARSHALL
volves a major omission within left space and an equally severe distortion within right space cannot adequately be explained by any of the three main accounts of neglect as they currently stand. CONCLUSION
FIG. 4. (a and b) Drawings by a patient with a right hemisphere stroke who considered both sketches to be complete.
considerable conceptual and indeed experiential insight into their deficit and its consequences (see Halligan and Marshall, 1998). Furthermore, unlike cancellation and line bisection where feedback is stimulus specific, the patient in a graphic task has access to a large body of long-term semantic and propositional memories that should serve as reminders of what to include in a drawing. One account of this aberrant perceptual awareness stresses the notion of “pathological completion,” a positive symptom whereby patients mentally “fill-in” the missing information (presumably on the basis of congruence between their current perceptual experience and semantic knowledge). Halfdrawings are thus believed to be fully complete (Halligan and Marshall, 1994b). This apparent lack of awareness of the deficit (Halligan and Marshall, 1998), even when the incongruity of the patients’ asymmetric productions is repeatedly pointed out to them stands in urgent need of explanation. Pathological left side “completion” is thought to be implicated in some, but not all, forms of impaired clock drawing (see Fig. 5). The patient who placed all 12 numerals along the right hemi-circumference of the clock-face insisted (in free vision and without time limit) that his reproduction looked just fine even when compared with that of a normal clock face. Such behavior which clearly in-
Visual neglect is no longer considered a single monolithic disorder but rather a multiplicity of cognitive deficits that collectively result in a lateralized disturbance of behavioral responses in space. Over the past decade, the use of conventional tests has highlighted the fact that they can produce differential and at times fully dissociable results within and between patients with neglect. This is hardly surprising given that these tests were never explicitly designed to selectively sample what is now considered the range of different cognitive impairments involved in visual neglect. These task-specific dissociations occur despite the fact that most of the traditional clinical tests are much less than pure in eliciting the contribution of specific cognitive deficits. Impaired performance on these “impure” tasks has been described as “perceptual,” “representational,” “attentional,” “intentional,” or “motor” neglect on the basis of a superficial characterization of the eliciting conditions. Graphic tasks appear simple but nonetheless require the coordination of many different skills with distinct cerebral substrates: the patient’s final production often involves interactions between his or her selective deficits and the struggle to compensate for impairment with whatever relevant functions remain. In drawing from memory, there must be an internal representation of the visual spatial features of the object to be delineated, as well as a mechanism for retrieving this knowledge. Complex visual perceptual and visual motor processes are then needed to translate the mental representation into an appropriate motor program for a “realistic” drawing. Further reconsideration of simple graphic performance by patients with visual ne-
FIG. 5. Two qualitatively different drawings of a clock face by patients with right brain damage.
GRAPHIC NEGLECT
glect may revitalize the search for the interplay between different cognitive processes and additional previously neglected pathological features implicated in this complex disorder. ACKNOWLEDGMENT P.W.H. and J.C.M. are supported by the Medical Research Council (UK).
REFERENCES Aldefender, M., and Blashfield, R. 1984. Cluster Analysis. Sage Publications, Beverly Hills. Agrell, B. M., Dehlin, O. I., and Dahlgren, C. J. 1997. Neglect in elderly stroke patients: A comparison of five tests. Psychiat. Clin. Neurosci. 51: 295–300. Barbieri, C., and De Renzi, E. 1989. Patterns of neglect dissociation. Behav. Neurol. 2: 13–24. Beschin, N., Cocchini, G., Della Sala, S., and Logie, R. H. 1997. What the eyes perceive, the brain ignores: A case of pure unilateral representational neglect. Cortex 33: 3–26. Binder, J., Marshall, R., Lazar, R., Benjamin, J., and Mohr, J. P. 1992. Distinct syndromes of hemineglect. Arch. Neurol. 49: 1187– 1194. Bisiach, E., and Luzzatti, C. 1978. Unilateral neglect of representational space. Cortex 14: 129 –133. Bisiach, E., Capitani, E., Luzzatti, C., and Perani, D. 1981. Brain and conscious representation of outside reality. Neuropsychologia 19: 543–551. Bisiach, E., and Berti, A. 1986. Representational impairment as a factor in neglect. In Neurology (K. Poeck, H. J. Freund, and H. Ganshirt, Eds.). Springer-Verlag, Berlin, Heidelberg. Bisiach, E. 1993. Mental representation in unilateral neglect and related disorders. The Twentieth Bartlett Lecture. Q. J. Exp. Psychol. 46: 435– 461. Bisiach, E., Pizzamiglio, L., Nico, D., and Antonucci, G. 1996. Beyond unilateral neglect. Brain 119: 851– 857. Cantagallo, A., and Della Sala, S. 1998. Preserved insight in an artist with extrapersonal spatial neglect. Cortex 34: 163–189. Caramazza, A., and Hillis, A. E. 1990. Levels of representations, co-ordinate frames and unilateral neglect. Cogn. Neuropsychol. 7: 391– 445. Cubelli, R., Nichelli, P., Bonito, V., De Tanti, A., and Inzaghi, M. G. 1991. Different patterns of dissociation in unilateral spatial neglect. Brain Cogn. 15: 139 –159. De Renzi, E. 1982. Disorders of Space Exploration and Cognition. Wiley, New York. Grusser, O.-J., and Landis, T. 1991. Visual Agnosias and Other Disturbances of Visual Perception and Cognition. Macmillan, London. Guariglia, C., Padovani, A., Pantano, P., and Pizzamiglio, L. 1993. Unilateral neglect restricted to visual imagery. Nature 15: 235– 237.
S97
Halligan, P. W., Marshall, J. C., and Wade, D. T. 1989. Visuo-spatial neglect: Underlying factors and test sensitivity. Lancet 2: 908 – 911. Halligan, P. W., Cockburn, J., and Wilson, B. A. 1991. The behavioural assessment of visual neglect. Neuropsychol. Rehabil. 1: 5–32. Halligan, P. W., and Marshall, J. C. 1992. Left visuo-spatial neglect: A meaningless entity? Cortex 28: 525–535. Halligan, P. W., and Marshall, J. C. 1994a. Toward a principled explanation of unilateral neglect. Cogn. Neuropsychol. 11: 167– 206. Halligan, P. W., and Marshall, J. C. 1994b. Completion in visuospatial neglect: A case study. Cortex 30: 685– 694. Halligan, P. W., and Marshall, J. C. 1997. The art of visual neglect. Lancet 350: 139 –140. Halligan, P. W., and Marshall, J. C. 1998. Neglect of awareness. Conscious. Cogn. 7: 356 –380. Heilman, K. M., and Valenstein, E. 1979. Mechanisms underlying hemispatial neglect. Ann. Neurol. 5: 166 –170. Kinsbourne, M. 1974. Mechanisms of hemispheric interaction in man. In Hemispheric Disconnection and Cerebral Function (M. Kinsbourne and W. L. Smith, Eds.), pp. 260 –285. C. C. Thomas, Springfield, IL. Kinsella, G., Oliver, J., Ng, K., Packer, S., and Stark, R. 1993. Analysis of the syndrome of unilateral neglect. Cortex 29: 135–140. McGlinchey-Berroth, R., Bullis, D. P., Milberg, W. P., Verfaellie, M., Alexander, M., and D’Esposito, M. 1996. Assessment of neglect reveals dissociable behavioural but not neuroanatomical subtypes. J. Int. Neuropsychol. Soc. 2: 441– 451. Marshall, J. C., and Halligan, P. W. 1996. Hemispheric antagonism in visuo-spatial neglect: A case study. J. Int. Neuropsychol. Soc. 2: 412– 418. Marshall, J. C., and Halligan, P. W. 1995. Within- and between-task dissociations in visuo-spatial neglect: A case study. Cortex 31: 367–376. Mattingley, J. B., Bradshaw, J. L., and Phillips, J. G. 1992. Reappraising unilateral neglect. Austr. J. Psychol. 44: 163–169. Mazzucchi, A., Pesci, G., and Trento, D. 1994. Cervello E Pittura. Fratelli Palombi Editori, Rome. Mesulam, M. 1985. Principles of Behavioural Neurology. F. A. Davis, Philadelphia. Ogden, J. A. 1985. Anterior–posterior interhemispheric differences in the loci of lesions producing visual hemineglect. Brain Cogn. 4: 59 –75. Pizzamiglio, L., Bergego, C., Halligan, P., Homberg, V., Robertson, I., Weber, E., Wilson, B., Zoccolotti, P., and Deloche, G. 1992. Factors affecting the clinical measurement of visuo-spatial neglect. Behav. Neurol. 5: 233–240. Peru, A., and Zapparoli, P. 1999. A new case of representational neglect. Ital. J. Neurol. Sci. 20: 243–246 Watson, R. T., Valenstein E., and Heilman, K. 1978. Nonsensory neglect. Ann. Neurol. 3: 505–508.
Graphic Neglect—More Than the Sum of the Parts Peter W. Halligan* ,1 and John C. Marshall† *School of Psychology, Cardiff University, Cardiff CF10 3YG, United Kingdom; and †University Department of Clinical Neurology, Radcliffe Infirmary, Oxford OX2 6HE, United Kingdom Received March 15, 2001
Patients who seem to “ignore” objects or people on one side of space have been described in the medical literature for well over a century. The term “visuospatial neglect” is now used to describe the cluster of behaviors whereby patients after unilateral cerebral lesions (most frequently of right parietal cortex) fail to attend or explore (predominantly) the side of space contralateral to the lesion. Although the condition comprises a complex disruption of space-related behaviors, the prevailing view was that the different symptoms could be accommodated in terms of damage to one of three different cognitive mechanisms mediating attention (e.g., K. M. Heilman and E. Valenstein, Ann. Neurol. 5: 166 –170, 1979), intention (R. T. Watson, E. Valenstein, and K. Heilman, Ann. Neurol. 3: 505–508, 1978), and/or representation (E. Bisiach, Q. J. Exp. Psychol. 46: 435– 461, 1993). The general consensus favors an attentional deficit but the notion of attention has always proved conceptually slippery and difficult to operationalize (P. W. Halligan and J. C. Marshall, Cogn. Neuropsychol. 11: 167–206, 1994a). In this paper, we consider how drawing performance after right brain damage in patients with “visual neglect” reveals the involvement and interplay of several cognitive deficits, including aspects of mental representation and spatial awareness. © 2001 Academic Press Key Words: spatial cognition; unilateral visual neglect; parietal cortex; visual arts.
INTRODUCTION One reason that so many competing accounts have been put forward to explain neglect stems from the relatively poor understanding of what constitutes the functional architecture of normal spatial cognition. Current notions of psychological “space” suggest a complex interplay between distinct (and often hemispheri1 To whom correspondence and reprint requests should be addressed at School of Psychology, Cardiff University, P.O. Box 901, Cardiff CF10 3YG, UK. Fax: 029 2087 4858. E-mail: [email protected].
cally lateralized) systems that deal with what is where, how to act accordingly, and what we are conscious of at any one time. Kinsbourne (1974) stressed that there are many lateralized systems in the normal brain that interact in the detection and exploration of “ambient changes in the environment.” The time-honored clinical syndromes into which impairments of visual and spatial functions are classified were not preestablished in any theoretically principled fashion. Given the association of severe and persistent neglect with large right hemisphere lesions, it is likely that the observed symptoms are due to interactions between “neglect(s)” and other damaged right hemisphere functions (De Renzi, 1982; Grusser and Landis, 1991). Yet relatively early in its description as a neuropsychological entity, unilateral neglect “became conceptually more and more segregated within a syndrome of which it constituted the most frequent but by no means the sole or even necessary feature” (Bisiach and Berti, 1986). On the assumption that the three main theoretical accounts of visuospatial neglect are not mutually exclusive, selective damage to attentional, intentional, and representational processes may provide the basis for a coarse-grained taxonomy of relevant behavioral symptoms. By adopting finer classifications, “it should be possible to identify subsystems of spatial cognition which are equal in complexity to those elucidated for language and memory processes ” (Mattingley et al., 1992). However, since large naturally occurring lesions do not respect post hoc selective functional explanations, differential damage to a range of specific cognitive processes will interact to bring about the characteristic behaviors traditionally associated with visuospatial neglect. It may accordingly be difficult to explain neglect in a particular patient by any single one of the three accounts referenced earlier (Caramazza and Hillis, 1990; Halligan and Marshall, 1994a). The problem, then, remains how to extrapolate meaningful subtypes of neglect in the absence of an explicit unifying model of normal visuospatial processing that can constrain the different interpretations of pathological performance.
S91
1053-8119/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.
S92
HALLIGAN AND MARSHALL
BACK TO BASICS One way of addressing this problem is to carry out a close(r) examination of some task-demands involved in the clinical tests used to diagnose and define neglect in the first place (Halligan and Marshall, 1992). Although it is often forgotten, once the diagnosis has been made, these tests continue to provide crucial operational definitions and entry criteria for theoretical speculation and further empirical study of neglect. Many of these tests are decidedly pre-theoretical and originate from clinical bedside procedures developed in the early parts of the 20th century: Examples include cancellation, line bisection, copying and spontaneous drawing. Group studies of visual neglect sometimes set the inclusion criteria for diagnosing neglect as impaired performance on any one of these tests (Ogden, 1985), although other investigations emphasize an aggregate score based on several tests (Halligan et al., 1991). Until the late 1980s most of these tests were considered measures of the same underlying disturbance, albeit with differential sensitivity (Halligan et al., 1989). Subsequent single case and group reports revealed a different picture (Barberi and De Renzi, 1989; Marshall and Halligan, 1995; Cubelli et al., 1991). Halligan and Marshall (1992) demonstrated a reliable classic double dissociation between line bisection and cancellation (two prototypical neglect tasks) in two patients with unilateral right hemisphere stroke. There must accordingly be a functional (and presumably a neurophysiological and anatomical) distinction between the processes involved in successful execution of these two tasks. Supporting evidence, with larger groups of patients, has been reported by Binder et al. (1992) and McGlinchey-Berroth et al. (1996). Reports of dissociations between the standard tasks used to define visual neglect should come as no surprise if one considers the different complex cognitive and spatial demands tapped by these individual tasks (Pizzamiglio et al., 1992; Halligan and Marshall, 1992; Kinsella et al., 1993). Although these tasks may share a common “face validity” with regard to visual processing in peripersonal space, simple task analysis illustrates how these conventional tests selectively sample different forms of spatially specific impairment. Factor analytic studies of large groups of right brain damaged patients support this claim. Using a battery of six traditional clinical tests of neglect, Kinsella et al. (1993) showed that the pattern of results could be accounted for by two main factors: attentional scanning and internal representation. An examination of the relationship between a similar group of neglect tests using a cluster analysis based on the average linkage method (Aldefender and Blashfield, 1984) demonstrated significant intratest differences (Pizzamiglio et al., 1992). In particular, this study concluded that drawing tasks, which depend heavily on mental spatial
FIG. 1. Sketch of a parrot by a right-brain-damaged patient with visuospatial neglect.
representations“ must be considered to be relatively independent of the other tests.” Thus far, most task analysis studies have focused on the differences between cancellation and bisection performance (Halligan and Marshall, 1992, 1998; Marshall and Halligan, 1996; Agrell et al., 1997). However, florid “visual neglect” on copying and drawing may provide more insight into the condition (see Fig. 1). The skill shown by this amateur sketcher in drawing what he does include could almost convince one that the omissions are a deliberate artistic strategy . . . except, for the fact he always neglects the left, including food located on the left side of his plate! For many textbooks of neuropsychology and behavioral neurology, such illustrations provide indisputable evidence of a deficit that “amputates or distorts” the mental representation of the visual world (Bisiach, 1993; Bisiach et al., 1996). Such graphic neglect is equally evident in the work of professional painters and sculptors who have suffered right parietal stroke (Halligan and Marshall, 1997). NEGLECT IN THE VISUAL ARTS Artistic expression requires the integration of many basic cognitive processes concerned with perceptual, motor, and representational systems. But practice and skill in these domains has little prophylactic value: The work of several professional artists has been significantly affected by their visual neglect following right hemisphere lesions (see Cantagallo and Della Sala,
S93
GRAPHIC NEGLECT
1998, and Mazzucchi et al., 1994 for review). In his book “Art, Mind and Brain: A Cognitive Approach to Creativity” (1982), Howard Gardner illustrates how the work of Lovis Corinth (1858 –1925), the German painter on the boundary between impressionism and expressionism, was influenced by neglect following his stroke in 1911. Other painters whose work was affected by visual neglect include Otto Dix (1891–1969) and Anton Raderscheidt (1892–1970). In 1990, we had an opportunity to study the effects of visual neglect on the work of the distinguished English artist Tom Greenshields. Tom expressed himself visually using his left hand, having lost the use of his right hand in a farming accident. In 1989 he suffered a stroke, which included a mild sensory–motor weakness of his left arm and leg. Despite making a good physical recovery, Tom (in addition to other visuoconstructive problems) was unable to attend appropriately to the left part of what he was attempting to draw and sculpt (Halligan and Marshall, 1997). Figure 2a shows the high degree of artistic skill he had acquired using his left hand prior to his stroke. By contrast, Fig. 2b shows his performance several months after his right hemisphere stroke; the figure is poorly proportioned and parts of the left side are evidently distorted or omitted. The lack of correct proportion shows an evident impairment in the ability to relate parts to the whole. This impairment is also seen in the addition of a supernumerary toe. His neglect performance did not result from a subclinical motor deficit or hypometria; when drawing Tom clearly demonstrated that he was able to reach and make use of the left side of the large page despite omitting salient features of objects and figures contained therein. Figure 3 provides a good illustration of “object-based” neglect. Although both candlestick holder and face clearly show neglect for their respective left sides (as viewed by the artist), the candlestick is not totally ignored despite being located to the left of the face. CLINICAL TESTS AND REPRESENTATIONAL NEGLECT Tests such as line bisection and cancellation provide quantifiable and reliable indications of deficits in spatial search and/or spatial judgement. The extent to which either depends on mental imagery or the internal generation of mental representations is, by comparison with drawing at least, minimal. Although drawing can establish the integrity of the patient’s mental representations, adequate performance also depends on relatively intact manual motor execution. The first persuasive clinical evidence that avoided the necessity of manual response arose from the pioneering study by Bisiach and Luzzatti in 1978: They found that right brain damaged patients with visual neglect
tended to omit details from the left side of their memorial recollections of familiar scenes. When asked to describe from memory these known views from a given vantage point, patients would fail to include what was on the left from that viewpoint. If the viewpoint was reversed, items on the left from this new vantage point were now neglected while details from the previously neglected “left side,” were now reported. This finding of “representational neglect” has been confirmed many times. Although lateral cueing in these imagery tasks (Bisiach et al., 1981) could improve or worsen performance, the data were interpreted as providing the first unequivocal evidence that internal analogue space representation had been compromised in patients with visuospatial neglect. Subsequent reports of “pure” cases of representational neglect without visual neglect on other traditional tasks (Guariglia et al., 1993; Beschin et al., 1997; Peru and Zapparoli, 1999) provide further support for the argument that representational and perceptual spatial domains are subserved by (at least partially) independent neural mechanisms. GRAPHIC TASK ANALYSIS Unlike more easily quantifiable tasks such as bisection and cancellation, how one performs drawing/copying tasks is not fully constrained by the stimulus presented. Graphic tasks typically require some attempt to produce relevant local parts (in sequence) that are appropriately related to the overall global pattern. Moreover, graphic tests permit online perceptual feedback which in turn can allow for cross-cueing and selfcorrection. When drawing from memory, the patient’s initial response comes before any stimulus is perceived. Unlike bisection, copying and cancellation, there is nothing on the page prior to the patient commencing the task. Finally, perceptual analysis of one’s ongoing drawing involves a direct comparison of the mental representation of what one wants to draw and what has been drawn at any one time. In some cases, omissions in drawing decrease dramatically when such perceptual feedback is reduced or removed (Halligan and Marshall, 1994a; Mesulam, 1985): The patients make more accurate drawings with their eyes closed than open! Impairments of graphic skill that might be attributed to neglect can therefore result from at least three distinct types of deficit: lateralized impairment to (1) manual motor performance (including hypometria); (2) memorial recall and/or representational image generation; and (3) online visual perceptual feedback from the drawing itself (attentional). Drawing tasks thus sample a wider set of potential cognitive impairments than do other traditional neglect tests and may accordingly help explain the reported differences in sensitivity and dissociations between drawing/copying and line bisection and
S94
HALLIGAN AND MARSHALL
FIG. 2. Drawings by a professional artist, Tom Greenshields, before (a) and after (b) the artist’s stroke.
S95
GRAPHIC NEGLECT
FIG. 3.
Drawing by Tom Greenshields showing “object-based neglect” for two objects (candlestick and face).
cancellation (Pizzamigia et al., 1992; Kinsella et al., 1993). Figures 4a and 4b show something of the range of neglect impairments revealed by drawing. The patient fails to complete local features on the left side [the left eye in (a) and the left upper torso, left eye and left contour of the left leg in (b)]. This pattern of performance is difficult to reconcile with a general impairment of attention or with hypometria since many leftsided features are included and the overall composition is relatively well preserved. Although the patient knows full well that bodies and faces have a left side (and can verbally express this fact), she repeatedly neglects features on the left when drawing or copying. Furthermore, when looking at her supposedly “finished” drawing, she fails to notice that she has not overtly completed the left side. The left side of her spatial (but not verbal) mental representation may have failed to be activated and hence is not expressed in her drawing. Such failure might result from a local fault in
the structure that implements internal space representation. Alternatively, the deficit could stem from failure to attend to parts of the analogical mental image and hence not transcribe those parts in the overt drawing. PERCEPTUAL AWARENESS Both the above accounts, however, beg the question as to why patients fail to notice what they have done (or rather not done). Current theoretical approaches to neglect cannot readily reconcile why many patients unimpaired on one aspect of a task are grossly impaired on another aspect involving the very same stimulus. To merely say that the patients have left neglect is of course a description and not an explanation of their problems. Although not always explicitly stated, the common assumption is that such patients are not “aware” of their neglect, but this is often not true: many patients with intractable chronic neglect have
S96
HALLIGAN AND MARSHALL
volves a major omission within left space and an equally severe distortion within right space cannot adequately be explained by any of the three main accounts of neglect as they currently stand. CONCLUSION
FIG. 4. (a and b) Drawings by a patient with a right hemisphere stroke who considered both sketches to be complete.
considerable conceptual and indeed experiential insight into their deficit and its consequences (see Halligan and Marshall, 1998). Furthermore, unlike cancellation and line bisection where feedback is stimulus specific, the patient in a graphic task has access to a large body of long-term semantic and propositional memories that should serve as reminders of what to include in a drawing. One account of this aberrant perceptual awareness stresses the notion of “pathological completion,” a positive symptom whereby patients mentally “fill-in” the missing information (presumably on the basis of congruence between their current perceptual experience and semantic knowledge). Halfdrawings are thus believed to be fully complete (Halligan and Marshall, 1994b). This apparent lack of awareness of the deficit (Halligan and Marshall, 1998), even when the incongruity of the patients’ asymmetric productions is repeatedly pointed out to them stands in urgent need of explanation. Pathological left side “completion” is thought to be implicated in some, but not all, forms of impaired clock drawing (see Fig. 5). The patient who placed all 12 numerals along the right hemi-circumference of the clock-face insisted (in free vision and without time limit) that his reproduction looked just fine even when compared with that of a normal clock face. Such behavior which clearly in-
Visual neglect is no longer considered a single monolithic disorder but rather a multiplicity of cognitive deficits that collectively result in a lateralized disturbance of behavioral responses in space. Over the past decade, the use of conventional tests has highlighted the fact that they can produce differential and at times fully dissociable results within and between patients with neglect. This is hardly surprising given that these tests were never explicitly designed to selectively sample what is now considered the range of different cognitive impairments involved in visual neglect. These task-specific dissociations occur despite the fact that most of the traditional clinical tests are much less than pure in eliciting the contribution of specific cognitive deficits. Impaired performance on these “impure” tasks has been described as “perceptual,” “representational,” “attentional,” “intentional,” or “motor” neglect on the basis of a superficial characterization of the eliciting conditions. Graphic tasks appear simple but nonetheless require the coordination of many different skills with distinct cerebral substrates: the patient’s final production often involves interactions between his or her selective deficits and the struggle to compensate for impairment with whatever relevant functions remain. In drawing from memory, there must be an internal representation of the visual spatial features of the object to be delineated, as well as a mechanism for retrieving this knowledge. Complex visual perceptual and visual motor processes are then needed to translate the mental representation into an appropriate motor program for a “realistic” drawing. Further reconsideration of simple graphic performance by patients with visual ne-
FIG. 5. Two qualitatively different drawings of a clock face by patients with right brain damage.
GRAPHIC NEGLECT
glect may revitalize the search for the interplay between different cognitive processes and additional previously neglected pathological features implicated in this complex disorder. ACKNOWLEDGMENT P.W.H. and J.C.M. are supported by the Medical Research Council (UK).
REFERENCES Aldefender, M., and Blashfield, R. 1984. Cluster Analysis. Sage Publications, Beverly Hills. Agrell, B. M., Dehlin, O. I., and Dahlgren, C. J. 1997. Neglect in elderly stroke patients: A comparison of five tests. Psychiat. Clin. Neurosci. 51: 295–300. Barbieri, C., and De Renzi, E. 1989. Patterns of neglect dissociation. Behav. Neurol. 2: 13–24. Beschin, N., Cocchini, G., Della Sala, S., and Logie, R. H. 1997. What the eyes perceive, the brain ignores: A case of pure unilateral representational neglect. Cortex 33: 3–26. Binder, J., Marshall, R., Lazar, R., Benjamin, J., and Mohr, J. P. 1992. Distinct syndromes of hemineglect. Arch. Neurol. 49: 1187– 1194. Bisiach, E., and Luzzatti, C. 1978. Unilateral neglect of representational space. Cortex 14: 129 –133. Bisiach, E., Capitani, E., Luzzatti, C., and Perani, D. 1981. Brain and conscious representation of outside reality. Neuropsychologia 19: 543–551. Bisiach, E., and Berti, A. 1986. Representational impairment as a factor in neglect. In Neurology (K. Poeck, H. J. Freund, and H. Ganshirt, Eds.). Springer-Verlag, Berlin, Heidelberg. Bisiach, E. 1993. Mental representation in unilateral neglect and related disorders. The Twentieth Bartlett Lecture. Q. J. Exp. Psychol. 46: 435– 461. Bisiach, E., Pizzamiglio, L., Nico, D., and Antonucci, G. 1996. Beyond unilateral neglect. Brain 119: 851– 857. Cantagallo, A., and Della Sala, S. 1998. Preserved insight in an artist with extrapersonal spatial neglect. Cortex 34: 163–189. Caramazza, A., and Hillis, A. E. 1990. Levels of representations, co-ordinate frames and unilateral neglect. Cogn. Neuropsychol. 7: 391– 445. Cubelli, R., Nichelli, P., Bonito, V., De Tanti, A., and Inzaghi, M. G. 1991. Different patterns of dissociation in unilateral spatial neglect. Brain Cogn. 15: 139 –159. De Renzi, E. 1982. Disorders of Space Exploration and Cognition. Wiley, New York. Grusser, O.-J., and Landis, T. 1991. Visual Agnosias and Other Disturbances of Visual Perception and Cognition. Macmillan, London. Guariglia, C., Padovani, A., Pantano, P., and Pizzamiglio, L. 1993. Unilateral neglect restricted to visual imagery. Nature 15: 235– 237.
S97
Halligan, P. W., Marshall, J. C., and Wade, D. T. 1989. Visuo-spatial neglect: Underlying factors and test sensitivity. Lancet 2: 908 – 911. Halligan, P. W., Cockburn, J., and Wilson, B. A. 1991. The behavioural assessment of visual neglect. Neuropsychol. Rehabil. 1: 5–32. Halligan, P. W., and Marshall, J. C. 1992. Left visuo-spatial neglect: A meaningless entity? Cortex 28: 525–535. Halligan, P. W., and Marshall, J. C. 1994a. Toward a principled explanation of unilateral neglect. Cogn. Neuropsychol. 11: 167– 206. Halligan, P. W., and Marshall, J. C. 1994b. Completion in visuospatial neglect: A case study. Cortex 30: 685– 694. Halligan, P. W., and Marshall, J. C. 1997. The art of visual neglect. Lancet 350: 139 –140. Halligan, P. W., and Marshall, J. C. 1998. Neglect of awareness. Conscious. Cogn. 7: 356 –380. Heilman, K. M., and Valenstein, E. 1979. Mechanisms underlying hemispatial neglect. Ann. Neurol. 5: 166 –170. Kinsbourne, M. 1974. Mechanisms of hemispheric interaction in man. In Hemispheric Disconnection and Cerebral Function (M. Kinsbourne and W. L. Smith, Eds.), pp. 260 –285. C. C. Thomas, Springfield, IL. Kinsella, G., Oliver, J., Ng, K., Packer, S., and Stark, R. 1993. Analysis of the syndrome of unilateral neglect. Cortex 29: 135–140. McGlinchey-Berroth, R., Bullis, D. P., Milberg, W. P., Verfaellie, M., Alexander, M., and D’Esposito, M. 1996. Assessment of neglect reveals dissociable behavioural but not neuroanatomical subtypes. J. Int. Neuropsychol. Soc. 2: 441– 451. Marshall, J. C., and Halligan, P. W. 1996. Hemispheric antagonism in visuo-spatial neglect: A case study. J. Int. Neuropsychol. Soc. 2: 412– 418. Marshall, J. C., and Halligan, P. W. 1995. Within- and between-task dissociations in visuo-spatial neglect: A case study. Cortex 31: 367–376. Mattingley, J. B., Bradshaw, J. L., and Phillips, J. G. 1992. Reappraising unilateral neglect. Austr. J. Psychol. 44: 163–169. Mazzucchi, A., Pesci, G., and Trento, D. 1994. Cervello E Pittura. Fratelli Palombi Editori, Rome. Mesulam, M. 1985. Principles of Behavioural Neurology. F. A. Davis, Philadelphia. Ogden, J. A. 1985. Anterior–posterior interhemispheric differences in the loci of lesions producing visual hemineglect. Brain Cogn. 4: 59 –75. Pizzamiglio, L., Bergego, C., Halligan, P., Homberg, V., Robertson, I., Weber, E., Wilson, B., Zoccolotti, P., and Deloche, G. 1992. Factors affecting the clinical measurement of visuo-spatial neglect. Behav. Neurol. 5: 233–240. Peru, A., and Zapparoli, P. 1999. A new case of representational neglect. Ital. J. Neurol. Sci. 20: 243–246 Watson, R. T., Valenstein E., and Heilman, K. 1978. Nonsensory neglect. Ann. Neurol. 3: 505–508.