Closing-in without severe drawing disorders: The “fatal” consequences of pathological attraction

c o r t e x 4 5 ( 2 0 0 9 ) 2 8 5 – 2 9 2

ava i lab le at www.sc ienced i rec t . com

journa l homepage : www. e lsev ier . com/ loca te / cor tex

Special issue: Research report

Closing-in without severe drawing disorders: The ‘‘fatal’’consequences of pathological attraction

Massimiliano Consona, Sara Salzanoa, Valentino Manzob,Dario Grossia,c and Luigi Trojanoa,d,*aNeuropsychology Laboratory, Department of Psychology, Second University of Naples, Caserta, ItalybDepartment of Neurology, A.O.R.N. Cardarelli, Naples, ItalycC.I.R.N., Inter-University Center for Research in Neurosciences, ItalydFondazione Maugeri, I.R.C.C.S., Telese, Italy

a r t i c l e i n f o

Article history:

Received 2 July 2007

Reviewed 24 July 2007

Revised 23 October 2007

Accepted 8 November 2007

Published online 11 July 2008

Keywords:

Closing-in

Drawing impairments

Frontal lobe

Executive functions

Corticobasal degeneration

* Corresponding author. Neuropsychology LCaserta, Italy.

E-mail address: [email protected] (L0010-9452/$ – see front matter ª 2008 Elsevidoi:10.1016/j.cortex.2007.11.013

a b s t r a c t

The closing-in phenomenon (CIP) is often observed in patients with severe drawing

disorders, but its cognitive bases are not well understood. We describe an experimental

investigation aimed to clarify the nature of closing-in and its relationships with drawing

disorders in a patient with corticobasal degeneration. In copying simple or complex stimuli

(Experiment 1), the patient showed adherent and near types of closing-in, not affected by

stimulus complexity, and produced distorted and often unrecognisable drawings. On the

contrary, in drawing to dictation (without any available model), patients’ performances

significantly improved with respect to copying (Experiment 2). These data were consistent

with the hypothesis that in some patients closing-in may develop from frontal-related

release of approach behaviour even in the absence of relevant visuoperceptual impair-

ments. By asking the patient to reproduce given spatial locations within circular frames

(Experiment 3), we could further demonstrate the sparing of visuospatial processing and

the frontal genesis of closing-in. These findings allowed us to speculate on the heteroge-

neous nature of closing-in.

ª 2008 Elsevier Srl. All rights reserved.

1. Introduction particularly in late stages of Alzheimer’s disease (Ober et al.,

In copying drawings, patients may show a tendency either to

draw in close proximity to the model, or to start from one or

more of the model’s elements, or to overlap the model, some-

times producing a scrawl. Such behaviours (often co-occuring)

can all be described with the term closing-in phenomenon (CIP)

and are usually reported in demented patients (Mayer-Gross,

1935; Ajuriaguerra et al., 1960; Gainotti, 1972; Kwak, 2004),

aboratory, Department o

. Trojano).er Srl. All rights reserved

1991; Rouleau et al., 1996). The presence of CIP has been

considered suggestive of a diagnosis of ‘‘primary degenerative

dementia’’ (Gainotti et al., 1992), and could enhance sensitivity

and specificity of diagnosis of Alzheimer’s versus subcortical

vascular dementia (Kwak, 2004). The CIP is quite rare in

patients with focal lesions (Gainotti, 1972; Grossi et al., 1996).

Some authors suggested that the CIP occurs when

patients unable to structure an empty space look for

f Psychology, Second University of Naples, Via Vivaldi 43, 81100

.

Table 1 – General neuropsychological assessment

Neuropsychological test Rawscores

Equivalentscores

Mini mental state examination 22/30 P

Visuospatial span (Spinnler

and Tognoni, 1987)

1 0

Word span (Spinnler and Tognoni, 1987) 4 3

15-Word learning test: immediate

recall (Caltagirone et al., 1995)

34/75 3

15-Word learning test: delayed

recall (Caltagirone et al., 1995)

7/15 4

Story recall test (Novelli et al., 1986) 14,5/28 3

Raven’s colored progressive

matrices (Caltagirone et al., 1995)

8/36 0

Visuoconstructional abilities

Copying of geometrical drawings

(Spinnler and Tognoni, 1987)

2/14 0

Rey complex figure test: immediate

copy (Caffarra et al., 2002)

9/36 0

Rey complex figure test: delayed

reproduction (Caffarra et al., 2002)

12/36 4

Limb apraxia (gesture imitation test)

Right arm 30/72 P

Left arm 34/72 P

Note. Equivalent scores are obtained by adjusting raw scores for age

and education and then comparing corrected scores with distri-

bution of Italian normative studies (see reference citations in the

table): an equivalent score 0 means below the normal range, 1

means within normal limits, 2–4 means normal range. Italian

norms for MMSE (Measso et al., 1993) and for the gesture imitation

test (De Renzi et al., 1980) do not provide equivalent scores, but only

cut-off scores for normal range; for these tests P means patholog-

ical performance, while N indicates normal performance.

c o r t e x 4 5 ( 2 0 0 9 ) 2 8 5 – 2 9 2286

a reference point to solve difficult constructional tasks (De

Renzi, 1959). In this case, patients might be forced to resort to

already existing frames, for example the border of the model

or the edge of the paper (Grossi and Trojano, 2001). In this

case, CIP could be considered as a compensative behaviour:

indeed, Lee et al. (2004) suggested that CIP occurs to

compensate for an impairment of spatial working memory.

According to Lee et al. (2004), this hypothesis was consistent

with findings in patients with Alzheimer’s disease, who

showed more frequent CIPs in copying complex rather than

simple figures; in this case, the effect of stimulus complexity

has been ascribed to higher demand on spatial working

memory, although in principle more complex stimuli also

require finer spatial perceptual and representational

competences.

According to an alternative view, the closing-in has been

considered as a primitive reflex in patients with diffuse

cognitive deterioration who are strongly attracted by the

model and are unable to detach from it (Ajuriaguerra et al.,

1960; Gainotti, 1972). In this perspective, the tendency to

close-in to visible targets could be explained within the model

of competitive tropisms introduced by Denny-Brown (1956,

1958; Denny-Brown and Chambers, 1958). Starting from the

observation of the ‘‘magnetic attraction’’ shown by patients

with frontal lesions towards salient stimuli, Denny-Brown

proposed that central nervous system lesions lead to an

imbalance between two main motor approaches, positive

exploratory and negative withdrawal ‘‘tropisms’’. Positive

tropisms, such as grasping behaviours, would be controlled by

the parietal cortex and released by frontal lobe lesions;

negative tropisms, consisting in pathological avoidance or

withdrawal behaviours, would instead be controlled by frontal

regions and released by a parietal lobe damage.

In line with this interpretation, a frontal pathological

approach behaviour could occur independently from drawing

impairments, but this possibility has not been reported yet. In

the present paper we describe a patient with a frontal–

subcortical dysfunction who showed closing-in in different

drawing tasks in the absence of severe visuoconstructional

impairments. We verified that the patient’s systematic

tendency to draw upon or in proximity of relevant visual

stimuli could be ascribed to a defective control of motor

intentional behaviours, consistent with recent observations

on the genesis of spatial transpositions in some neglect

patients (Grossi et al., 2004; Lepore et al., 2005). We could also

document that the CIP was not dependent from severe

visuospatial disturbances.

2. Case report

L.M. is a 75 year-old right-handed woman, who had worked as

a teacher and was affected by corticobasal degeneration. Her

MR scan revealed a moderate diffuse cortical atrophy,

whereas the Tc99 m SPECT showed a marked hypoperfusion

in the basal ganglia region.

The patient underwent a formal neuropsychological

assessment (Table 1) that showed pathological scores on

visuospatial and drawing tasks, and on gesture imitation,

while verbal short-term memory, verbal learning and verbal

comprehension were normal. Specific assessment of execu-

tive functions (Table 2) did not reveal defects of abstract

thinking, cognitive flexibility and verbal fluency, while the

patient failed on tasks assessing monitoring and inhibition of

automatic responses (Stroop test and inverse motor learning

test).

L.M. showed several episodes of utilization behaviour, that

is non-requested but coherent activities in presence of trig-

gering environmental stimuli (Brazzelli and Spinnler, 1998);

for instance, L.M. lifted up the telephone receiver on several

occasions even if she was asked not to do so. Moreover, she

presented several kinds of approach behaviour towards

irrelevant visual stimuli; for example, on Raven’s Colored

Progressive Matrices she was strongly attracted by the salient

blank space within the model, often pointing to it as the

correct choice (see Suzuki et al., 2003, for an analogous

observation).

When asked to copy geometrical figures, L.M. systemati-

cally started to draw from the model or progressively

approached to it and produced grossly distorted drawings.

Since the tendency to close-in to the models was consistent

across repeated examinations several days apart, we decided

to investigate its causal mechanisms and its relationships

with drawing disorders. The patient did not show CIP on

gesture imitation tasks; therefore, the special neuro-

psychological assessment was targeted on drawing only.

Table 2 – Assessment of executive functions

Neuropsychological test Rawscores

Equivalentscores

Categorical verbal fluency

(Spinnler and Tognoni, 1987)

10.75 1

Phonological verbal fluency

(Caltagirone et al., 1995)

23 1

Abstract verbal judgment test

(Spinnler and Tognoni, 1987)

36/60 1

Cognitive estimation test 17 N

Stroop test: color naming

(Barbarotto et al., 1998)

15 4

Stroop test: interfered color

naming (Barbarotto et al., 1998)

5 0

Attentional matrices (Spinnler

and Tognoni, 1987)

16/60 0

Trail making test: part A

(Giovagnoli et al., 1996)

536 0

Trail making test: part B

(Giovagnoli et al., 1996)

1025 0

Trail Making Test: part B–A

(Giovagnoli et al., 1996)

489 0

Inverse motor learning test: right hand

(Spinnler and Tognoni, 1987)

6/24 0

Inverse motor learning test: left hand

(Spinnler and Tognoni, 1987)

5/24 0

Note. Equivalent scores are obtained by adjusting raw scores for age

and education and then comparing corrected scores with distri-

bution of Italian normative studies (see reference citations in the

table): an equivalent score 0 means below the normal range, 1

means within normal limits, 2–4 means normal range. Italian

norms for cognitive estimation test (Della Sala et al., 2003) do not

provide equivalent scores, but only cut-off scores for normal range;

for this test P means pathological performance, while N indicates

normal performance.

c o r t e x 4 5 ( 2 0 0 9 ) 2 8 5 – 2 9 2 287

3. Special neuropsychological assessment

Our experimental investigation was aimed to assess whether

CIP was related to defective visuospatial abilities or could be

ascribed to a frontal pathological approach behaviour. To

this purpose we assessed the effect of stimulus complexity

on copying geometrical figures (Experiment 1), compared

copying and drawing from memory tasks (Experiment 2),

and investigated the effect of spatial layout on reproduction

of simple spatial information (Experiment 3). The experi-

mental study started approximately one week after the

general neuropsychological assessment and was conducted

during the following two weeks, with three sessions per

week.

3.1. Experiment 1: reproduction of stimuli with twolevels of complexity

As a first step we aimed to assess frequency and qualitative

aspects of the CIP by means of a copying task in which the

patient had to reproduce simple and complex stimuli. After

Lee et al. (2004), the manipulation of stimulus complexity

could provide some clues about the nature of CIP in our

patient.

3.1.1. Materials and methodsWe used geometrical figures and symbols with two levels of

complexity. The 10 ‘‘simple’’ stimuli consisted of basic two-

dimensional figures or simple alphanumeric symbols (i.e.,

square, circle, rectangle, triangle, ellipse, pentagon, schematic

star, ‘‘X’’, upside-down ‘‘U’’, cross); the 10 ‘‘complex’’ stimuli

consisted of two-dimensional rendering of three-dimensional

figures or composite two-dimensional figures (i.e., cube,

cylinder, triangular pyramid, square pyramid, tetrahedron,

eight-armed star, cone, coil, square prism, pentagonal prism).

In each trial, a single stimulus (about 5 cm long, and 3 cm

wide) was printed in the upper half of a vertically arranged A4

sheet of paper, and aligned along the patient’s midline axis.

L.M. was required to copy the stimulus on the lower half of the

same sheet, starting from a small black dot (1 mm diameter)

printed at about 5 cm from the lower edge of the paper. The

examiner completed his verbal instructions pointing to the

small black dot which indicated the starting position, and

actually placing the pen in its correspondence. These

instructions were provided for each trial. All stimuli were

presented twice, for a total of 40 trials; simple and complex

stimuli were presented in randomized order.

Two scores were calculated: accuracy in reproducing the

model and the number of CIPs. With respect to the accuracy,

a response was scored 2 points when it was a perfect repro-

duction of the model, 1 point when the copy was distorted but

still recognizable, and 0 points when the model was not

identifiable in the patient’s response.

As regards CIPs, we adapted Kwak’s (2004) classification

criteria and included in this class of errors both the responses

where the patient overlapped at least one part of the model

(adherent type), and those in which the patient ignored the

starting point and drew in the space between the starting

point and the model, without touching it (near type). The

number of the two CIPs was recorded.

3.1.2. Results and commentL.M. was significantly more accurate in reproducing simple

(mean¼ 1; S.D.¼ .6) than complex (mean¼ .3; S.D.¼ .5)

stimuli (Wilcoxon signed ranks test, Z¼�6.121, p¼ .001).

However, the frequency of CIPs (82.5%; 33/40) did not differ

between simple (16/20) and complex (17/20) figures (Fisher’s

exact test, p¼ .5). Adherent (simple stimuli: 9/16; complex

stimuli: 8/17) and near (simple: 7/16; complex: 9/17) closing-in

types occurred with similar frequencies in the two levels of

stimulus complexity (Fisher’s exact test, p¼ .1; see Fig. 1).

Neither overall frequency nor the type of closing-in were

thus affected by stimulus complexity, and this is not predicted

by the interpretation based on working memory defects (Lee

et al., 2004). On the contrary, the present findings seem more

in line with the hypothesis that the phenomenon can be

ascribed to a pathological approach behaviour towards the

model, independently from the kind of the target to be

reproduced. In the same vein, Kwon et al. (2002) interpreted

the lack of a significant effect of task complexity as a finding

supporting the frontal nature of CIP in gesture imitation tasks.

According to this hypothesis, the mechanisms underlying CIP

are based on the release of an automatic behaviour, rather

than representing an attempt at a compensation of visuo-

spatial deficits. In this case, a reduction of drawing disorders

Fig. 1 – Adherent (A and C) and near (B and D) closing-in

types in copying simple (upper row) and complex (lower

row) figures.

c o r t e x 4 5 ( 2 0 0 9 ) 2 8 5 – 2 9 2288

could be expected in conditions when the CIP cannot be eli-

cited, e.g., in drawing without a visible model. This prediction

was tested in Experiment 2.

Fig. 2 – Patient’s performance on copying (A) and drawing

to dictation (B). L.M. was more accurate in drawing to

dictation than in copying.

3.2. Experiment 2: drawing to dictation versus copying

To verify the hypothesis that the patient’s closing-in was

relatively independent from severe drawing disorders, we

compared the drawing of the same stimuli with a model to be

reproduced (copying condition) and without it (drawing to

dictation). Improved accuracy in drawing to dictation would

suggest that closing-in could contribute to generate drawing

errors, thus indirectly supporting the frontal versus the

compensative hypothesis of the CIP in our patient.

3.2.1. Materials and methodsWe asked the patient to draw the same 20 stimuli in two

different conditions, i.e., copying and drawing from memory.

Each stimulus consisted in 2 two-dimensional geometrical

figures arranged in a given pattern (e.g., a square on the left of

a circle, a triangle on the right of a rhombus, a rectangle below

a pentagon, and circle above a rectangle).

In the copying task, each stimulus was printed in the upper

half of an A4 sheet, as in Experiment 1, and had to be repro-

duced in the lower half of the same sheet, starting from

a small black dot (1 mm diameter). In drawing to dictation,

L.M. was presented with blank A4 sheets and required to draw

the pattern verbally described by the examiner (e.g., ‘‘draw

a square on the left of a circle’’, or ‘‘draw a circle above

a rectangle’’).

As in Experiment 1, before each trial of the copying task,

L.M. was instructed to start from the given position, both

verbally and by placing the pen above it.

The two tasks were presented in randomized order, and we

calculated accuracy and the number of closing-in as described

in Experiment 1; however, since each pattern consisted in two

elements, scoring for accuracy took into account the whole

response rather than reproduction of its single elements (i.e.,

if the patient’s response contained only one correctly drawn

element, it was scored 0, since the stimulus could not be

recognized by a naive observer).

3.2.2. Results and commentThe patient was significantly more accurate in drawing to

dictation (mean¼ 1.5; S.D.¼ .7) than in copying (mean¼ .4;

S.D.¼ .6; Wilcoxon Signed Ranks Test, Z¼�4.121, p¼ .001; see

Fig. 2). The closing-in occurred with a frequency of 75% (15/20)

in the copying task (adherent¼ 7/15; near¼ 8/15). In drawing

to dictation, where no model was available, on two occasions

the patient showed a peculiar tendency to overlap the second

geometrical figure to the first one produced by herself, thereby

producing closing-in (Fig. 3; see also Suzuki et al., 2003); both

these items were rated 0 points on the accuracy score.

The drastic improvement of patient’s drawing perfor-

mances upon dictation suggested that closing-in could inter-

fere with the patient’s drawing performance. Actually, in

drawing to dictation L.M. could accurately reproduce several

stimuli that had been heavily deformed in copying. Although

findings from Experiment 1 demonstrated some impairments

in reproducing complex stimuli, CIP seemed to contribute to

this patient’s drawing difficulties.

The present results, together with those showing that

stimulus complexity affected drawing accuracy but not the

frequency of CIP in Experiment 1, would be consistent with

Fig. 3 – In drawing to dictation, on two occasions (A: to

draw a square above a rectangle, B: to draw a circle on the

right of a rectangle) the patient showed the tendency to

overlap the second geometrical figure to the first one

produced by herself, thereby leading to a peculiar kind of

closing-in.

c o r t e x 4 5 ( 2 0 0 9 ) 2 8 5 – 2 9 2 289

the idea that the CIP was independent from visuospatial

processing in our patient. Therefore, available evidence did

not support the hypothesis that CIP represented a compensa-

tion for defective spatial processing, and favoured the inter-

pretation based on a pathological approach behaviour.

Table 3 – Number and direction of horizontal (positivevalues mean rightward errors) and radial (positive valuesmean upward errors) transpositions in the fourconditions of Experiment 3

Transpositions Target location

Above Below Left Right

Horizontal 0 0 �11 12

Radial 10 �10 0 0

3.3. Experiment 3: reproduction of spatial positionswithin a circular frame

From the previous experiments we collected data suggesting

that CIP was determined by a response bias towards relevant

visual stimuli, independently from visuospatial processing

defects. This hypothesis would be supported by a task capable

both to quantify the pathological approach behaviour and to

demonstrate at least a relative sparing of visuospatial coding

abilities. For this purpose we adopted Lepore et al.’s (2005) task

requiring to reproduce single spatial locations within blank

circular frames; in this task the position of the model with

respect to the drawing space is systematically manipulated.

By means of this task, Lepore et al. (2005) observed trans-

positions in multiple directions of space, but always towards

the model, in a patient with a frontal–subcortical lesion; this

pattern of response was considered an instance of CIP due to

a released approach behaviour. On these bases, our patient

was expected to show an analogous pattern of spatial trans-

positions, but with spared abilities to compute stimulus

locations within the model.

3.3.1. Materials and methodsEach stimulus was centred on an A4 sheet and consisted of

a circle (12.5-cm diameter) with a 1-cm bar printed in one of

the positions usually occupied by numbers on a clock face.

The patient was required to reproduce in a blank circle, cen-

tred on a separate A4 sheet, the position of the little bar pre-

sented in the model. The two circles were always placed about

8 cm apart, but their disposition varied systematically: in two

radial conditions, the model was located above or below the

response sheet, and in two horizontal conditions, the target

was located on the left or the right of the response sheet.

For each condition, 12 stimuli (each corresponding to the

position of one of the 12 h of a clock face) were given three

times, in different testing sessions 3 days apart (36 stimuli per

condition); the order of the conditions was balanced across

sessions. Task instructions did not make any reference to the

clock face.

Number and direction of spatial transpositions (considered

as responses located in the half of the dial opposite the actual

position) were calculated for each condition.

3.3.2. Results and commentThe absolute number of errors was similar across the four

experimental conditions, but L.M. made only radial trans-

positions in the two radial conditions and only horizontal

transpositions in the horizontal conditions (Table 3 and Fig. 4),

always towards the model. Nonetheless, the patient’s

responses were not all grouped in close proximity to the target

but maintained some regular spatial distribution within the

circular frame. Indeed, L.M. correctly reproduced stimuli

located in the half close to the model and transposed only the

stimuli presented in the opposite half, systematically

respecting their relative location within the half-frame (see

Appendix for a geometrical model describing the systematic

relationships between target locations and patient’s actual

responses). These findings would thus confirm that L.M. was

able to represent the spatial location of stimuli within the

circular frame of reference, whereas spatial transpositions

mainly derived from the pathological approach behaviour

towards the model, consistent with expectations and with

Lepore et al.’s findings in a patient with a frontal lesion

(2005).

4. Discussion

Classical studies on demented patients have demonstrated

that frequency and severity of closing-in increase with disease

progression (Mayer-Gross, 1935), thus suggesting that severity

of CIP would parallel progressive spatial disorders (see also De

Renzi, 1959). More recently, Lee et al. (2004) pointed out that

higher frequency of CIP for more complex stimuli could be

ascribed either to visuospatial or to working memory deficits,

consistent with the idea that the CIP may represent

a compensative attempt at coping with visuospatial disorders.

In the present paper we described a patient showing

frequent CIP in drawing tasks. We collected evidence that

stimulus complexity decreased L.M.’s drawing accuracy but

did not affect the frequency and the type of closing-in

(Experiment 1), thus suggesting that CIP was at least relatively

independent from the presence of drawing disorders. This

claim was supported by the remaining findings of the special

neuropsychological assessment: in drawing to dictation

Fig. 4 – Graphical representation of patient’s performance in the four conditions of Experiment 3. All responses were

digitalised and superimposed on a template of the response sheet, separately for each condition. The two radial conditions

are represented at the top, where the model was positioned above (A) or below (B) the response sheet, and the two

horizontal conditions at the bottom, where the model was positioned on the left (C) or the right (D) of the response sheet.

The locations of L.M.’s responses were identified by numbers that referred to the original stimulus locations (represented as

in analogic clock hours).

c o r t e x 4 5 ( 2 0 0 9 ) 2 8 5 – 2 9 2290

(Experiment 2), where no relevant visual stimuli were avail-

able to sight, L.M.’s drawing performances improved signifi-

cantly; therefore, the complex cognitive operations

subserving drawing tasks seemed to be interfered with by the

occurrence of CIP. Last, results of Experiment 3 demonstrated

that the tendency to draw close to the model systematically

distorted the output of relatively spared spatial coding

mechanisms (see Fig. 4).

These findings did not fit with an interpretation of closing-

in based on a defect of spatial abilities, but were compatible

with the hypothesis of a released approach behaviour. Start-

ing from the approach-avoidance model of Denny-Brown

(1956, 1958; Denny-Brown and Chambers, 1958), it is possible

to ascribe the defective suppression of automatic motor

responses towards available environmental stimuli to frontal

dysfunction. Indeed, in L.M. the CIP was associated with

utilization behaviour during the clinical examination and with

several kinds of approaching behaviours in tasks where she

had to pay attention and act upon visual stimuli, as reported

in patients with medial frontal lesions (Archibald et al., 2001;

Boccardi et al., 2002). Consistent with this, L.M. failed in

executive tasks known to tap cognitive processes subserved

by medial frontal structures, such as the Stroop task and

a reversed-learning motor task (Stuss and Levine, 2002).

This interpretation is also consistent with findings from

recent neuropsychological studies in neglect patients

demonstrating that the release of pathological approach

behaviour may represent a component of the complex

mechanisms underlying productive drawing phenomena,

such as perseverations on line cancellation tasks or spatial

transpositions on constructional tasks (Grossi et al., 2004;

Lepore et al., 2005; Rusconi et al., 2002; Vallar et al., 2006; Pia

et al., 2009, this issue). In particular, Grossi et al. (2004)

described a neglect patient with a right frontal lesion in whom

the release of a pathological approach behaviour could

account for the systematic tendency to transpose drawings

towards the side of the page where he started to draw. An

analogous interpretation has been proposed to explain spatial

transpositions towards different directions of space in

another neglect patient with a right frontal lesion (Lepore

et al., 2005). Therefore, the present findings would extend

such a theoretical framework to comprehension of drawing

phenomena in non-neglect patients, and would demonstrate

the extent to which frontal dysfunctions may impair visuo-

spatial performances.

Taken together, these findings suggest that the patient

could build up sufficiently accurate spatial representation of

the figure to be produced, and that the lack of frontal inhibi-

tory mechanisms determined a capture of the patient’s

attention and action towards the model, with production of

distorted spatial relationships in copying tasks. In such tasks,

subjects have to construct and maintain a mental

c o r t e x 4 5 ( 2 0 0 9 ) 2 8 5 – 2 9 2 291

representation of the model in short-term memory before

starting reproduction, and then they have to continuously

check their own reproduction to verify their accuracy (Pisella

and Mattingley, 2004; but see Tchalenko and Miall, 2009, this

issue, for a different point of view). A frontal impairment

might interfere with such a process that could provide stable

representations of visual stimuli across shifts of spatial

attention, and update their spatial locations across eye

movements.

In conclusion, we have described a patient with a frontal–

subcortical dysfunction which showed closing-in of likely

frontal origin. The release of automatic approach behaviours

seemed to determine the patient’s tendency to act in direction

of relevant visual stimuli giving rise to CIP in copying tasks, in

spite of relatively spared spatial processing abilities. This

observation would support the original interpretation

proposed by Gainotti (1972), according to whom the CIP might

arise from a failure to inhibit attraction of actions towards the

focus of visual attention in demented patients, as well as in

children (see also Ambron et al., 2009, this issue, for an

experimental study on normal children). However, the

present case study on a patient affected by a degenerative

disease cannot allow strong neuroanatomical inferences in

support of the functional relation between anterior and

posterior areas suggested by Denny-Brown’s model. To

support this model a systematic investigation on homoge-

neous patient groups with either frontal or parietal focal

lesions is warranted. Notwithstanding this limitation, since

L.M. presented a range of other utilization behaviours, it could

be possible to speculate that the closing-in is one of the

automatic response schemata released by the frontal lesion.

In this theoretical framework, pathological approach behav-

iours, such as utilization behaviour (Lhermitte et al., 1986) or

more complex imitative behaviours (Conchiglia et al., 2007),

might derive from the lack of inhibition on the action simu-

lation system (see Baldissera et al., 2001). By analogy, it could

be hypothesized that in the context of drawing, the frontal

dysfunction would make evident the covert process of

sensory–motor representation that precedes the development

of complex mental representation of space in children

(Gainotti, 1972). This interpretative hypothesis would fit the

construct of the embodied simulation theory (Gallese and

Lakoff, 2005), according to which abstract concepts derive

from the direct sensory–motor interaction with the external

world. Further investigations would be necessary to disen-

tangle action simulation release from visual grasping

hypotheses in patients with ‘‘frontal’’ closing-in.

Acknowledgement

We are grateful to A. Cammarota, University of Salerno (Italy),

for computation of geometrical relations.

Appendix

To verify that the patient accurately processed spatial loca-

tions, and that the actual response systematically incorpo-

rated this information, although transposed in the half-frame

closest to the model, we formulated a geometrical model of

this behaviour and correlated it with patient’s actual

responses. The model could be expressed by the following

geometrical relations:�a ¼ q for 90� þ f � q < 270� þ 4

a ¼ 2ð90þ fÞ � q for � 90� þ f � q < 90� þ f

where a is the angular coordinate of the response expressed in

degrees, q is the angular coordinate of the bar within the

stimulus, and 4 is the angular coordinate of the stimulus

location with respect to the response sheet (0� means that the

response sheet is above the stimulus, 90� means that the

response sheet is on the right of the stimulus, and so on).

Responses’ angular coordinates computed by these relations

were significantly correlated with observed responses (Pear-

son’s r¼ .754, p< .001).

r e f e r e n c e s

Ajuriaguerra J, Muller M, and Tissot R. Apropos of some problemsposed by apraxia in dementia. Revue Neurologique, 102: 640–642, 1960.

Ambron E, Della Sala S, and McIntosh RD. Animal magnetism:evidence for an attraction account of closing-in behaviour inpre-school children. Cortex, 45: 278–284, 2009.

Archibald SJ, Mateer C, and Kerns KA. Utilization behaviour:clinical manifestations and neurological mechanisms.Neuropsychology Review, 11: 117–130, 2001.

Baldissera F, Cavallari P, Craighero L, and Fadiga L. Modulation ofspinal excitability during observation of hand actions inhumans. European Journal of Neuroscience, 13: 190–194, 2001.

Barbarotto R, Laiacona M, Frosio R, Vecchio M, Farinato A, andCapitani E. A normative study on visual reaction times andtwo Stroop colour-word tests. Italian Journal of NeurologicalSciences, 19: 161–170, 1998.

Boccardi E, Della Sala S, Motto C, and Spinnler H. Utilisationbehaviour consequent to bilateral SMA softening. Cortex, 38:289–308, 2002.

Brazzelli M and Spinnler H. An example of lack of frontalinhibition: the ‘utilization behaviour’. European Journal ofNeurology, 5: 347–353, 1998.

Caffarra P, Vezzadini G, Dieci F, Zonato F, and Venneri A.Rey–Osterrieth complex figure: normative values in anItalian population sample. Neurological Sciences, 22: 443–447,2002.

Caltagirone C, Gainotti G, Carlesimo GA, and Parnetti L. Batteriaper la valutazione del deterioramento mentale (parte I):descrizione di uno strumento per la diagnosineuropsicologica. Archivio di Psicologia, Neurologia e Psichiatria,55: 461–470, 1995.

Conchiglia G, Della Rocca G, and Grossi D. On a peculiarenvironmental dependency syndrome in a case with frontal–temporal damage: Zelig-like syndrome. Neurocase, 13: 1–5,2007.

De Renzi E. Osservazioni semeiogenetiche in tema di aprassiacostruttiva. Rivista Sperimentale di Freniatria, 58: 231–256,1959.

De Renzi E, Motti F, and Nichelli P. Imitating gestures:a quantitative approach to ideomotor apraxia. Archives ofNeurology, 37: 6–10, 1980.

Della Sala S, Macpherson SE, Phillips LH, Sacco L, and Spinnler H.How many camels are there in Italy? Cognitive estimatesstandardised on the Italian population. Neurological Sciences,24: 10–15, 2003.

c o r t e x 4 5 ( 2 0 0 9 ) 2 8 5 – 2 9 2292

Denny-Brown D. Positive and negative aspects of cerebral corticalfunction. North Carolina Medical Journal, 17: 295–303, 1956.

Denny-Brown D. The nature of apraxia. Journal of Nervous andMental Disease, 126: 9–32, 1958.

Denny-Brown D and Chambers RA. The parietal lobe andbehavior. Association for Research in Nervous and Mental Disease,36: 35–117, 1958.

Gallese V and Lakoff G. The brain’s concepts: the role of thesensory–motor system in reason and language. CognitiveNeuropsychology, 22: 455–479, 2005.

Gainotti G. A quantitative study of the ‘‘closing-in’’ symptom innormal children and in brain-damaged patients.Neuropsychologia, 10: 429–436, 1972.

Gainotti G, Parlato V, Monteleone D, and Carlomagno S.Neuropsychological markers of dementia on visuospatialtasks: a comparison between Alzheimer’s type and vascularforms of dementia. Journal of Clinical and ExperimentalNeuropsychology, 14: 239–252, 1992.

Giovagnoli R, Del Pesce M, Mascheroni S, Simoncelli M,Capitani E, and Laiacona M. Italian norms for trail making test.Italian Journal of Neurological Sciences, 17: 305–309, 1996.

Grossi D and Trojano L. In Behrmann M (Ed), Constructional andvisuospatial disorders. Handbook of neuropsychology, vol. 4.Amsterdam: Elsevier, 2001. Sec.

Grossi D, Correra G, Calise C, and Trojano L. Selectiveconstructional disorders after right subcortical stroke. Aneuropsychological premorbid and follow-up study. ItalianJournal of Neurological Sciences, 14: 23–33, 1996.

Grossi D, Di Cesare G, and Trojano L. Left on the right orviceversa: a case of ‘‘alternating’’ constructional allochiria.Cortex, 40: 511–518, 2004.

Kwak YT. ‘‘Closing-in’’ phenomenon in Alzheimer’s disease andsubcortical vascular dementia. BMC Neurology, 4: 3, 2004.

Kwon JC, Kang SJ, Lee BH, Chin J, Heilman KM, and Na DL. Manualapproach during hand gesture imitation. Archives of Neurology,59: 1468–1475, 2002.

Lee BH, Chin J, Kang SJ, Kim EJ, Park KC, and Na DL.Mechanism of the closing-in phenomenon in a figurecopying task in Alzheimer’s disease patients. Neurocase, 10:393–397, 2004.

Lepore M, Conson M, Grossi D, and Trojano L. Multidirectionaltranspositions suggesting pathologic approach behaviourafter frontal stroke. Neurology, 64: 1615–1617, 2005.

Lhermitte F, Pillon B, and Serdaru M. Human autonomy and thefrontal lobes. Part I: imitation and utilization behaviour:

a neuropsychological study of 75 patients. Annals of Neurology,19: 326–334, 1986.

Mayer-Gross W. Some observations on apraxia. Proceedings of theRoyal Society of Medicine, 28: 1203–1212, 1935.

Measso G, Cavarzeran F, Zappala G, Lebowitz BD, Crook TH,Pirozzolo FJ, et al. The mini-mental state examination:normative study of an Italian random sample. DevelopmentalNeuropsychology, 9: 77–85, 1993.

Novelli G, Papagno C, Capitani E, Laiacona M, and Cappa SF. Tretest clinici di memoria verbale a lungo termine: Taratura susoggetti normali. Archivio di Psicologia, Neurologia e Psichiatria,47: 278–296, 1986.

Ober BA, Jagust WJ, Koss E, Delis DC, and Friedland RP.Visuoconstructive performance and regional cerebral glucosemetabolism in Alzheimer’s disease. Journal of Clinical andExperimental Neuropsychology, 13: 752–772, 1991.

Pia L, Folegatti A, Guagliardo M, Genero R, and Gindri P. Aredrawing perseverations part of the neglect syndrome? Cortex,45: 293–299, 2009.

Pisella L and Mattingley JB. The contribution of spatial remappingimpairments to unilateral visual neglect. Neuroscience andBiobehavioural Reviews, 28: 181–200, 2004.

Rouleau I, Salmon DP, and Butters N. Longitudinal analysis ofclock drawing in Alzheimer’s disease patients. Brain andCognition, 31: 17–34, 1996.

Rusconi ML, Maravita A, Bottini G, and Vallar G. Is the intact sidereally intact? Perseverative responses in patients withunilateral neglect: a productive manifestation.Neuropsychologia, 40: 594–604, 2002.

Spinnler H and Tognoni G. Standardizzazione e taratura italianadi tests neuropsicologici. Italian Journal of Neurological Sciences,6: 8–96, 1987.

Stuss DT and Levine B. Adult clinical neuropsychology: lessonsfrom studies of the frontal lobes. Annual Reviews of Psychology,53: 401–433, 2002.

Suzuki K, Otsuka Y, Endo K, Ejima A, Saito H, Fujii T, et al.Visuospatial deficits due to impaired visual attention:investigation of two cases of slowly progressive visuospatialimpairment. Cortex, 39: 327–341, 2003.

Tchalenko J and Miall C. Eye–hand strategies in copying complexlines. Cortex, 45: 368–376, 2009.

Vallar G, Zilli T, Gandola M, and Bottini G. Productive anddefective impairments in the neglect syndrome: graphicperseveration, drawing productions and optic prism exposure.Cortex, 42: 911–920, 2006.