RESEARCH ARTICLE

Line and word bisection in right-brain-damaged patients with leftspatial neglect

Laura Veronelli • Giuseppe Vallar •

Chiara V. Marinelli • Silvia Primativo •

Lisa S. Arduino

Received: 6 February 2013 / Accepted: 27 September 2013 / Published online: 22 October 2013

� Springer-Verlag Berlin Heidelberg 2013

Abstract Right-brain-damaged patients with left unilate-

ral spatial neglect typically set the mid-point of horizontal

lines to the right of the objective center. By contrast, healthy

participants exhibit a reversed bias (pseudoneglect). The

same effect has been described also when bisecting ortho-

graphic strings. In particular, for this latter kind of stimulus,

some recent studies have shown that visuo-perceptual

characteristics, like stimulus length, may contribute to

both the magnitude and the direction bias of the bisec-

tion performance (Arduino et al. in Neuropsychologia

48:2140–2146, 2010). Furthermore, word stress was shown

to modulate reading performances in both healthy

participants, and patients with left spatial neglect and neglect

dyslexia (Cubelli and Beschin in Brain Lang 95:319–326,

2005; Rusconi et al. in Neuropsychology 18:135–140, 2004).

In Experiment I, 22 right-brain-damaged patients (11 with

left visuo-spatial neglect) and 11 matched neurologically

unimpaired control participants were asked to set the sub-

jective mid-point of word letter strings, and of lines of

comparable length. Most patients exhibited an overall dis-

proportionate rightward bias, sensitive to stimulus length,

and similar for words and lines. Importantly, in individual

patients, biases differed according to stimulus type (words

vs. lines), indicating that at least partly different mechanisms

may be involved. In Experiment II, the putative effects on the

bisection bias of ortho-phonological information (i.e., word

stress endings), arising from the non-neglected right hand

side of the stimulus were investigated. The orthographic cue

induced a rightward shift of the perceived mid-point in both

patients and controls, with short words stressed on the

antepenultimate final sequence inducing a smaller rightward

deviation with respect to short words stressed on the penul-

timate final sequence. In conclusion, partly different mech-

anisms, including both visuo-spatial and lexical factors, may

support line and word bisection performance of right-brain-

damaged patients with left spatial neglect, and healthy

participants.

Keywords Unilateral spatial neglect � Word

bisection � Line bisection � Word stress � Lexical

variables � Neglect dyslexia

Introduction

When setting the mid-point of horizontal lines, right-brain-

damaged patients with left unilateral spatial neglect

L. Veronelli � G. Vallar

Department of Psychology, Universita di Milano-Bicocca,

Milan, Italy

L. Veronelli (&)

Department of Neurorehabilitation, Casa Cura Policlinico,

Via Dezza 48, 20144 Milan, Italy

e-mail: laura.veronelli@gmail.com

G. Vallar

Neuropsychological Laboratory, IRCCS Istituto Auxologico

Italiano, Milan, Italy

C. V. Marinelli � S. Primativo

Department of Psychology, La Sapienza Universita di Roma,

Rome, Italy

C. V. Marinelli � S. Primativo

Neuropsychological Research Centre, IRCCS Foundation

Hospital Santa Lucia, Rome, Italy

L. S. Arduino

Universita LUMSA, Rome, Italy

L. S. Arduino

ISTC-CNR, Rome, Italy

123

Exp Brain Res (2014) 232:133–146

DOI 10.1007/s00221-013-3726-7

typically transect the line to the right of the objective mid-

point, ipsilateral to the side of the hemispheric lesion

(ipsilesional, see e.g., Bisiach et al. 1976, 1983). The

rightward bisection bias is a main manifestation of the

syndrome of unilateral spatial neglect (Vallar 1998;

Bisiach and Vallar 2000; Vallar and Bolognini 2013),

reflecting a defective perceptual representation of the lat-

eral extent of the line, which is interpreted as truncated, or

disproportionately compressed in the side contralateral to

the hemispheric lesion (contralesional), namely the left

hand side in right-brain-damaged patients (Bisiach et al.

1983; Bisiach and Vallar 2000). On the contrary, healthy

participants show a leftward bias (pseudoneglect), which

may reflect a leftward orientation, supported by a relative

prevalence of a leftward attentional bias by the right

hemisphere, and modulated by a number of factors,

including age and scanning direction (Bowers and Heilman

1980; Jewell and McCourt 2000, for review).

In recent years, the bisection task has been considered a

useful paradigm for investigating spatial representations,

not only of lines (e.g., Bisiach et al. 1976; Vallar et al.

2000; review in Jewell and McCourt 2000), but also of

written words (i.e., meaningful letter strings). Neurologi-

cally unimpaired participants show a leftward bias (i.e.,

pseudoneglect), when bisecting orthographic material,

including words, pseudowords (i.e., legal, pronounceable,

non-words), letter, and symbol strings (Fischer 1996,

2000). Automatic lexical access may involve an attentional

focusing on the word beginning (Fischer 1996, 2000,

2004), in order to establish a cohort of potential entries in

the mental lexicon (Paap et al. 1982). Consequently, par-

ticipants who read from left to right, may over-attend/

represent the lateral extent of the initial part of the word,

with a systematic leftward bias, when bisecting the letter

string (Attentional Scaling Hypothesis; Fischer 1996, 2000,

2004). This spatial distortion could be even partially

responsible for the leftward bias in oculomotor behavior

during reading tasks: The optimal viewing position may be

located just to the left of the middle of the word, reducing

gaze duration, lexical decision, and word reading time

(O’Regan et al. 1984; O’Regan and Jacobs 1992).

Stimulus length influences the bisection performance of

Italian healthy participants in a different fashion for lines

and orthographic strings (Arduino et al. 2010): Leftward

biases for short and long lines, and a more varied pattern

for orthographic strings (mainly, leftward biases for long

words, pseudowords, consonant, and symbol strings;

rightward biases for short such stimuli). Word and line

bisection may then undergo different types of processing

not only linguistic, but also visuo-perceptual, based on the

different structural and perceptual features of the two types

of stimuli, namely discrete (letter strings), for words, ver-

sus continuous, for lines. In one study (Lee et al. 2004; see

also Na et al. 2000, for a detailed description of the stim-

uli), Korean right-brain-damaged patients with left spatial

neglect and neurologically unimpaired participants circle

the target closest to the mid-point of long linear arrays (24

targets and 21 distracter in a 242-mm string) of character

and symbol strings. All participants exhibit a rightward

deviation (greater in right-brain-damaged neglect patients)

with both character and symbol strings, as compared with

solid lines. In another study (Mohr and Leonards 2007),

native English healthy participants show greater rightward

biases, when bisecting long letter strings, in which emo-

tional (as compared to neutral) words are inserted on the

right or on the left of the veridical center. This rightward

deviation may be explained by a left-hemispheric activa-

tion, related to greater demands on left-hemisphere-sup-

ported attentional resources, devoted to local processing

(Martin 1979). In the case of real words, inserted in letter

strings, lexical-semantic processing may also contribute to

a left-hemispheric activation (see Bowers and Heilman

1980).

Due to the few available neuropsychological studies on

word bisection, we first aimed at assessing, in Italian right-

brain-damaged patients with left spatial neglect, whether

stimulus length modulates the error in a similar way as that

reported for line bisection, namely a greater rightward

deviation with longer lines (see, e.g., Bisiach et al. 1983;

Vallar et al. 2000; Daini et al. 2002, for such an evidence in

right-brain-damaged patients from line bisection perfor-

mance). Accordingly, we predicted a greater rightward bias

for long words than for short ones. Secondly, with word

and line bisection involving partially independent mecha-

nisms (see Arduino et al. 2010), double dissociations

(Vallar 2000) may be predicted in the patients’ perfor-

mances with respect to the line and word stimuli. Specifi-

cally, unilateral spatial neglect may impact on the bisection

of words and lines in a different fashion in different

patients, supporting the existence of independent mecha-

nisms, which could be compromised selectively by brain

damage.

Secondly, we aimed at investigating the effects of the

lexical status of the stimulus on the magnitude of the bias.

Lexical variables modulate the reading performance of

brain-damaged patients with unilateral spatial neglect

(review in Vallar et al. 2010). As for word bisection, no

effects of word written frequency have been found so far

(Arduino et al. 2010; Fischer 1996).

In Italian, word stress is a good candidate to exploit the

influence of lexical factors in the patients’ reading perfor-

mance, since the main source of information for setting

word stress position is to look at their final (right-sided)

part (Burani and Arduino 2004). Right-brain-damaged

patients with left neglect dyslexia are able to read aloud

letters in the right hand side of a word. Results from both

134 Exp Brain Res (2014) 232:133–146

123

healthy participants, and patients with left spatial neglect

and neglect dyslexia, show that word stress modulates

reading (Cubelli and Beschin 2005, the presence of an

accent mark; Rusconi et al. 2004, preserved stress position

in the response). Italian words with three or more syllables

have two main stress patterns: They can be stressed on the

penultimate (last but one) syllable (e.g., matıta, pencil), or

on the antepenultimate (last but two) syllable (e.g., bıbita,

drink), with the former being the more frequent stress

pattern (dominant or regular), the latter the less frequent

one (non-dominant or irregular). Different word final

sequences show different proportions of words with dom-

inant or non-dominant stress. For instance, most words

ending in -oro are regularly stressed on the penultimate

syllable (about 81 %), with only 19 % of words taking the

irregular stress. The final word sequence could then be used

by Italian readers as an informational cue for attributing the

correct word stress pattern (Burani and Arduino 2004;

Sulpizio et al. 2013). This, in turn, may modulate the

bisection performance of right-brain-damaged patients with

left spatial neglect. Based on the above-mentioned evi-

dence suggesting that stress information modulates reading

performance, the rightward bisection bias for words stres-

sed on the antepenultimate syllable, as compared with

words stressed on the penultimate syllable may be reduced,

both in brain-damaged neglect patients, as well as in

healthy participants.

Experiment I

Materials and methods

Participants

Participants were recruited from the inpatient population of

the Fondazione S. Lucia IRCCS, Rome (Research Center

on Neuropsychology), and from the Casa Cura Policlinico,

Milan (Department of Neurorehabilitation), Italy. A total of

11 right-hemisphere-damaged patients with left unilateral

spatial neglect (N?), 11 right-hemisphere-damaged

patients without left spatial neglect (N-), and 11 unim-

paired control participants (C) took part in the study. The

N? patients had suffered a stroke (10 ischemic, 1 ischemic

with hemorrhagic infarction). The sample included 4

females and 7 males with a mean age of 67.90 years

(SD = ±9.25, range 48–79), and a mean education of

9.81 years (SD = ±5.19; range 5–18). Mean duration of

disease of the 11 patients was 2 months (SD = ±1.46,

range 1–6). In the N- group, 9 patients had suffered an

ischemic stroke and 2 an ischemic stroke with hemorrhagic

infarction. The sample included 4 females and 7 males

(mean age = 69.82 years, SD = ±9.97, range 44–79;

mean education = 11.27 years, SD = ±4.90, range 2–18).

The mean duration of the disease was 1.41 months

(SD = ±.70, range 1–3). C participants were matched for

sex (6 females), age (mean = 68.73 years, SD = ±7.56,

range 57–80), and education (mean = 10.73 years,

SD = ±3.90, range 5–17) with the patients. One-way

analyses of variance (ANOVAs) showed that age

(F2,30 = .12; p = .88; pg2 = .01) and educational level

(F2,30 = .27; p = .77; pg2 = .02) did not differ among

groups (N?, N-, and C). Furthermore, duration of the

disease did not differ between N? and N- patients

(t20 = 1.21; p = .24). Lesion site was assessed for each

right-brain-damaged patient by CT or MRI scan and drawn

manually using the MRIcro software (Rorden and Brett

2000) onto selected horizontal slices of a standard template

brain. Figure 1 shows the overlapped lesion maps of 21 out

of 22 right-brain-damaged patients (N? and N- groups).

Scan images were unavailable for N- patient RL, with

neuroradiological medical records reporting an ischemic

lesion involving the right fronto-temporal region and the

insula. In N? patients lesions superimposed in the right

putamen and in the white matter underneath the insula, the

rolandic operculum and the frontal-inferior operculum (11

patients); in N- patients a maximum overlap was observed

in the white matter underneath the insula (5 patients).

Average lesion volume (SD) was greater in N? patients

(105.16 cm3 ± 73.61) than in N- patients (17.31 ± 15.37;

t11 = 3.87, p = .003), in line with previous findings

(Cattaneo et al. 2012; Leibovitch et al. 1998).

Patients were right-handed on a standard questionnaire

(Oldfield 1971), had normal or corrected-to-normal

vision, and no history of previous neurological and

psychiatric disorders. All patients showed preserved

visual fields on Goldmann perimetry or on the confron-

tation test, but N? patient MA, who exhibited left

hemianopia. All patients were given a Mini Mental State

Examination (MMSE, Folstein et al. 1975) and their

scores (N?: mean 26.53, SD ±1.28, range 25–29; N-:

26.76 ± 1.91, range 25–30) were above the adjusted

cutoff of Magni et al. (1996). Demographic and neuro-

logical information for N? and N- patients are sum-

marized in Table 1. Informed consent was obtained from

all participants (Declaration of Helsinki, British Medical

Journal, 302: 1194, 1991).

Baseline neuropsychological assessment

The presence and severity of left spatial neglect were

assessed by a diagnostic battery, which included: line

(Albert 1973), letter (Diller and Weinberg 1977; Vallar

et al. 1994), and star (Wilson et al. 1987; Ronchi et al.

2009) cancellation; line bisection (Fortis et al. 2010);

complex figure drawing (Gainotti et al. 1972; Ronchi et al.

Exp Brain Res (2014) 232:133–146 135

123

2009); clock drawing from memory (Mondini et al. 2003);

sentence reading (Zoccolotti et al. 1989). Finally, a single

word reading test was given. Stimuli included two lists of

38 words and 38 orthographically pseudowords (Vallar

et al. 1996). Stimuli were presented separately in a random-

fixed order, with a time-limited computerized procedure.

Participants sat about 50 cm away from a 15.400 PC screen

in a quiet room, with the center of the screen being aligned

with the mid-sagittal plane of the participant’s body.

E-Prime v 2.0 software was used to display the stimuli.

Each trial began with a 300-ms fixation point (a cross,

black, 30-pt, Arial font) followed by the appearance of a

single word (black letters uppercase, 37-pt, Arial font) in

the center of the screen, with a 500-ms exposure. The

participant’s task was to read aloud each letter string, with

answers being manually registered by the experimenter. If

a participant failed to read correctly all of the first 5 stimuli

or 8 out of the first 10 stimuli, time exposure was increased

to 750 ms. In case of failure with this time exposure of the

stimulus (scored with the same criteria), a 1,000-ms

exposure was used. No feedback was given as for the

participant’s accuracy. Errors were classified as ‘‘neglect’’

errors using the ‘‘neglect point’’ criterion of Ellis et al.

(1987). A pathological score in at least one test was con-

sidered as an index of spatial neglect. Table 2 shows the

results of the baseline assessment. Notably, pseudoword

reading performance resulted a more sensible indicator of

neglect dyslexia, as compared with word reading (Vallar

et al. 2010, for review; Martelli et al. 2011).

Stimuli

Three sets of 20 Italian nouns were selected from a corpus

of Italian written language of 3 million tokens (Laudanna

et al. 1995). Words varied in length (5–10–13 letters), and

had a medium word token frequency of 81 (range 5–304).

All words were morphological simple, with the dominant

(or regular) stress pattern and without double consonants or

contextual rules. For 5- and 10-letter words, stimuli were

Fig. 1 Superimposition of the right-hemispheric lesions in 11 right-

brain-damaged N? patients (a) and in 10 right-brain-damaged N-

patients (b). MNI coordinates for the shown axial slices are given.

The number of overlapping lesions is indicated by different colors,

coding increasing frequencies (from violet, n = 1, to red, n = 11).

Regions specifically damaged in right-brain-damaged N? patients

mainly involved the right putamen and the white matter underneath

the insula, the rolandic operculum and the frontal-inferior operculum

Table 1 Demographic and neurological data of 11 right-brain-dam-

aged N? patients and 11 right-brain-damaged N- patients

Patient Sex/age/

education

Etiology/lesion

site

Duration of disease

(months)

N?

DF F/69/8 I/TFP 1

MV M/79/5 I/FT In 2

BA M/68/5 I/TF In 1

PF M/70/17 I/FTO In 2

LE F/72/13 I/P 1

LA M/74/5 I/P 1

DMF M/57/8 I/ FP In ic 1

MP M/61/8 I/TFP 2

GP F/78/5 I/T In Bg 2.5

MA M/48/16 I/PTO 2.5

VI F/71/8 I H/TFP 6

N-

DI F/44/13 I H/FT In Bg 1

CE M/78/13 I/P 1

DA M/75/13 I/ic 1

LG M/67/18 I H/ic Bg 1.5

RA M/73/10 I/P 2.5

VA M/61/13 I/t ic 1

RL M/70/8 I/FT In 1.5

FA F/76/2 I/ec Bg 1

LL M/73/5 I/P 3

PG F/72/11 I/TP 1

FMT F/79/18 I/Bg 1

M/F male/female, I/H ischemic/hemorrhagic lesion, F frontal,

P parietal, T temporal, O occipital, In insula, ic internal capsule, ec

external capsule, Bg basal ganglia

136 Exp Brain Res (2014) 232:133–146

123

no-suffixed nouns, while for 12- or 13-letter words very

few suffixed nouns were included, due to the lack of a

sufficient number of no-suffixed Italian words matched for

length (see ‘‘Appendix’’). All 3 categories of stimuli were

approximately matched for number and position (initial vs.

final) of ascending and descending letters (Fischer 1996).

Each word was printed in lowercase, 54-pt, Arial font. The

following word lengths (mean mm ± SD, range) were

used: 44.6 ± 5.6, 34–55, for 5-letter words; 88.2 ± 6.9,

75–100, for 10-letter words; 104.4 ± 7.9, range 94–126,

for 13-letter words. The distance between the leftward

extremity of the first letter and the rightward extremity of

the last letter was measured to the nearest millimeter.1 A

total of 60 lines (2-pt, 1 mm in width) matched one by one

for length (in millimeter) with the words were designed.

Procedure

Stimuli, both words and lines, were presented in the center

of an A4 horizontally oriented sheet, each page containing

four stimuli located one above the other, with the center of

the sheet being aligned with the mid-sagittal plane of the

participant’s body at a viewing distance of about 40 cm. A

moveable window was used in order to present each

stimulus one at a time. Participants were free to move their

head and eyes throughout the task.

Words and lines were divided into two parts, in which

the three lengths were equally represented. An ABBA

order (words-lines-lines-words) was used for half of the

Table 2 Baseline assessment for left visuo-spatial neglect

Line

cancellation

Letter

cancellation

Star

cancellation

Sentence

reading

Complex figure

drawing

Line

bisection

Clock

drawing

Single word reading

L R L R L R Word Non-word

N?

DF 9/11a 0/10a 53/53a 47/51a 5/30a 0/26a 5/6a 2/5a -6.44 9/10 0 (0/2) .57 (4/7)b

MV 6/11a 0/10a 43/53a 20/51a 30/30a 8/26a 5/6a 3/5a 21.36a 7/10 0 (0/2) .5 (2/4)

BA 0/11 0/10 33/53a 7/51a 0/30 1/26 1/6a 1/5a 13.36a 3/10a 0 (0/1) .33 (4/12)

PF 0/11 0/10 6/53a 3/51a 0/30 1/26 0/6 0/5 4.16 7/10 0 (0/1) .2 (3/15)

LE 8/11a 0/10a 24/53a 16/51a 2/30a 0/26a 1/6a 0/5 12.51a 9/10 0 (0/2) .22 (2/9)

LA 0/11 0/10 34/53a 5/51a n.a. n.a. 1/6a 1/5a n.a. 2.5/10a .17 (1/6) 0 (0/12)

DMF 0/11 0/10 53/53a 31/51a 0/30 0/26 0/6 0/5 10.5a 2.5/10a 0 (0/0) 0 (0/4)

MP 7/11a 0/10a 39/53a 4/51a 15/30a 2/26a 5/6a 0/5 -5.18 10/10 1 (1/1) .5 (1/2)

GP 1/11a 0/10a 53/53a 47/51a 3/30 3/26 2/6a 0/5 -3.24 1/10a 0 (0/3) .21 (3/14)

MA 11/11a 1/10a 8/53a 0/51a 10/30 11/26 5/6a 0/5 16.22a n.a. .77 (10/13)b .91 (31/34)b

VI 4/11a 0/10a 53/53a 22/51a 30/30a 11/26a 4/6a 2/5a 3.89 3.5/10 0 (0/3) .29 (4/14)

N-

DI 0/11 0/10 0/53 0/51 1/30 0/26 0/6 0/5 2.81 9.5/10 0 (0/0) 0 (0/2)

CE 0/11 0/10 0/53 0/51 0/30 0/26 0/6 0/5 4.96 7/10 0 (0/7) .16 (3/18)

DA 0/11 0/10 0/53 0/51 1/30 0/26 0/6 0/5 5.54 10/10 0 (0/2) .16 (1/6)

LG 0/11 0/10 0/53 1/53 0/30 0/26 0/6 0/5 -1.93 9.5/10 0 (0/0) 0 (0/0)

RA 0/11 0/10 0/53 0/51 0/30 0/26 0/6 0/5 .99 7/10 0 (0/0) 0 (0/3)

VA 0/11 0/10 1/53 0/51 0/30 0/26 0/6 0/5 2.69 10/10 0 (0/0) 0 (0/1)

RL 0/11 0/10 2/53 3/51 1/30 0/26 0/6 0/5 3.51 9/10 0 (0/0) 0 (0/0)

FA 0/11 0/10 2/53 0/51 0/30 1/26 0/6 0/5 2.26 6/10 .28 (2/7) .14 (3/22)

LL 0/11 0/10 0/53 0/51 0/30 0/26 0/6 0/5 2.09 8.5/10 0 (0/2) 0 (0/3)

PG 0/11 0/10 0/53 8/51 2/30 3/26 0/6 0/5 3.39 6/10 0 (0/3) .31 (4/13)

FMT 0/11 0/10 0/53 0/51 0/30 0/26 0/6 n.a. 4.82 6.5/10 0 (0/0) 0 (0/2)

Cancellation tasks: number of omissions in the left/right (L/R) hand side of the display. Sentence reading: number of correct responses. Complex

figure drawing: 0/5 indicates errorless performance. Clock drawing: total score ranged from 0 to 10. Line bisection: deviation in millimeter (-/?

leftward/rightward deviation); percentage and number (in brackets) of neglect errors out of the total errors in word and non-word reading

n.a. not assesseda Defective performance, as compared with normative data, indicating left neglectb Neglect errors

1 In the case of words starting with the letters ‘‘t’’ and ‘‘f’’, the

measure began from the vertical bar rather than the horizontal one.

Exp Brain Res (2014) 232:133–146 137

123

patients, and a BAAB for the other half. The experiment

was performed in a single session. Patients were individ-

ually tested in a quiet room, with the experimenter sitting

in front of them. The patients’ task was to bisect each

stimulus, marking its mid-point with a pencil, using the

unaffected ipsilesional right hand. Participants were

informed that the bisection mark could be made in any

point of the word, irrespective of whether it might fall

between two letters or go through a letter. No time limits

were imposed, but participants were instructed not to count

letters. No feedback was given with respect to the accuracy

of the response. The distance between the left end of each

stimulus and the participant’s mark was measured to the

nearest millimeter. Each measure in millimeter was con-

verted into a standardized score (measured left half minus

objective half/objective half 9 100), in order to equate the

participants’ error with respect to stimulus length. This

percent deviation yielded positive values for rightward

deviations with respect to the objective mid-point of the

stimulus, negative values for leftward deviations (Rode

et al. 2006).

Statistical analyses

Average standardized scores were analyzed by repeated-

measures analyses of variance (ANOVAs), with two

within-subjects factors (type of stimulus and length), and

one between-subjects factor (the groups of participants). In

order to control for a possible effect of lesion size in

determining the amount of the patients’ bisection error, and

its direction, an analysis of covariance (ANCOVA) on the

mean deviations was carried out, with group (N?, N-) as

the between-subjects factor, type of stimulus and length as

the within-subjects factors, and lesion size as the covariate

(mean-centered prior to the analysis). The Greenhouse-

Geisser correction for repeated-measures analyses

(Greenhouse and Geisser 1959) was used, in order to cor-

rect for violations of the sphericity assumption whenever

necessary. For every analysis, we calculated the partial Eta

Squared (pg2), which measures the proportion of the total

variance that was attributable to a main factor or to an

interaction (Cohen 1973). Whenever necessary, pairwise

comparisons were performed with Student-Newman-Ke-

uls’ post hoc multiple comparisons. Mean bisection per-

formances were compared to accurate bisection through

one-sample t tests against zero. Subsequently, to assess for

any significant defective performance with different types

of stimuli in individual patients with left unilateral spatial

neglect, we compared the average deviation errors of each

patient in the word and line tests with the mean deviations

of healthy participants and of the N- patients. The anal-

yses were performed by t tests (Crawford and Garthwaite

2002), testing whether the score of an individual patient

was significantly different from the mean score of control

participants. Using the Revised Standardized Different Test

(Crawford and Garthwaite 2005), we also tested for each

patient whether the difference between the standardized

scores on word and line bisection was significantly dif-

ferent from the correspondent mean differences found in

the control samples. Finally, Pearson’s correlations, with

Bonferroni correction for multiple comparisons, were used

in order to measure the association between bisection

errors for words and lines in the three groups.

Results

Since a previous analysis showed the lack of significant

differences between medium and long stimuli for all

groups of participants, these two conditions were merged

together and compared with short stimuli. Figure 2 shows

the average percentage error by group (N?, N-, and C

participants), stimulus (words, lines), and lengths (short

and long). N? patients made a greater rightward mean

error with long stimuli with respect to short ones, as

compared to those made by N- patients and control

participants.

The ANOVA revealed significant main effects of length

(short, 5 letters; long, 10–13 letters) (F1,30 = 9.50;

p = .004, pg2 = .24), while the main effects of group (N?

patients, N- patients, and control participants)

(F2,30 = 2.56; p = .09; pg2 = .15) and type of stimulus

(words vs. lines) were not significant (F1,30 = .72;

p = .40; pg2 = .02). The length by group (F2,30 = 7.17;

p = .003; pg2 = .32) and the type of stimulus by length

(F1,30 = 5.49; p = .03; pg2 = .15) interactions were sig-

nificant. The type of stimulus by group (F2,30 = .05;

Fig. 2 Experiment I: mean percent deviation error (±SE) by group

[11 N? patients, 11 N- patients, and 11 control participants (C)],

stimulus type (words and lines), and stimulus length (short: light gray

column; long: dark gray column)

138 Exp Brain Res (2014) 232:133–146

123

p = .95; pg2 = .003) and the type of stimulus by length by

group (F2,30 = .62; p = .54; pg2 = .04) interactions were

not significant. Post hoc multiple comparisons revealed that

N? patients showed a ‘‘length effect,’’ with a mean smaller

rightward error in bisecting short (1.05 %) versus long

(10.55 %) stimuli. Average bisection scores of N- patients

did not differ by stimulus length (short: 1.57 % and long:

2.52 %), as well as those of C participants (short: -.005 %

and long: -.02 %). Rightward deviations made by N?

patients with long stimuli are larger than the biases

exhibited by N- and C participants with all types of

stimuli, both lines and words, with no significant differ-

ences for short stimuli. This means that N? patients

bisected long stimuli more to right when compared to both

C participants and N- patients. No significant differences

were found between lengths, when comparing the perfor-

mances of the N- and C groups. As the type of stimulus by

length interaction, while mean deviations for short words

(.74 %) did not differ from that for short lines (1.01 %),

mean deviations for long words (3.25 %) were smaller than

those for long lines (5.45 %): This means that long lines

were globally bisected more to the right than long words,

independently from the group of participants.

The ANCOVA revealed a significant main effect of the

covariate lesion size (F1,18 = 14.12; p = .001; pg2 = .44).

Importantly, the main effect of length (F1,18 = 9.49;

p = .006; pg2 = .35) and the length by group interaction

(F1,18 = 5.51; p = .03; pg2 = .23) were significant. The

main effect of type of stimulus, of group and all the other

interactions were not significant.

In the N? group, mean rightward bisection errors dif-

fered from zero (i.e., accurate bisection) for long lines

(11.44 %, t10 = 3.25, p = .009); they were almost signif-

icant for long words (9.65 %, t10 = 2.21, p = .05), while

no differences were found for the other types of stimuli:

short words (.50 %, t10 = .12, p = .90) and short lines

(1.60 %, t10 = .41, p = .69). In the N- group, mean

rightward bisection errors differed from zero for long

(4.08 %, t10 = 4.10, p = .002), but not for short (1.68 %,

t10 = 1.51, p = .16) lines, as well as for words of both

lengths (short: 1.46 %, t10 = 1.29, p = .23; long: .96 %,

t10 = .95, p = .36). In the C group, mean bisection errors

did not differ from zero for all types of stimulus.

The performances of the individual N? patients were

compared with those of both N- patients and C partici-

pants for long stimuli, since the above reported ANOVAs

had shown no differences between groups for short stimuli

(see Fig. 3). Comparing the difference between word and

line bisection in the individual patient with the same

average difference in the control groups (both N- patients

and C participants), patient MV bisected long lines more

rightwards than long words, while three patients out of 11

exhibited larger rightward displacements for long words

than for long lines (MP, GP, and MA).

In the C group, Bonferroni-corrected (p \ .008) Pear-

son’s correlations performed on bisection deviations were

significant between short and long words (r = .76,

p \ .008) and short and long lines (r = .81, p \ .008),

with all other correlations not attaining the significance

level. In the N? group, correlations were not significant for

all types of comparison, as well as in the N- group.

In sum, data from single-patient analyses concur to

suggest that left spatial neglect affects in a differential

fashion the patients’ bisection of words and lines.

Experiment II

As discussed in the Introduction, Italian words with three

or more syllables have two main stress patterns: they can

Fig. 3 Experiment I: mean percent deviation error by stimulus type

(words and lines) and length (long), made by each of the 11 N?

patients. The mean error (±SE) of the 11 N- patients and the 11 C

participants is shown in the bottom row. *Significant difference

between each patient’s score and the mean score of the C group for

each stimulus condition. **Significant difference between each

patient’s score and both the mean score of the N- and of the C

groups for each stimulus condition. Circle and brackets: significant

difference between word and line bisection in the individual patient,

compared to the average difference in the C group. Double circle and

brackets: significant difference between word and line bisection in the

individual patient, compared to both the average difference in the N-

and in the C groups

Exp Brain Res (2014) 232:133–146 139

123

be stressed on the penultimate (last but one) syllable (e.g.,

matıta, pencil, ‘‘dominant’’ stress pattern), or on the ante-

penultimate (last but two) syllable (e.g., bıbita, drink,

‘‘non-dominant’’ stress pattern). Moreover, also the final

letters of the word provide informational cues for attrib-

uting the correct word stress pattern. Accordingly, right-

brain-damaged patients with left neglect, who typically

attend the right hand side of a word more than the left hand

side, may manifest this effect, possibly even strongly (see

Daini et al. 2002, for related evidence the bisection per-

formance of the Brentano version of the Muller-Lyer by

right-brain-damaged neglect patients). In particular, the

rightward final sequence, which includes information as to

the probability that the word is stressed on the penultimate

or the antepenultimate syllable, could modulate bisection

accuracy: specifically, guessing that the word is likely to be

irregular, patients could orient toward its initial (left-sided)

letters, thus reducing the rightward, neglect-related, bias.

To this aim, both patients and neurologically unimpaired

control participants set manually the mid-point of words

with the two different stress patterns and final sequences,

which, as discussed earlier, can be used as orthographic

cues for predicting word stress.

Stimuli and procedure

The stimuli of Burani and Arduino (2004) were used.

Stimuli were divided into two lists, according to the type of

final sequences: (1) 30 words containing final sequences as

-oro, which characterized more regularly stressed words

(penultimate final sequence words); (2) 30 words contain-

ing final sequences as -ola, which characterized more

irregularly stressed words (antepenultimate final sequence

words). Stimuli were 6 to 9 letters long, and were of low

frequency (Istituto di Linguistica Computazionale

1989) (see ‘‘Appendix’’). As in Experiment I, each word

was printed in lowercase, 54-pt, Arial font: mean length of

the stimuli was 55.2 mm (SD = ±9.49, range 41–76) for

penultimate final sequence words, and 55.2 mm

(SD = ±10.57, range 42–73) for antepenultimate final

sequence words. Procedures and scoring modality were

identical to those of Experiment I.

Statistical analyses

The data were analyzed by repeated-measures ANOVAs

with a within-subjects factor (type of stimulus), and one

between-subjects factor (the three groups of participants).

Mean bisection performances were compared to accurate

bisection through one-sample t tests against zero. The

performances of the individual patients were analyzed by

t tests (Crawford and Garthwaite 2002), comparing each

patient score with the mean score of control participants.

Pearson’s correlations were carried out in order to measure

the association between errors for penultimate and ante-

penultimate final sequence words in the three groups of

participants.

Results

As shown in Fig. 4, all participants showed a greater rel-

ative rightward deviation with penultimate final sequence

words with respect to antepenultimate final sequence

words. The ANOVA revealed a significant main effect of

the type of stimulus (F1,30 = 5.91, p = .02, pg2 = .16),

while the main effect of group (F2,30 = .40; p = .67,

pg2 = .03), and the type of stimulus by group interaction

(F2,30 = 1.68; p = .20, pg2 = .10) were not significant. In

order to control whether length could have an effect even

with these types of stimuli, a post hoc analysis was con-

ducted dividing words by number of letters (6–7 letters vs.

8–9 letters). The ANOVA revealed a significant main

effect of length (F1,30 = 6.94, p = .01, pg2 = .19), while

the main effect of group (F2,30 = 1.42; p = .26,

pg2 = .09) and of the type of stimulus were not significant

(F1,30 = 1.91; p = .18, pg2 = .06). The length by group

(F2,30 = 13.37; p \ .001; pg2 = .47) and the type of

stimulus by length (F1,30 = 4.61; p = .04; pg2 = .13)

interactions were significant. The type of stimulus by group

(F2,30 = .18; p = .84; pg2 = .01) and the type of stimulus

by length by group (F2,30 = .35; p = .71; pg2 = .02)

interactions were not significant. Post hoc multiple com-

parisons revealed that N? patients showed a ‘‘length

effect,’’ with a mean smaller rightward error in bisecting

short (2.09 %) versus long (8.13 %) words, replicating

results of Experiment I. Average bisection scores of N-

patients did not differ by stimulus length (short: 1.80 %

Fig. 4 Experiment II: mean percent deviation error (±SE) by group

(N? and N- patients, and C participants), and by stimulus condition

(penultimate final sequences: dark gray column; ante-penultimate:

light gray column)

140 Exp Brain Res (2014) 232:133–146

123

and long: 2.08 %), as well as those of C participants (short:

.44 % and long -.99 %). No other comparisons were

significant. As the type of stimulus by length interaction,

the mean deviation for short antepenultimate final sequence

words (.32 %) was smaller than those for short penultimate

final sequence words (2.57 %), confirming the lexical

effect found in the previous analysis of variance for short

stimuli. The mean deviation for long antepenultimate final

sequence words (3.19 %) did not differ from those for long

penultimate final sequence words (2.95 %). Being a post

hoc analysis, the lack of lexical effect with long words

could be due to the fact that long stimuli are very few with

respect to short ones: the small number of items could be

insufficient to permit the emergence of such an effect.

However, caution is needed and more controlled studied

should be conducted in order to investigate the relationship

between lexical factors and length in word bisection.

In the N? group, the mean bisection error did not differ

from zero (i.e., accurate bisection) for both types of stimuli

(penultimate, 4.38 %, t10 = 1.09, p = .30; antepenulti-

mate: 1.92 %, t10 = .54, p = .60), as well as in the C group

(penultimate: .33 %, t10 = 2.88, p = .78; antepenultimate:

.16 %, t10 = .16, p = .87). In the N- group, penultimate

(2.42 %, t10 = 2.87, p = .02), but not antepenultimate

(1.31 %, t10 = 1.31, p = .22) final sequence words differed

from zero.

Comparing the mean shifts in the penultimate and ante-

penultimate final sequence word conditions for each patient,

with those of both N- patients and healthy participants,

patients GP and LA showed significant rightward biases for

both types of stimuli (GP penultimate, ?33.86 %,

N- patients: t10 = 10.75, p \ .001; healthy participants:

t10 = 8.51, p \ .001; GP antepenultimate, ?23.06 %,

N- patients: t10 = 6.25, p \ .001; healthy participants:

t10 = 6.79, p \ .001; LA penultimate, ?10.55 %,

N- patients: t10 = 2.78, p = .02; healthy participants:

t10 = 2.59, p = .02; LA antepenultimate, ?12.94 %,

N- patients: t10 = 3.34, p = .007; healthy participants:

t10 = 3.80, p = .003). BA exhibited leftward deviations for

both types of stimuli (penultimate, -23.74 %, N- patients:

t10 = -8.94, p \ .001; healthy participants: t10 = -6.11,

p \ .001; antepenultimate, -25.40 %, N- patients: t10 =

-7.68, p \ .001; healthy participants: t10 = -7.60, p \ .001).

Errors made by all other N? patients did not differ from those

of N- patients and healthy control participants (p [ .05, for

all comparisons). For patient GP, the difference between the

bisection errors in penultimate and antepenultimate final

sequence words was significant (t29 = 2.84, p = .008),

while for patient LA (t29 = .65, p = .52) it was not. Pear-

son’s correlations performed on bisection deviations were

significant between penultimate and antepenultimate final

sequence words in all groups of participants (N?: r = .95,

p \ .001; N-: r = .79, p \ .005; C: r = .88, p \ .001).

Discussion

The present study had the twofold aim of investigating:

(1) how perceptual and lexical features modulate the

bisection of orthographic strings in right-brain-damaged

patients and neurologically unimpaired control partici-

pants; (2) whether unitary or independent processes sup-

port the bisection of lines and orthographic material (see

for discussion of this issue Arduino et al. 2010). To these

aims, the effects of stimulus type (lines vs. words),

stimulus length (short vs. long), and lexical features (word

stress) on bisection accuracy were assessed. In Experiment

I (bisection of lines and words of different lengths), right-

brain-damaged patients with left spatial neglect make, as a

group, an average rightward error larger than those of

patients without neglect and healthy participants, and

show a length effect, with a greater rightward directional

bias for longer stimuli, comparable for lines and words.

These results are in line with previous findings from line

bisection tasks in right-brain-damaged patients with left

spatial neglect (see Bisiach et al. 1983; Marshall and

Halligan 1989; Vallar et al. 2000; Daini et al. 2002), and

extend them to the bisection of words. Furthermore, long

lines are bisected more rightwards when compared to long

words, independently from the group of participants. This

result is in line with previous findings on word bisection

(Fischer 1996, 2000), and could be attributed to reading

strategy activated by the type of stimulus (i.e., words)

even in patients with spatial neglect, possibly involving

automatic lexical access, which focuses attention on the

word beginning (Attentional Scaling Hypothesis; Fischer

1996, 2000, 2004). However, a perusal of the perfor-

mances of individual patients shows that left spatial

neglect may affect differently bisection deviations for

words and lines. In particular, patient DF exhibits a larger

rightward bias with words, as compared with lines, as

patients MP, GP, and MA. Conversely, in patient MV, the

rightward bias is greater with lines. The different impact

of spatial neglect on word and line bisection is supported

by the significant correlations, in control participants, of

the bisection errors for different lengths of the same

stimulus type (words, lines), but not between the two

different types of stimuli (words vs. lines). These findings

suggest that the processes involved in word and line

bisection are likely to be, at least in part, independent (see

also Arduino et al. 2010). At first, there is a difference

between lines and strings emerging early on in a lower-

level processing stage: Lines and words differ perceptu-

ally, being composed by continuous (lines) versus discrete

(letter strings, words) elements. In line with this view,

healthy participants show comparable deviations in the

bisection of both orthographic and symbolic strings (Ar-

duino et al. 2010). Although the bisection task explicitly

Exp Brain Res (2014) 232:133–146 141

123

requires a global processing of stimulus extent, in the case

of discrete strings, a processing of the local features may

be involved (see Gallace et al. 2008).

When focusing on the group of controls matched by age

with the patients, they do not show pseudoneglect, being

accurate with both lines and words. These results are in line

with a reduced pseudoneglect exhibited by elder partici-

pants in line bisection (see Jewell and McCourt 2000, for

review). Age may explain the lack of a left deviation even

for words, since the sample studied by Fischer (1996,

2004)—who reported a leftward bisection bias for words in

healthy participants—is younger than the one enrolled in

the present study. The reduction in the leftward deviation

could be interpreted in light of the hypothesis of a dis-

proportionate aging of the right hemisphere compared the

left one (Dolcos et al. 2002; Schmitz and Peigneux 2011).

Furthermore, right-brain-damaged N- patients exhibit a

rightward deviation for long lines, being accurate with all

the other types of stimuli, and this is in accordance with a

right-hemispheric involvement in the bisection task. Con-

versely, N? patients exhibit rightward deviations for lines

and words (the last almost significant), while they are

accurate with short stimuli.

A number of studies have investigated visuo-spatial

processing of verbal material with different results. In a

task requiring participants to circle the central letter of very

long character and symbol strings, as compared with line

bisection, neurologically unimpaired participants exhibit a

rightward deviation, which, in right-brain-damaged

patients with left neglect, is disproportionate (Lee et al.

2004). As compared with solid lines, greater demands on

local processing, left-hemisphere-supported, attentional

resources (Martin 1979) may be posed by the task of

bisecting strings of characters, and, more generally, strings

of discrete items. This may result in a task-induced left-

hemispheric activation, which in turn may bring about in

unimpaired participants a rightward bias, and exacerbate

the pathological rightward deviation made by right-brain-

damaged patients with left spatial neglect. Similarly, neu-

rologically unimpaired participants may show a rightward

deviation in bisecting long letter strings in which no word

is inserted, or in which a word (neutral, emotional) is

placed in the right or in the left hand side of the letter string

(Mohr and Leonards 2007). These findings argue for a role

of left-hemisphere-mediated semantic verbal processing

(Bowers and Heilman 1980) in letter and line bisection. In

the present study, as in Fischer (1996, 2000, 2004), the type

of stimuli—i.e., single words—could have activated a

reading strategy (from left to right in both English and

Italian language), once again mediated by the left hemi-

sphere and responsible of the leftward shifting of attention

and of the consequent bisection bias. As already discussed,

the differences with Fischer’s studies could be explained

by the age of the samples enrolled. The activation of a

reading strategy may be due to various factors, such as

stimulus readability, stimulus length, and the type of task

required. Fischer reported a leftward bisection error also

for pronounceable pseudowords and letter strings of com-

parable length to real words (Fischer 1996, 2000, 2004),

but not when vowels are replaced by x’s (Fischer 1996).

Furthermore, the stimuli used by Lee et al. (2004) and

Mohr and Leonards (2007) are very long (45–48 letters)

with respect to ones used in Fischer’s (1996, 2000, 2004)

and in the present study: The length of the stimuli would

prevent the processing of the letter strings as words, thus

not activating the reading strategy. Finally, the type of task

required by Lee et al. (2004) is to identify the target closest

to the mid-point, combining the demands of length judg-

ment with the visual search required in target cancellation.

As stated by the authors, this would require additional local

feature analyses mediated by the left hemisphere, in addi-

tion to global attention requested in mid-point judgment.

More in general, the double dissociation found in the

present study suggests that the processes involved in word

and line bisection are likely to be, at least in part, inde-

pendent: Other than the linguistic mechanism supported by

a reading strategy, a difference between continuous and

discrete material could be crucial (see also Arduino et al.

2010).

Experiment II assessed the impact of linguistic infor-

mation on word bisection in right-brain-damaged patients

with and without visuo-spatial neglect and control par-

ticipants. A reduced rightward bias when bisecting words

containing those final sequences characterizing irregu-

larly stressed stimuli (namely, with stress on the ante-

penultimate syllable) was found, at least for short words,

in the three groups of participants, with the different

types of words correlating within each group. The

overall lack of differences among groups may possibly

reflect the short (55 mm) length of the stimuli used.

There is wide evidence that the magnitude of the

patients’ bisection error is modulated by stimulus length,

with larger rightward errors with longer lines (see, e.g.,

Vallar et al. 2000 and references therein). In fact, the

mean length of both types of words (55.2 mm) is

broadly comparable to the mean length of the short

(44.6 mm) stimuli used in Experiment I, in which group

differences emerge only with longer words. Nevertheless,

patient GP, who exhibits left spatial neglect with both

types of stimuli, as compared to control participants,

shows a minor rightward deviation with antepenultimate

compared to penultimate final sequence words, with a

directional pattern similar to that of patients without

neglect and control participants. Orthographic and pho-

nological information present in the right hand side of

the stimulus appears to be able to induce a leftward bias.

142 Exp Brain Res (2014) 232:133–146

123

In conclusion, the present study provides evidence from

right-brain-damaged patients with left spatial neglect that

the processing of the lateral extent of line and word stimuli

is supported by at least partially independent processes,

which can be differentially affected by brain damage, and

that word bisection is influenced by lexical variables. More

generally, the processes which support the explicit esti-

mation of the lateral extent of linguistic material through

the setting of the mid-point of a horizontal letter string

(admittedly, a laboratory task, whose role in ecological

linguistic and visuo-spatial performances is not definite so

far), appear –as suggested by the double dissociation found

in individual patients between disproportionate rightward

deviations in word and line bisection– to rely on visuo-

spatial components specific for written language repre-

sentations. These, in turn, modulate bisection performance,

as shown by the results of Experiment II. These conclu-

sions are in line with the established dissociation between

visuo-spatial neglect (as indexed by cancellation and line

bisection tasks Bisiach and Vallar 2000, for review) and

neglect dyslexia (Vallar et al. 2010, for review), and extend

it to the computation of extent, as assessed by bisection of

different materials (line vs. letter strings). The present

pattern of double dissociations suggests that at least partly

different processes contribute to bisection performance of

letter strings and lines, which may be differentially affected

by right brain damage, bringing about selective patterns of

left spatial neglect.

Acknowledgments This study was performed in partial fulfillment

of the PhD program of L. V.. G. V. was supported in part by FAR

Grants from the University of Milano-Bicocca, and by Ricerca Cor-

rente Grants from the IRCCS Istituto Auxologico Italiano.

Appendix

Word sets of Experiment I

5-letter-long: tasca, posta, gesto, monte, spesa, vetro, disco,

poeta, fiore, borsa, cesto, perla, nuora, spiga, tarlo, mensa,

avena, fungo, amaca, talpa.

10-letter-long: conferenza, equilibrio, patrimonio, mag-

istrato tecnologia, entusiasmo, inflazione, dimensione,

contributo, superficie, parsimonia, pneumatico, mascalz-

one, crepuscolo, camaleonte, partigiano, didascalia, pro-

boscide, rimprovero, coriandolo.

12/13-letter-long: distribuzione, proprietario, informaz-

ione, testimonianza, disperazione, comunicazione, circo-

lazione, imprenditore, parlamentare, responsabile,

mobilitazione, verniciatura, prestigiatore, centralinista,

destinatario, registratore, razionalismo, orfanotrofio, trep-

idazione, comunicatore.

Experiment I: list of words, number of letters (N) and lengths of

words and comparable lines expressed in millimeter (L)

Word N L

1 fiore 5 34

2 tasca 5 43

3 poeta 5 45

4 posta 5 44

5 vetro 5 41

6 monte 5 51

7 disco 5 43

8 borsa 5 45

9 spesa 5 49

10 gesto 5 45

11 spiga 5 44

12 fungo 5 45

13 avena 5 50

14 amaca 5 55

15 Tarlo 5 35

16 talpa 5 38

17 perla 5 40

18 cesto 5 44

19 nuora 5 46

20 mensa 5 55

21 tecnologia 10 84

22 inflazione 10 78

23 contributo 10 82

24 dimensione 10 96

25 entusiasmo 10 96

26 patrimonio 10 87

27 equilibrio 10 75

28 conferenza 10 92

29 superficie 10 80

30 magistrato 10 87

31 camaleonte 10 97

32 mascalzone 10 100

33 didascalia 10 83

34 crepuscolo 10 91

35 parsimonia 10 91

36 partigiano 10 82

37 pneumatico 10 96

38 proboscide 10 91

39 coriandolo 10 86

40 rimprovero 10 89

41 disperazione 12 107

42 testimonianza 13 115

43 imprenditore 12 104

44 distribuzione 13 105

45 circolazione 12 99

46 parlamentare 12 110

47 proprietario 12 95

Exp Brain Res (2014) 232:133–146 143

123

Word sets from Experiment II

Penultimate final sequence words: alloro, ardita, canora,

castoro, decoro, dimora, eremita, fallita, malora, papiro,

parassita, ristoro, salita, sonoro, traforo, acrobata, angora,

bibita, canfora, cernita, decrepita, forfora, fosforo, logoro,

onnivoro, pecora, porpora, satiro, tacita, tortora.

Antepenultimate final sequence words: atomica, bambola,

bussola, fertile, fossile, muscolo, organica, ostacolo, pas-

colo, pugile, rettile, sogliola, tattile, tessile, tipica, badile,

barile, capriola, cazzuola, fienile, fucile, mollica, ortica,

ostile, pignolo, pistola, sedile, tritolo, usignolo, vescica.

Experiment II: list of words, stressed final sequence (StrSeq): pen-

ultimate (p) versus antepenultimate (a); lengths of words and com-

parable lines expressed in millimeter (L)

Word StrSeq L

1 alloro p 45

2 ardita p 45

3 canora p 56

4 castoro p 61

5 decoro p 57

6 dimora p 56

7 eremita p 61

8 fallita p 41

9 malora p 55

10 papiro p 51

11 parassita p 74

12 ristoro p 51

13 salita p 42

14 sonoro p 57

Word StrSeq L

15 traforo p 53

16 acrobata p 72

17 angora p 57

18 bibita p 43

19 canfora p 61

20 cernita p 55

21 decrepita p 76

22 forfora p 52

23 fosforo p 56

24 logoro p 51

25 onnivoro p 72

26 pecora p 55

27 porpora p 63

28 satiro p 45

29 tacita p 42

30 tortora p 52

31 atomica a 65

32 bambola a 70

33 bussola a 63

34 fertile a 44

35 fossile a 52

36 muscolo a 69

37 organica a 71

38 ostacolo a 69

39 pascolo a 63

40 pugile a 49

41 rettile a 45

42 sogliola a 62

43 tattile a 43

44 tessile a 51

45 tipica a 42

46 badile a 49

47 barile a 44

48 capriola a 65

49 cazzuola a 73

50 fienile a 47

51 fucile a 42

52 mollica a 56

53 ortica a 45

54 ostile a 43

55 pignolo a 59

56 pistola a 52

57 sedile a 48

58 tritolo a 44

59 usignolo a 69

60 vescica a 62

Word N L

48 comunicazione 13 126

49 informazione 12 107

50 responsabile 12 106

51 trepidazione 12 101

52 razionalismo 12 105

53 comunicatore 12 113

54 centralinista 13 100

55 prestigiatore 13 103

56 verniciatura 12 97

57 destinatario 12 97

58 orfanotrofio 12 95

59 registratore 12 94

60 mobilitazione 13 109

144 Exp Brain Res (2014) 232:133–146

123

References

Albert ML (1973) A simple test of visual neglect. Neurology

23:658–664

Arduino LS, Previtali P, Girelli L (2010) The centre is not in the

middle: evidence from line and word bisection. Neuropsycho-

logia 48:2140–2146

Bisiach E, Vallar G (2000) Unilateral neglect in humans. In: Boller F,

Grafman J, Rizzolatti G (eds) Handbook of neuropsychology,

2nd edn. Elsevier Science, Amsterdam, pp 459–502

Bisiach E, Capitani E, Colombo A, Spinnler H (1976) Halving a

horizontal segment: a study on hemisphere-damaged patients

with cerebral focal lesions. Schweiz Arch Neurol Neurochir

Psychiatr 118:199–206

Bisiach E, Bulgarelli C, Sterzi R, Vallar G (1983) Line bisection and

cognitive plasticity of unilateral neglect of space. Brain Cogn

2:32–38

Bowers D, Heilman KM (1980) Material-specific hemispheric

activation. Neuropsychologia 18:309–319

Burani C, Arduino LS (2004) Stress regularity or consistency?

Reading aloud Italian polysyllables with different stress patterns.

Brain Lang 90:318–325

Cattaneo Z, Fantino M, Mancini F, Mattioli F, Vallar G (2012)

Listening to numbers affects visual and haptic bisection in

healthy individuals and neglect patients. Neuropsychologia

50:913–925

Cohen J (1973) Eta-squared and partial eta-squared in fixed factor

ANOVA designs. Educ Psychol Meas 33:107–112

Crawford JR, Garthwaite PH (2002) Investigation of the single case in

neuropsychology: confidence limits on the abnormality of test

scores and test score differences. Neuropsychologia 40:1196–1208

Crawford JR, Garthwaite PH (2005) Testing for suspected impair-

ments and dissociations in single-case studies in neuropsychol-

ogy: evaluation of alternatives using Monte Carlo simulations

and revised tests for dissociations. Neuropsychology 19:318–331

Cubelli R, Beschin N (2005) The processing of the right-sided accent

mark in left neglect dyslexia. Brain Lang 95:319–326

Daini R, Angelelli P, Antonucci G, Cappa SF, Vallar G (2002)

Exploring the syndrome of spatial unilateral neglect through an

illusion of length. Exp Brain Res 144:224–237

Diller L, Weinberg J (1977) Hemi-inattention in rehabilitation. The

evolution of a rational remediation program. Raven Press, New

York

Dolcos F, Rice HJ, Cabeza R (2002) Hemispheric asymmetry and

aging: right hemispheric decline or asymmetry reduction.

Neurosci Biobehav Rev 26:819–825

Ellis AW, Flude BM, Young AW (1987) Neglect dyslexia and the

early visual processing of letters in words and non words. Cogn

Neuropsychol 4:439–464

Fischer MH (1996) Bisection performance indicates spatial word

representation. Cogn Brain Res 4:163–170

Fischer MH (2000) Word centre is misperceived. Perception

29:337–354

Fischer MH (2004) Orthographic contributions to perceived word

center. Brain Lang 88:321–330

Folstein MF, Folstein SE, McHugh PR (1975) Mini-mental state: a

practical method for grading the cognitive state of patients for

the clinician. J Psychiatr Res 12:189–198

Fortis P, Maravita A, Gallucci M, Ronchi R, Grassi E, Senna I, Olgiati

E, Perucca L, Banco E, Posteraro L, Tesio L, Vallar G (2010)

Rehabilitating patients with left spatial neglect by prism exposure

during a visuomotor activity. Neuropsychology 24:681–697

Gainotti G, Messerli P, Tissot R (1972) Qualitative analysis of

unilateral spatial neglect in relation to laterality of cerebral

lesions. J Neurol Neurosurg Psychiatr 35:545–550

Gallace A, Imbornone E, Vallar G (2008) When the whole is more

than the sum of the parts: evidence from visuospatial neglect.

J Neuropsychol 2:287–413

Greenhouse SW, Geisser S (1959) On methods in the analysis of

profile data. Psychometrika 24:95–112

Istituto di Linguistica Computazionale, CNR (1989) Corpus di

Italiano contemporaneo. Unpublished manuscript, Pisa

Jewell G, McCourt ME (2000) Pseudoneglect: a review and meta-

analysis of performance factors in line bisection tasks. Neuro-

psychologia 38:93–110

Laudanna A, Thornton AM, Brown G, Burani C, Marconi L (1995)

Un corpus dell’italiano scritto contemporaneo dalla parte del

ricevente. Cisu, Roma

Lee BH, Kang SJ, Park JM, Son Y, Lee KH, Adair JC (2004) The

character-line bisection task: a new test for hemispatial neglect.

Neuropsychologia 42:1715–1724

Leibovitch FS, Black SE, Caldwell CB, Ebert PL, Ehrlich LE, Szalai

JP (1998) Brain-behavior correlations in hemispatial neglect

using CT and SPECT: the Sunnybrook Stroke Study. Neurology

50:901–908

Magni E, Binetti G, Bianchetti A, Rozzini R, Trabucchi M (1996)

Mini-mental state examination: a normative study in Italian

elderly population. Eur J Neurol 3:198–202

Marshall JC, Halligan PW (1989) When right goes left: an

investigation of line bisection in a case of visual neglect. Cortex

25:503–515

Martelli ML, Arduino LS, Daini R (2011) Two different mechanisms

for omission and substitution errors in neglect dyslexia. Neuro-

case 17:122–132

Martin M (1979) Hemispheric specialization for local and global

processing. Neuropsychologia 17:33–40

Mohr C, Leonards U (2007) Rightward bisection errors for letter lines:

the role of semantic information. Neuropsychologia 45:295–304

Mondini S, Mapelli D, Vestri A, Bisiacchi PS (2003) Esame

neuropsicologico breve. Una batteria di test per lo screening

neuropsicologico. Raffaello Cortina Editore, Milano

Na DL, Adair JC, Choi SH, Seo DW, Kang Y, Heilman KM (2000)

Ipsilesional versus contralesional neglect depends on attentional

demands. Cortex 36:455–467

O’Regan JK, Jacobs AM (1992) Optimal viewing position effect in

word recognition: a challenge to current theory. J Exp Psychol

Hum Percept Perform 18:185–197

O’Regan JK, Levy-Schoen A, Pynte J, Brugaillere B (1984) Conve-

nient fixation location within isolated words of different length

and structure. J Exp Psychol Hum Percept Perform 10:250–257

Oldfield RC (1971) The assessment and analysis of handedness: the

Edinburgh inventory. Neuropsychologia 9:97–113

Paap KR, Newsome SL, McDonald JE, Schvaneveldt RW (1982) An

activation-verification model for letter and word recognition: the

word-superiority effect. Psychol Rev 89:573–594

Rode G, Michel C, Rossetti Y, Boisson D, Vallar G (2006) Left size

distortion (hyperschematia) after right brain damage. Neurology

67:1801–1808

Ronchi R, Posteraro L, Fortis P, Bricolo E, Vallar G (2009)

Perseveration in left spatial neglect: drawing and cancellation

tasks. Cortex 45:300–312

Rorden C, Brett M (2000) Stereotaxic display of brain lesions. Behav

Neurol 12:191–200

Rusconi ML, Cappa SF, Scala M, Meneghello F (2004) A lexical

stress effect in neglect dyslexia. Neuropsychology 18:135–140

Schmitz R, Peigneux P (2011) Age-related changes in visual

pseudoneglect. Brain Cogn 76:382–389

Sulpizio S, Arduino LS, Paizi D, Burani C (2013) Stress assignment

in reading Italian (polysyllabic) pseudowords. J Exp Psychol

Learn Mem Cogn 39:51–68

Exp Brain Res (2014) 232:133–146 145

123

Vallar G (1998) Spatial hemineglect in humans. Trends Cogn Sci

2:87–97

Vallar G (2000) The methodological foundations of human neuro-

psychology: studies in brain-damaged patients. In: Boller F,

Grafman J, Rizzolatti G (eds) Handbook of neuropsychology,

2nd edn. Elsevier Science, Amsterdam, pp 305–344

Vallar G, Bolognini N (2013) Unilateral spatial neglect. In: Nobre

AC, Kastner S (ed) Oxford handbook of attention. Oxford

University Press, Oxford

Vallar G, Rusconi ML, Fontana S, Musicco M (1994) Tre test di

esplorazione visuo-spaziale: taratura su 212 soggetti normali.

Arch Psicol Neurol Psichiatr 55:827–841

Vallar G, Guariglia C, Nico D, Tabossi P (1996) Left neglect dyslexia

and the processing of neglected information. J Clin Exp

Neuropsychol 18:1–14

Vallar G, Daini R, Antonucci G (2000) Processing of illusion of

length in spatial hemineglect. A study of line bisection.

Neuropsychologia 38:1087–1097

Vallar G, Burani C, Arduino LS (2010) Neglect dyslexia: a review of

the neuropsychological literature. Exp Brain Res 206:219–235

Wilson B, Cockburn J, Halligan PW (1987) Behavioural inattention

test. Thames Valley Test Company, Titchfield

Zoccolotti P, Antonucci G, Judica A, Montenero P, Pizzamiglio L,

Razzano C (1989) Incidence and evolution of the hemi-neglect

disorder in chronic patients with unilateral right brain-damage.

Int J Neurosci 47:209–216

146 Exp Brain Res (2014) 232:133–146

123