Lexical and gestural symbols in left-damaged patients

Q10

Q1 Q2

www.sciencedirect.com123456789

1011121314151617181920212223242526272829303132333435363738394041424344454647484950515253545556575859606162636465

c o r t e x x x x ( 2 0 1 2 ) 1e1 1

CORTEX899_proof ■ 2 October 2012 ■ 1/11

Available online at

Journal homepage: www.elsevier.com/locate/cortex

66676869707172737475767778

Research report

Lexical and gestural symbols in left-damaged patients
79808182838485
Liuba Papeo* and Raffaella I. Rumiati

Scuola Internazionale di Studi Superiori Avanzati e SISSA, Via Bonomea 265, 34136 Trieste, Italy

8687888990919293949596979899

100101102103

a r t i c l e i n f o

Article history:

Received 20 February 2012

Reviewed 22 March 2012

Revised 7 August 2012

Accepted 5 September 2012

Action editor Georg Goldenberg

Published online xxx

Keywords:

Action understanding

Aphasia

Apraxia

Communicative symbolic gestures

Language

* Corresponding author. Present address: DCambridge, MA 02138, USA.

E-mail addresses: [email protected]

104105106107108109110111

Please cite this article in press as: Papeohttp://dx.doi.org/10.1016/j.cortex.2012.09

0010-9452/$ e see front matter ª 2012 Elsevhttp://dx.doi.org/10.1016/j.cortex.2012.09.003

a b s t r a c t

Motor activations reported during action-word understanding have raised the question as

to whether the system for motor production contains semantically-relevant information.

Cognitive neuropsychologists have provided compelling evidence that damage to the

system for production of object-directed (transitive) actions does not necessarily lead to

detrimental changes in the individuals’ ability to understand the corresponding action

words, and vice versa. We addressed this question focusing on intransitive symbolic

gestures (emblems; e.g., waving goodbye), which are known to engage different resources,

or neural representations, than object-directed actions, and are thought to enjoy a special

relationship with language, due to a lexicalized relation between form (the gesture) and its

meaning. We tested 12 left-damaged patients (and 17 healthy controls) on praxis (imitation

and gesturing-to-verbal-command) and lexicalesemantic tasks (naming and wordepicture

matching) involving the same emblems. With the group-level analyses, we replicated

correlations between praxis and language deficits typically observed in left-damaged

patients. The analyses of patients’ performance at the single-case level, however,

revealed double dissociations between the ability to produce emblems and the ability to

retrieve and recognize their lexicalesemantic definition. Double dissociations, even in the

event of positive group-level correlations across tasks, imply that the motor representation

of a gesture and the lexicalesemantic representation of the corresponding word rely on

functionally independent system. This study is the first systematic neuropsychological

investigation of the relationship between the lexicalesemantic and the motor represen-

tation of emblems, the closest counterpart of words in the gestural domain.

ª 2012 Elsevier Srl. All rights reserved.
112113 114 115116 1. Introduction et al., 2009, 2011), and facilitates motor behavior (Glenberg 117118119120121122123124
A number of empirical phenomena suggest that the neural

systems for language understanding and action production

are closely interactive: understanding action-related words

correlates with activity in fronto-parietal motor regions (Hauk

et al., 2004; Tettamanti et al., 2005; Tomasino et al., 2007),

enhances corticospinal excitability (Oliveri et al., 2004; Papeo

epartment of Psycholog

u, [email protected]

L, Rumiati RI, Lexical a.003

ier Srl. All rights reserved

and Kaschak, 2002; Scorolli and Borghi, 2006; Zwaan and

Taylor, 2006).

A key question is whether the system for action production

contains semantically-relevant information necessary for

action-word understanding. A current popular view is that

understanding words such as grasping relies on the basic

ability to perform the corresponding physical action, which

y, Harvard University, William James Hall, 33 Kirkland Street,

m (L. Papeo).

125126127128129130

nd gestural symbols in left-damaged patients, Cortex (2012),

.

mailto:[email protected]

mailto:[email protected]

www.sciencedirect.com/science/journal/00109452

www.elsevier.com/locate/cortex

http://dx.doi.org/10.1016/j.cortex.2012.09.003



Q3

c o r t e x x x x ( 2 0 1 2 ) 1e1 12

131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195

196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260


makes available the internal simulation of that action during

conceptual processing: motor representations sustaining

action performance would thus be a central part in the lex-

icalesemantic representation of action-related words

(Rizzolatti and Arbib, 1998; Rizzolatti and Craighero, 2004).

This view predicts that damage to the system for action

production should lead to detrimental changes in the patients’

ability to understand action words.

However, cognitive neuropsychologists have shown that

damage to the left mouth/hand/foot sensorimotor cortices

could leave unaffected the patients’ ability to recognize verbs

and nouns related to mouth/hand/foot actions (Arevalo et al.,

2012). Moreover, large-scale studies of patients’ brain lesions

revealed no single case with focal damage to motor/premotor

cortices and impaired lexicalesemantic processing of action-

words (Kemmerer et al., 2012). Along this line, in a recent

multiple-single case study, we found that left-damaged

patients could have normal comprehension of action-verbs

and tool-nouns, even though they had lost the ability to

imitate the implied actions and use tools (Papeo et al., 2010;

see Papeo and Hochmann, 2012, for a more extensive review).

In that study (Papeo et al., 2010), we tested patients’ ability to

understand and produce meaningful object-directed actions.

The focus on object-directed actions, in ours aswell as inmost

studies in the field, is implicitlymotivated by the proposal that

the motor substrates recruited during conceptual tasks

encode actions not just in terms of means (i.e., specific motor

sequences), but primarily in terms of goals (Johnson Frey et al.,

2003; Rizzolatti and Craighero, 2004).

Object-directed actions perhaps represent the most

obvious category of goal-directed actions. Emblems such as

“thumbs up” to mean “OK” too, fall in the category of mean-

ingful goal-directed actions: these are symbolic, culturally-

defined gestures that, while not related to a physical object,

are nevertheless directed to a (communicative) goal, as they

are intentionally used in social interaction to convey mean-

ings and evoke behavioral responses in other individuals

(Frey, 2008). In this perspective, emblems meet the criteria of

the actions that are held to inducemotor resonance, or activity,

in the human brain.

On the other hand, considering their praxis features,

object-directed actions and emblems fall into the different

categories of transitive and intransitive (non-object-direct)

gestures, respectively, which might differ in terms of repre-

sentational properties (e.g., one is constrained by the object-

features and the other by the socio-cultural context) and

procedures for production (Bartolo et al., 2003; Cubelli et al.,

2000; Mozaz et al., 2002; Ochipa et al., 1989), and/or in terms

of the implicated cognitive resources (Carmo and Rumiati,

2009; Kroliczak and Frey, 2009). This observation leaves open

a possibility that neuropsychological results on object-

directed actions might not be readily generalized to emblems.

Emblems also differ from other intransitive gestures, such

as co-speech (i.e., gestures that spontaneously accompany

speech production; Goldin-Meadow, 1999), as they have

meaning independent of speech and can occur on their own

(Ekman and Friesen, 1969); they are symbols, in that their

meaning results from a conventional and arbitrary relation

between form (the sign) and referent. These properties assign

to emblems a language-like aspect, to the extent that they may

Please cite this article in press as: Papeo L, Rumiati RI, Lexical ahttp://dx.doi.org/10.1016/j.cortex.2012.09.003

be represented in amental lexicon (McNeill, 1992), and encoded

in a way analogous to word recognition. For instance, pro-

cessing meaningful and meaningless emblems elicits

a difference in the event-related potential component N400,

analogous to the electrophysiological correlate of the

distinction between words and pseudowords (Gunter and

Bach, 2004; see also Wu and Coulson, 2005). A magneto-

encephalography study showed that the processing of

emblems involves two stages, at w230 and 370 msec, remi-

niscent of the lexical-access andmeaning-selection stages for

word recognition (Nakamura et al., 2004). Further, activity in

the classic perisylvian language network (inferior frontal and

posterior temporal cortices) has been found during the

observation of communicative symbolic gestures (Xu et al.,

2009). These circumstances suggest that emblems could

enjoy a special relationship with language (McNeill, 1992).

This view is emphasized in some evolutionary accounts of

language, whereby the brain system for manual communi-

cation is regarded as the direct precursor of the speech

architecture (Gentilucci and Corballis, 2006; Rizzolatti and

Arbib, 1998).

We reasoned that, if the system for action production is

part of the system maintaining the lexicalesemantic defini-

tion of gestures (i.e., action-words), the link between the two

systems could be particularly strong in the case of emblems,

as they share more properties with words, relative to other

gestural categories. Using the cognitive neuropsychology

approach, we tested whether damage to the mechanism for

emblem production necessarily results in damage to the lex-

icalesemantic representation of the corresponding words.

While a parallel between symbolic gesturing and language

has been widely documented in infants’ development (Bates

and Dick, 2002; Hill, 2001), neuropsychology to date has not

yet provided a clear contribution to this enterprise. Available

studies involved groups of aphasics (individuals with

language dysfunction) showing either weak association

between the conceptual processing of symbolic gestures and

the ability to reproduce them (Gainotti and Ibbia, 1972;

Gainotti and Lemmo, 1976), or strongly correlated perfor-

mances on verbal and gestural communication competence

(e.g., auditory language comprehension and production of

conventional gesture; Wang and Goodglass, 1992), attributed

either to the severity of aphasia (Glosser et al., 1986) or to the

general loss of intellectual efficiency (Goodglass and Kaplan,

1963).

In the current study, we tested 12 consecutive left-

damaged patients (and 17 healthy controls) on praxis and

language tasks, involving the very same emblems. Praxis was

tested by presenting a set of emblems for imitation and verbal

commands to trigger execution; the same emblemswere used

for a naming task and the corresponding words (i.e., verbs)

were presented for wordepicture matching. Different

modalities of stimulus presentation and response (visual

stimuli for naming and spoken words for recognition) allowed

us to assess the semantic level of word representations

(Caramazza and Shelton, 1998;Warrington and Shallice, 1984).

Patients’ performance on praxis and language tasks was

analyzed both at the group-level, to test whether it was

possible to reproduce the correlations, commonly reported in

groups of left-damaged patients, between aphasic and apraxic




c o r t e x x x x ( 2 0 1 2 ) 1e1 1 3

261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325

326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364


deficits (i.e., disorders in the execution of purposeful actions),

and at the single-case level, to investigate potential dissocia-

tions in those patients whose performance might be opposite

to, but masked by, the group-level trend (Caramazza, 1986). In

the latest years, the multiple single-case approach has earned

new impetus, as it takes into account individual perfor-

mances, while it allows replicating significant observations

within the same sample. A difference in performance on two

tasks, however, might reflect a bias due to one task being, for

any reason, more demanding than the other. To exclude task-

specific bias in the interpretation of patients’ performance, it

is important to find a Case 1, who performs pathologically on

task A and normally (classic dissociation) or significantly

better (strong dissociation) on task B, and Case 2 who shows

the opposite pattern. This double dissociation provides solid

evidence that the abilities underlying the two tasks are

specialized to functionally independent systems (Shallice,

1988).

Thus, taking advantage of the cognitive neuropsychology

methodology, we assessed whether the motor representation

of an emblem is a necessary substrate not only for the

production of that act but also for the representation of the

lexical unit expressing its meaning; in other words, whether

the system for motor production is the locus of overlap

between language and action. A positive answer to this

question requires demonstrating that not only deficits in

language and praxis tasks correlate at the group-level, but also

that each single patient with impaired emblem performance,

has also impaired processing of the lexicalesemantic defini-

tion of emblems (i.e., verbs), and vice versa. Alternatively,

should a double dissociation be found (even in the event of

a positive group-level correlation across tasks), this would be

the demonstration that the lexicalesemantic processing of

action-related words is independent of the state (damaged or

intact) of the system for the production of the implied-

language gestures.

1 The mean score for age of acquisition was 1.8. Ratings for ageof acquisition were collected in a sample of 19 healthy partici-pants (11 female, mean age 25 years), on a 7-point Likert scalewith 1 corresponding to acquisition within the third year of life,2 between the fourth and fifth year of life, and so on up to 7,corresponding to acquisition after 13 years.

365366367368369370371372373374375376377378379380381382383384385386387388389390

2. Material and methods

2.1. Participants

The experimental group included patients, with a single focal

unilateral lesion to the left-hemisphere, >5 years of educa-

tion, right-handed (Oldfield, 1971), native Italian speakers,

with normal or corrected-to-normal vision, no hearing diffi-

culties, and psychiatric history. All patients were clear of any

sign of cognitive decline due to progressive neuro-

degenerative disorders, or other premorbid neurological

conditions, including previous cerebrovascular accident.

Twelve consecutive patients (mean age 70 � 8 years, mean

education 10 � 4 years) who met these criteria were included

in the study in order to obtain a random sample, representa-

tive of the population of interest.

Patients were administered a full neuropsychological

screening including standardized tests to evaluate aphasia,

apraxia, visuo-spatial abilities, visual recognition, executive

functions and memory. Importantly, given that in the current

study visual stimuli were employed, patients performed

within the normal range on the screening test of the Visual


Object and Space Perception (VOSP) battery (Warrington and

James, 1991), assessing visual sensory efficiency (the score

on this test is not available for Case 1, as the evaluation had to

be interrupted for accidental reasons). Table 1 summarizes

the demographic variables and the scores obtained by each

patient in each screening test. The same group of patients was

involved in a previous independent study by Papeo et al.

(2010), to which we refer for a description of their lesional

sites.

Seventeen healthy adults, matched for age (mean 69 � 8

years) and education (mean 12 � 4 years), took part in the

study as the control group. All controls were clear of any sign

of cognitive decline (Mini Mental State Examination; Folstein

et al., 1975).

All participants confirmed their voluntary participation

signing the informed consent. The study was approved by

SISSA Ethics Committee.

2.2. Experimental design and materials

Stimuli were selected with a norming study including two

phases. First, 45 healthy adults (age 21e86 years, education

5e18 years) were asked to name 72 color video-clips of 3 sec

each, in which a male actor performed emblems. Second,

naming latencies were collected in a group of 30 new healthy

participants (age 21e41 years), for those items that in the

former phase were named with the same verb by at least the

85% of participants. We thus selected 15 emblems with high-

conventional meaning consistently (>85% of agreement;

mean % of agreement: 91%) and readily (>3 sec) denoted by

a verb (Fig. 1). Another verb assigned to an item by at least 5%

of the panel was taken as an alternative correct response in

the experimental phase (e.g., “sentire” was the acceptable

alternative for “ascoltare”, both verbs meaning “to listen”).

Lexical items (verbs) corresponding to the 15 selected

emblems had mean oral frequency of 87.8 (oral word

frequencywas taken from a corpus of Italian words; DeMauro

et al., 1993), mean length of 7.9 phonemes, and were acquired,

on average, within the fourth year of life.1

The 15 video-clips of the emblems were presented for the

imitation and naming task. The most salient frame of each

video-clip was presented for the verb recognition task

(wordepicture matching), in the form of color photograph,

together with two distractor photographs, depicting a gesture

semantically related and a gesture visually similar to the

target. The visual distractor was obtained by modifying

a spatial feature in the target gesture, such as the hand/arm

orientation. In the wordepicture matching task, photographs

were preferred to video-clips to enable the simultaneous

presentation of the three items, and thus to reduce the load on

participants’ working memory. The three photographs

appeared on the screen while the name of the target (i.e., the

verb denoting the emblem’s meaning) was spoken aloud by




Table 1 e Demographic variables and patients’ scores on the neuropsychological evaluation. Q6

Part 1

Case Sex Age Education(years)

Testingpost-onset(months)

EHI LTMwords

LTMfaces

Spanforward

Spanbackward

Corsi TMT A TMT B Weigl Raven’sCPM

C1 M 72 18 3 100 N.a. 10 e e 4 e e T.i. T.i.

C2 M 60 8 1.5 100 43 21 4 4 4 36 144 10 32

C3 M 80 8 1.5 100 32 23 6 3 5 92 T.i. 7 24

C4 M 77 17 1 100 N.a. e e N.a. N.a. 194 e 3 20

C5 F 62 10 1 100 39 e 5 5 4 47 106 13 29

C6 F 58 13 11 100 35 25 e e 4 48 76 13 31

C7 F 71 8 2 100 34 24 e e 4 117 T.i. 5 18

C8 M 70 8 2 62 31 20 3 2 5 45 285 4 e

C9 F 75 5 1 100 35 21 5 2 4 125 N.a. 3 11

C10 F 83 8 1 100 36 19 4 2 4 89 T.i. 6 13

C11 F 64 8 7 100 28 21 N.a. N.a. 6 94 N.a. 9 28

C12 M 71 6 4 83 29 9 4 N.a 3 N.a. N.a. 5 19

Maximum

score

100 e e e e e e e 15 36

Part 2

Case AATtoken

AATrepetition

AATwrittenlanguage

AATnaming

AATcomprehension

Lexicaldecision

Reading Picturenaming

VOSPscreen

VOSPo.d.

IMA IA Type ofaphasia

C1 46 66 10 23 49 T.i. N.a. N > V T.i. 9 60 10 Severe global

C2 15 127 85 94 110 144 109 N > V 20 20 64 14 Mild amnesic

C3 14 139 82 109 109 134 114 N > V 20 18 70 14 Severe unfluent Q7

C4 50 N.a. N.a. N.a. 22 122 N.a. N.a. 19 12 31 7 Severe

Wernicke’s

C5 2 149 90 117 117 144 116 N > V 19 18 58 14 e

C6 23 111 77 95 111 138 110 V > N 20 16 66* 14 Mild anomic

C7 28 65 27 60 93 128 23 V¼N 20 16 43 11 Mild Broca’s

C8 17 110 73 100 86 142 108 N > V 20 12 66 14 Mild Broca’s

C9 20 139 37 93 90 126 108 N > V 18 13 52 10 e

C10 5 143 62 105 100 141 115 N > V 20 16 46 9 Mild amnesic

C11 28 137 53 69 93 125 100 N.a. 20 13 e e Severe Broca’s

C12 29 132 31 93 84 101 65 N > V 20 17 48 14 Mild Broca’s

Maximum

score

50** 120 90 120 120 144 116 e 20 17 72 14

Standardized tests used in the neuropsychological assessment: EHI¼ Edinburgh Handedness Inventory (Oldfield, 1971); LTMwords¼ long-term

memory, word recognition (Warrington, 1984); LTM faces ¼ long-term memory, face recognition (Warrington, 1996); Span forward ¼ digit span

forward for verbal short-term memory; Span backward ¼ digit span backward for short-term memory and working memory; Corsi ¼ Corsi test

for spatial short-term memory (Spinnler and Tognoni, 1987); TMT A ¼ Trail making tests for attention; TMT B ¼ Trail making test for executive

functions (Giovagnoli et al., 1996); Weigl ¼ Weigl’s tests for executive functions (Spinnler and Tognoni, 1987); Raven’s CPM ¼ Raven Colored

Progressive Matrices for general intelligence (Carlesimo et al., 1996); AAT ¼ Aachener Aphasie Test, Italian norms (Luzzatti et al., 1986); AAT

token ¼ subtest for comprehension; AAT repetition ¼ subtest for repetition; AAT written language ¼ subtests for reading and writing; AAT

name ¼ subtest for production; AAT comprehension ¼ subtests for auditory and visual comprehension; Lexical decision (Luzzatti et al., 2002);

Reading (Toraldo et al., 2006); Picture naming ¼ object and action picture-naming (Crepaldi et al., 2006); VOSP screen ¼ Visual object and space

perception, screening task; VOSP o.d. ¼ VOSP object decision task (Warrington and James, 1991); IMA ¼ Test for ideomotor apraxia (Tessari and

Rumiati, 2004), * ¼ test by De Renzi et al. (1980); IA¼ Ideational Apraxia (De Renzi and Lucchelli, 1988). Patients are sorted alphabetically by their

initials. F ¼ female. M ¼ male. T.i. ¼ test interrupted because the patient was not able to perform the task. N.a. ¼ test not administered.

Pathological scores are reported in bold. Maximum scores are reported where applicable (** Maximum error).

c o r t e x x x x ( 2 0 1 2 ) 1e1 14

391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455

456457458459460461462463464465466467468469470471472473474475476477478479480481482483484485486487488489490491492493494495496497498499500501502503504505506507508509510511512513514515516517518519520


the experiment. Fifteen verbal commands (the verbs corre-

sponding to the target emblems) were spoken by the experi-

menter to trigger the execution of emblems in the gesturing-

to-verbal-command task.

Naming and pictureeword matching tasks were both

used to examine the lexicalesemantic representation of

emblems independent of the specific input and output

modality of the task (Caramazza and Shelton, 1998; Shallice,

1988; Warrington and Shallice, 1984). This approach is

particularly warranted as a failure in naming might be due


to a breakdown at the level of word processing other than

lexicalesemantics (e.g., articulatory, lexical or phonological

retrieval). Therefore, pathological naming performance was

considered here as a reliable indication of lexicalesemantic

deficit, only when accompanied by a failure in performing

the pictureeword matching task. The same logic applies to

the praxis tasks: failure in gesturing-to-verbal-command,

a task critically involving auditory verbal comprehension

component, was taken as a sign of apraxia only when

corroborated by failure on imitation.




Fig. 1 e The selected 15 emblems used as experimental stimuli. These emblems correspond to the following lexical

definitions (from the upper line, from left to right): dormire (to sleep), contare (to count), telefonare (to phone), camminare (to

walk), alzare (to lift), applaudire (clapping), andare via (to leave), sentire (to hear), fermare (to stop), mangiare (to eat), pregare (to

pray), salutare (to salute), salutare (waving), sparare (to shout), tagliare (to cut).

c o r t e x x x x ( 2 0 1 2 ) 1e1 1 5

521522523524525526527528529530531532533534535536537538539540541542543544545546547548549550551552553554555556557558559560561562563564565566567568569570571572573574575576577578579580581582583584585

586587588589590591592593594595596597598599600601602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649650


Notice also how employing the same stimuli across

different tasks minimized the variability due to the stimuli, in

participants’ performance on different tasks. This method

also allowed controlling for the possible confound of concur-

rent deficits at the level of input processing: indeed, the same

perceptual abilities were required to process the visual stimuli

(video-clips) in the naming task and in the imitation task (and

the photographs in the pictureeword matching task), and to

process auditory verbal inputs (verbs) given for wordepicture

matching and gesturing-to-verbal-command.

2.3. Procedures

The four experimental tasks (see below) were presented in

counterbalanced order across participants, while the order of

trials in each task was kept fixed. Five controls performed the

praxis tasks with the non-dominant-left hand to match them

with the patients’ group including three caseswith right-sided


hemiparesis due to left-hemisphere injury. All other patients

and controls performed using the dominant-right hand.

2.3.1. Praxis tasksVideo-clips of emblems were presented, one at a time, on

a computer screen placed in front of the participant who was

instructed to imitate the gesture as soon as the clip ended. If

the participant failed to imitate, the clip was shown again for

a maximum of two times. In the gesturing-to-verbal-

command task participants were given the verbal cue to

produce the emblems.

During both tasks, participants’ were videotaped for off-

line qualitative error analysis carried out by one author (LP)

and two neuropsychologists, unaware of the hypotheses of

the study. Each trial was judged as “correct” or “incorrect” by

the three raters, independently. Trails judged as “correct” by

at least two raters were assigned 2 points; the errors in

“incorrect” trials (i.e., trials judged so by at least two raters)




Q4

Table 2 e Results of the correlational analysis computedover the t scores.

Imitation Naming Recognition

Gesturing-to-verbal-

command

.79a .92a .88a

Imitation .74a .52

Naming .76a

a Correlation is significant at <.01 level.

c o r t e x x x x ( 2 0 1 2 ) 1e1 16

651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708709710711712713714715

716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767768769770771772773774775776777778779780


were classified according to the criteria described in Rumiati

et al. (2001) and adopted in previous studies (Negri et al.,

2007; Papeo et al., 2010; Tessari et al., 2007). One point was

assigned when the participant made spatial errors (i.e.,

misorientation of hand/arm) but the action was clearly

recognizable; “0” was given in case of semantic error (body-

part use as a tool, or substitution with a semantically-related

action), visual error (substitution with a visually-related

action), omission, or unrecognizable gesture. Therefore, each

trial could be assigned 0, 1 or 2 points, the maximum score of

each task being 30. All emblems, but two (clapping and pray-

ing), were unimanual gestures. In hemiparetic patients, the

two bimanual gestures were assessed based on the execution

with the effective left-hand. As both clapping and praying

involve identical and symmetrical movements of the two

hands, the unimanual performance could provide reliable

indication of whether the correct motor representation was

being retrieved and produced.

2.3.2. Language tasksIn the naming task, participants were instructed to name the

emblems showed in thevideo-clips appearing, one at a time, on

a computer screen. Correct responses, including self-repairs,

dialect forms and phonological errors in which the target was

clearly recognizable, were scored “1”. Incorrect responseswere

scored “0”. Semantic paraphasias, circumlocutions and laten-

cies longer than 5 secwere consideredas errors. Themaximum

score was therefore 15. In the wordepicture matching task,

each verb was spoken aloud by the experimenter, while three

color photographs depicting the same actor performing the

target-emblem and two distracters, appeared on the computer

screen. The relative positione left, center, right of the screene

of target and distracters was counterbalanced across trials.

Participants were asked to point at the photograph referring to

the spoken verbmeaning. Correct (target) and incorrect (either

distractor) responses were scored “1” and “0”, respectively, for

a maximum total score of 15.

2.4. Analysis

At the group-level, we compared the performances of the

patients against those of controls in each task (each individual

performance was expressed as the percentage of correct

responses), and computed correlations between patients’

performances on praxis and linguistic tasks. As patients’ data

in the four tasks were not normally distributed (Shapiroe

Wilk’s W-test, ps > .05), non-parametric tests were used

(ManneWhitney U test for between-groups comparisons and

Spearman’s correlations).

To test dissociations, a multiple single-case analysis was

carried out using the Revised Standardized Difference Test

(RSDT, Crawford and Garthwaite, 2006). In the RSDT, indi-

vidual scores in each task are transformed into t scores. The

RSDT determines: (1) the abnormality of each t score defined

as the percentage of the normal population that would obtain

a lower score, and (2) whether a significant difference between

two t scores obtained in two different tasks, reflects classical

dissociation or strong dissociation (Crawford et al., 2003;

Shallice, 1988). The computation implemented in the RSDT

considers directly the controls’ performance on two tasks


(means and standard deviations) and the correlation between

them, in order to ensure that the difference between two tasks

giving rise to dissociation is significantly higher than it could

be randomly observed in the normal population. The revised

version of the Crawford and Garthwaite’s method, used here,

is particularly appropriate for the analysis of small-sized

samples. With this inferential method, we compared single

patient’s performance in each praxis task (imitation and

gesturing-to-verbal-command) with her/his performance in

each language task (naming and word recognition).

3. Results

Group-level. Relative to controls, patients were impaired in all

four experimental tasks. The difference between the two

groups was significant in gesturing-to-verbal command

[mean percentage of correct responses � SEM: patients

(Mp) ¼ 57.5 � 9.1, controls (Mc) ¼ 88.4 � 2.7, ManneWhitney

U ¼ 21.5, p < .001], imitation (Mp ¼ 76.1 � 5.4, Mc ¼ 88.4 � 2.7,

ManneWhitney U ¼ 31, p ¼ .001), and verb recognition

(Mp ¼ 82.7 � 5.2, Mc ¼ 98.8 � .9, ManneWhitney U ¼ 33,

p ¼ .001). While, on average, patients performed worse than

controls in naming too (Mp ¼ 65.5 � 10.5, Mc ¼ 89 � 2.3), this

difference did not reach significance ( p > .05).

The correlational analysis carried out over the percentages

of correct responses, revealed significant correlations

between patients’ performance on the two praxis tasks

(gesturing-to-verbal command and imitation, N ¼ 12,

Spearman R ¼ .81, p ¼ .001), and on the two linguistic tasks

(naming and wordepicture matching, N ¼ 12, Spearman

R ¼ .75, p < .01). More importantly, patients’ performance on

naming positively correlated with their performance in both

gesturing-to-verbal-command (N ¼ 12, Spearman R ¼ .93,

p > .0001) and imitation (N ¼ 12, Spearman R ¼ .74, p < .01).

Patients’ performance on wordepicture matching correlated

with their ability to produce gestures to verbal command

(N ¼ 12, Spearman R ¼ .88, p ¼ .0001), and showed a trend

toward a positive correlation with the performance on imita-

tion (N ¼ 12, Spearman R ¼ .53, p ¼ .07). The very same results

were replicated when correlations were performed over

Crawford’s t scores (Table 2).

Single-case level analysis (Fig. 2). Two patients (C7 and C10)

performed significantly better on tasks assessing lex-

icalesemantic processes than on praxis tasks. C10 exhibited

a deficit in both imitation and naming, although he was

significantly more impaired on the former than on the latter

(strong dissociation). Consistently with this observation, C10

performed qualitatively better on naming than on gesturing-




Fig. 2 e (A) Double dissociation between gesture naming

and imitation (significant difference at the RSDT; Crawford

and Garthwaite, 2006: C1 and C4, ps < .0001; C10,

p [ .001). (B) Double dissociation between word

recognition (wordepicture matching) and imitation

(significant difference: C1, C4, ps < .0001; C3, C7, C9,

ps £ .001; C10, p [ .01). (C) Double dissociation between

word recognition and gesturing-to-verbal-command

(significant difference: C3, C7, ps < .05; C9, p [ .0001).

Beside patients’ initials: *indicates classical dissociation,

**indicates strong dissociation.

2 Spearman’s rank correlation coefficient for each pair of tasks:naming versus gesturing-to-verbal-command, N ¼ 15, SpearmanR ¼ .34, p ¼ .2; naming versus imitation, N ¼ 15, Spearman R ¼ �.04, p ¼ .8; wordepicture matching versus gesturing-to-verbalcommand, N ¼ 15, Spearman R ¼ .22, p ¼ .4; wordepicturematching versus imitation, N ¼ 15, Spearman R ¼ .02, p ¼ .09.

c o r t e x x x x ( 2 0 1 2 ) 1e1 1 7

781782783784785786787788789790791792793794795796797798799800801802803804805806807808809810811812813814815816817818819820821822823824825826827828829830831832833834835836837838839840841842843844845

846847848849850851852853854855856857858859860861862863864865866867868869870871872873874875876877878879880881882883884885886887888889890891892893894895896897898899900901902903904905906907908909910


to-verbal command, but this difference did not reach signifi-

cance. Moreover, C10 was more impaired in imitation also

relative to wordepicturematching (strong dissociation). In the

same direction, C7 exhibited a classical dissociation: her

ability to recognize verbs (in the wordepicture matching) was

normal and significantly better than her pathological ability to

imitate emblems as well as to produce emblems after verbal


command. Notice that, in this patient, naming was as

impaired as the two praxis tasks (i.e., performances across

tasks were not significantly different). The absence of word

comprehension deficit in this patient proves that impaired

naming performance reflected damage to a level of word

processing, different than the semantic one, and that

impaired gesturing-to-verbal command reflected a genuine

apraxic deficit.

In four patients (C1, C3, C4, C9), we found that praxis was

spared or significantly less pathological than lexicalesemantic

abilities. C1 and C4 exhibited a strong dissociation, with

naming being significantly more impaired than imitation.

These two patients, for whom a less severe impairment on

gesturing-to-verbal command relative to naming would be

expected, could not be tested on the former task due to their

inability to comprehend verbal commands (i.e., the task had to

be interrupted after a few trials). The severe verbal compre-

hension disorder of these two patients was confirmed by the

following finding: both C1 and C4 exhibited a strong dissocia-

tion between imitation and wordepicturematching, the latter

being significantly more impaired than imitation. The same

dissociation was found in C9 (strong dissociation) and in C3,

who performed normally on imitation and pathologically on

wordepicture matching (classical dissociation). Consistent

results were found when verb recognition was compared with

gesturing-to-verbal-command in C9 (strong dissociation) and

C3 (classic dissociation). In these two patients, naming

performance did not differ from praxis.

The same comparisons were carried out for all the

remaining cases that are not discussed here, as their behavior

did not give rise to any dissociation. Table 3 reports the score

of all patients in all experimental tasks.

3.1. Additional analysis and results

Ourhypothesis that apatientwith impairedpraxis abilities can

still process successfully (or less pathologically) action-related

words, and vice versa, implies that the lexicalesemantic defi-

nitionofanemblemcanbespared, even if theability to execute

that emblem is damaged. Double dissociations already suggest

that this item-related hypothesis holds true: in fact, no double

dissociation would be found, if each item that cannot be

imitated, cannot be lexically processed either. This stancewas

further confirmed in a by-tem analysis, where we derived

rankings of item difficultly across the four tasks, and then

evaluated the relationship between the two ranked series of

items in each critical pair of tasks (for a total of four correla-

tions), by computing the Spearman’s rank correlation coeffi-

cient. Rank order values of item difficulty were defined

according to the number of patients who hit the target in each

task. No correlation was found to be significant (all ps > .05),2

confirming that items with lower/higher successful perfor-

mance were not the same across tasks.




Table 3 e Individual performance (t scores; Crawford andGarthwaite, 2006) of all patients across all experimentaltasks.Q8Q9

Gesturing-to-verbal-command

Imitation Naming Wordepicture

matching

C1 T.i. L4.074 L8.919 L12.559

C2 �.155 �.73 1.1 .326

C3 �.741 .704 .432 L3.354

C4 T.i. L3.118 L8.919 L14.398

C5 �.448 �.73 .432 .326

C6 �.155 .226 .432 .326

C7 L3.96 L5.984 L4.911 �1.513

C8 �.741 �1.684 1.1 �1.513

C9 L3.84 L3.596 �1.571 L8.876

C10 L4.548 L8.374 L4.243 L5.196

C11 L2.205 L4.074 L2.239 L3.354

C12 �1.033 �1.207 �.904 L3.354

Note: Numbers in bold denote t scores significantly below the

controls’ mean.

c o r t e x x x x ( 2 0 1 2 ) 1e1 18

911912913914915916917918919920921922923924925926927928929930931932933934935936937938939940941942943944945946947948949950951952953954955956957958959960961962963964965966967968969970971972973974975

976977978979980981982983984985986987988989990991992993994995996997998999

100010011002100310041005100610071008100910101011101210131014101510161017


Moreover, while we argue that dissociations were the

result of the interaction between the patients’ cognitive

profiles and the task tapping eithermotor or lexicalesemantic

representations related to emblems, we are assuming that the

results were not affected by the specific items selected for the

study. To demonstrate this, we computed the variance of

patients’ responses (expressed as 0 for incorrect and 1 for

correct responses) that was due to the items, using a multi-

level logistic regression. Since our 15 items were assumed to

be randomly selected from a population of virtually infinite

possible levels, we treated “items” as a random effect, hier-

archically nested within the “patient” and “task” factors (fixed

effects), to take into account the non-independence of the

observations. This analysis, applied to the four pairs of tasks

that gave rise to dissociations, showed that the variance of

patients’ responses was only marginally (on average, 4.6%)

explained by the “Item” factor.3

4 As already observed in the “Results” section, the apraxicdeficit of C10 could not be appreciated in the gesturing-to-verbal-command task, as his deficit in verbal comprehension was suchthat he could not understand the commands. In C7, while word

10181019102010211022102310241025102610271028102910301031103210331034

4. Discussion

Empirical evidence for motor activations during conceptual

tasks, such as action-word understanding, has raised the

question as to whether the system for motor production

contains semantically-relevant information. We addressed

this question by investigating whether deficits in gesture

production after left-brain injury were necessarily accompa-

nied by changes in the patients’ ability to understand words

denoting the impaired gestures, and vice versa. In particular,

this study provides the first systematic neuropsychological

investigation of the relationship between the lex-

icalesemantic and the motor representation of emblems,

3 Type I mean squares estimates of the variance componentsfor items in (1) naming versus gesturing-to-verbal-command:.012, random error .12; (2) naming versus imitation: .004,random error .16; (3) wordepicture matching versus gesturing-to-verbal command: .007, random error .13; and (4) wordepicturematching versus imitation: .002, random error .16.


a category that is regarded as the closest counterpart of words

in the gestural domains.

The behavior of 12 left-damaged patients was analyzed in

emblem imitation and gesturing-to-verbal-command (praxis

tasks) and in emblem naming and emblem-related verb

recognition (lexicalesemantic tasks). At the group-level,

naming correlated with both imitation and gesturing-to-

verbal-command, in line with the typical association of

apraxic and aphasic symptoms in left-damaged patients.

Wordepicture matching (word recognition) correlated with

gesturing-to-verb-command but not significantly with imita-

tion behavior. The closer relationship of wordepicture

matching with gesturing-to-verbal-command, than with

imitation could in the first place reflect the fact that gesturing-

to-verbal-command also involves auditory comprehension.

However, auditory comprehension cannot fully account for

the patients’ gesturing-to-verbal-command: in effect, their

performance on this task tightly correlated with their imita-

tive ability, and all patients with impaired gesturing-to-

verbal-command were also impaired in imitation (see

Table 3), which does not involve auditory comprehension. The

lack of correlation (although there was a trend in the direction

of a positive correlation) between wordepicture recognition

and imitation could reflect the limited power of our analysis

including a small-size sample, although even larger-scale

studies sometimes failed to report correlation between

praxis and semantic tasks (e.g., Tessari et al., 2007).

While it cannot be unequivocally established why two

functions correlate or do not, as any third factor can be

responsible for this outcome, double dissociations provide

solid evidence that they are served by independent systems or

subsystems (Shallice, 1988). Our single-case analysis revealed

six performance profiles that generalize to the domain of

communicative symbolic gestures (emblems), previous

evidence of dissociation between semantic and praxis pro-

cessing, investigated on object-directed actions (Negri et al.,

2007; Papeo et al., 2010).

Two cases of apraxia were observed with spared (C7) or

significantly less impaired (C10) word recognition: C10 was

impaired in imitation more than in the two lexicalesemantic

tasks; and C7 was impaired in the two praxis tasks and had

normal word recognition (wordepicturematching).4 Four cases

of lexicalesemantic deficitwith spared (C9) or less impaired (C1,

C3 and C4) praxis abilities were found. While C1 and C4’s

semantic deficit was evident both in naming and word recog-

nition (as well as in gesturing-to-verbal-command, where they

failedtocomprehend thecommands),C3andC9’swereseverely

impaired inverb recognition (wordepicturematching), but their

naming ability was not worse than their praxis. While they tap

comprehension was entirely spared, her naming ability was notsignificantly different than her pathological praxis, this patternsuggesting a damage to a level of word processing, different thanthe semantic one (e.g., articulatory or lexicalephonologicalretrieval; e.g., Berndt et al., 1997; Caramazza, 1997). Theseobservations emphasize the importance of assessing the samerepresentation through different testing modalities.

103510361037103810391040




c o r t e x x x x ( 2 0 1 2 ) 1e1 1 9

10411042104310441045104610471048104910501051105210531054105510561057105810591060106110621063106410651066106710681069107010711072107310741075107610771078107910801081108210831084108510861087108810891090109110921093109410951096109710981099110011011102110311041105

11061107110811091110111111121113111411151116111711181119112011211122112311241125112611271128112911301131113211331134113511361137113811391140114111421143114411451146114711481149115011511152115311541155115611571158115911601161116211631164116511661167116811691170


the sameconceptual representation, naming andwordepicture

matching differ in many respects. One possible reason for the

advantage of naming over word recognition (wordepicture

matching) in these two patients might be a particular suscepti-

bility to the interference-effect of distracters, that in the latter

task appeared together with the target-emblem (perhaps due to

impaired lexical selection; Levelt et al., 1999). The susceptibility

to “semantic intrusion” isa commonsignofdamage to semantic

processing (Warrington and McCarthy, 1987).

Before considering the theoretical implications of our

results, wewill discuss how our experimental design took care

of confounds that represent a threat in neuropsychological

investigation. In particular, the inter-individual variability is

one major concern, when considering the performances of

single patients. First, there is the possibility that, in some

patients, perceptual deficits could exacerbate the observed

pathological retrieval of motor and/or lexicalesemantic

representations of emblems. However, by using the same

input-stimuli in praxis and lexicalesemantic tasks, we made

sure that a perceptual deficit was not a relevant factor in the

comparison between two abilities in one patient, and between

two patients (e.g., in the case when patient X ismore impaired

in praxis task and patient Y is more impaired in lex-

icalesemantic task).

Second, some patients experienced attentional or execu-

tive control disorders, which could implicate a loss of cogni-

tive resources, more severe than in other patients. This

condition, however, would be expected to impact the perfor-

mance in a general manner, i.e., not differently across tasks,

unless one task is more demanding than another. This

possibility is ruled out by the occurrence of double dissocia-

tions: if one patient performs better on task A than on task B,

and another patient does the opposite, performance differ-

ences cannot be ascribed to one task being significantly more

difficult than the other.

Third, while it is normal that the same individual performs

differently on different tasks, the inferential method we used

to assess dissociations in single cases specifically tests that

the difference between two tasks in a patient (i.e., dissocia-

tion) is significantly higher than it could randomly be

observed in the normal population.

Finally, our patients were diagnosed with different aphasic

syndromes and symptoms. Although the type of aphasia was

not a factor in our experimental design, we can exclude,

a posteriori, that it constituted a bias in our analysis. Broca’s

aphasia was the most common syndrome in our sample (see

Table 1). If any specific aspect of this syndrome critically

affected the relationship between motor and lex-

icalesemantic representations, compatible performances

should have been observed in all Broca’s aphasics; this,

however, was not the case (e.g., C7 was impaired in naming

emblems, but less severely than in praxis tasks; C11 had

pathological lexicalesemantic performance and intact praxis;

see also Broca’s aphasics C8 and C12 with diverse perfor-

mance patterns). Likewise, the nouneverb discrepancy,

a common symptom in aphasics (mostly in favor of noun

retrieval; De Bleser and Kauschke, 2003; see also the result of

the Picture Naming test in Table 1), did not seem to play any

role in the assessed relationship between praxis and language

representations. Otherwise, we would have found analogous


patterns of performance in the eight patients with a poorer

performance on verbs versus nouns; but we did not.

Many circumstances suggest a close link between word

processing and symbolic communicative gesture production.

They are related in normal (Bates and Dick, 2002; Piaget, 1954;

Werner and Kaplan, 1963) and pathological development (Thal

et al., 1997), and in neurological conditions (Kempler et al.,

1995; Wang and Goodglass, 1992; and the current study);

relative to other gestures, the production of communicative

symbolic gestures shows a stronger right-hand bias, taken as

an indication of language-like hemispheric lateralization (Bates

et al., 1986); and the systems for motor control and language

are anatomically connected and possibly overlapping in some

regions (Pulvermuller, 2005; Xu et al., 2009).

The current results neither disclaim any of these obser-

vations, nor the idea that a common substrate might exist for

symbolic gestures and language (McNeill, 1992). They rather

demonstrate that this link is not in the system for action

production: this system can be impaired with no conse-

quences on the individuals’ ability to understand action

words; in other terms, emblem production and word

comprehension are functionally independent. This conclu-

sion opens to alternative hypotheses about the way the two

systems may be linked and the neural locus of their interac-

tion (see e.g., Xu et al., 2009).

Moreover, it is still possible that the system for action

performance contributes to semantic processing: the physical

experiencing of an actionmayhelp infants tomapnovel action-

words onto their referent and, in adults, the retrieval of stored

representations of this experience may enrich the comprehen-

sion of words (Mahon and Caramazza, 2008; Papeo and

Hochmann, 2012). Our results do not exclude that some

nuances of action-words’ meanings have gone lost in apraxic

patients, or that their representation of action meanings was

qualitatively different than in normal individuals, just like an

expert dancer and a neophytemay encode a ballet in a different

way. Yet, whatever information the system formotor execution

conveys during semantic analysis, this is not a core attribute of

action-wordmeanings. In the adult brain, the abilities to trans-

late a visual input into amotor output (imitation) and to retrieve

stored motor programs for execution (gesturing-to-verbal-

command) are not necessary to assign the physical referent to

an action-word and the correct lexical definition to an action.

Acknowledgments

The authors are grateful to Dr. Antonella Zadini for allowing

us to see the patients in her ward, to Pietro Chiarello for his

help in the process ofmaking the experimental stimuli, and to

Jean-Remy Hochmann for his valuable comments on

a previous version of this manuscript.

r e f e r e n c e s

Arevalo AL, Baldo JV, and Dronkers NF. What do brain lesions tellus about theories of embodied semantics and the humanmirror neuron system? Cortex, 48(2): 242e254, 2012.




c o r t e x x x x ( 2 0 1 2 ) 1e1 110

11711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235

12361237123812391240124112421243124412451246124712481249125012511252125312541255125612571258125912601261126212631264126512661267126812691270127112721273127412751276127712781279128012811282128312841285128612871288128912901291129212931294129512961297129812991300


Bartolo A, Cubelli R, Della Sala S, and Drei S. Pantomimes arespecial gestures which rely on working memory. Brain andCognition, 53(3): 483e494, 2003.

Bates E and Dick F. Language, gesture, and the developing brain.Developmental Psychobiology, 40(3): 293e310, 2002.

Bates E, O’Connell B, Vaid J, Sledge P, and Oakes L. Language andhand preference in early development. DevelopmentalNeuropsychology, 2(1): 1e15, 1986.

Berndt RS, Mitchum CC, Haendiges AN, and Sandson J. Verbretrieval in aphasia: 1. Characterizing single wordimpairments. Brain and Language, 56(1): 68e106, 1997.

Caramazza A. How many levels of processing are there in lexicalaccess? Cognitive Neuropsychology, 14(1): 177e208, 1997.

Caramazza A. On drawing inferences about the structure ofnormal cognitive systems from the analysis of patterns ofimpaired performance: The case for single-patient studies.Brain and Cognition, 5(1): 41e66, 1986.

Caramazza A and Shelton JR. Domain-specific knowledgesystems in the brain the animateeinanimate distinction.Journal of Cognitive Neuroscience, 10(1): 1e34, 1998.

Carlesimo GA, Caltagirone C, and Gainotti G. The mentaldeterioration battery: Normative data, diagnostic reliabilityand qualitative analyses of cognitive impairment. EuropeanNeurology, 36(6): 378e384, 1996.

Carmo JC and Rumiati RI. Imitation of transitive and Intransitiveactions in healthy individuals. Brain and Cognition, 69(3):460e464, 2009.

Crawford JR and Garthwaite PH. Methods of testing for a deficit insingle case studies: Evaluation of statistical power by MonteCarlo simulation.CognitiveNeuropsychology, 23(6): 877e904, 2006.

Crawford JR, Garthwaite PH, and Gray CD. Wanted: Fullyoperational definitions of dissociations in single-case studies.Cortex, 39(2): 357e370, 2003.

Crepaldi D, Aggujaro S, Arduino LS, Zonca G, Ghirardi G,Inzaghi MG, et al. Nouneverb dissociation in aphasia: The roleof imageability and functional locus of the lesion.Neuropsychologia, 44(1): 73e89, 2006.

Cubelli R, Marchetti C, Boscolo G, and Della Sala S. Cognition inaction: Testing a model of limb apraxia. Brain and Cognition,44(2): 144e165, 2000.

De Bleser R and Kauschke C. Acquisition and loss of nouns andverbs: Parallel or divergent patterns? Journal of Neurolinguistics,16(2e3): 213e229, 2003.

De Mauro T, Mancini F, Vedovelli M, and Voghera M. Lessico difrequenza dell’italiano parlato. Milano: EtasLibri, 1993.

De Renzi E and Lucchelli F. Ideational apraxia. Brain, 111(Pt 5):1173e1185, 1988.

De Renzi E, Motti F, and Nichelli P. Imitating gestures: Aquantitative approach to ideomotor apraxia. Archives ofNeurology, 37(1): 6e10, 1980.

Ekman P and Friesen WV. The repertoire of nonverbal behavior:Categories, origin, usage, and coding. Semiotica, 1(1): 49e98,1969.

Folstein MF, Folstein SE, and McHugh PR. “Mini-Mental State”: Apractical method for grading the cognitive state of patients forthe clinician. Journal of Psychiatric Research, 12(3): 189e198, 1975.

Frey SH. Tool use, communicative gesture and cerebralasymmetries in the modern human brain. PhilosophicalTransactions of the Royal Society London B Biological Sciences,363(1499): 1951e1957, 2008.

Gainotti G and Ibbia R. La comprensione del significato dei gestisimbolici negli afasici. Minerva Psichiatria e Psicologica, 13: 1e9,1972.

Gainotti G and Lemmo MS. Comprehension of symbolic gesturesin aphasia. Brain and Language, 3(3): 451e460, 1976.

Gentilucci M and Corballis MC. From manual gesture to speech: Agradual transition. Neuroscience & Biobehavioral Reviews, 30(7):949e960, 2006.


Giovagnoli AR, Del Pesce M, Mascheroni S, Simoncelli M,Laiacona M, and Capitani E. Trail making test: Normativevalues from 287 normal adult controls. Italian Journal ofNeurological Sciences, 17(4): 305e309, 1996.

Glenberg AM and Kaschak MP. Grounding language in action.Psychonomic Bulletin & Review, 9(3): 558e565, 2002.

Glosser G, Winer M, and Kaplan E. Communicative gestures inaphasia. Brain and Language, 27(2): 345e359, 1986.

Goldin-Meadow S. The role of gesture in communication andthinking. Trends in Cognitive Sciences, 3(11): 419e429, 1999.

Goodglass H and Kaplan E. Disturbance of gesture andpantomime in aphasia. Brain, 86(4): 703e720, 1963.

Gunter TC and Bach P. Communicating hands: ERPs elicited bymeaningful symbolic hand postures. Neuroscience Letters,372(1e2): 52e56, 2004.

Hauk O, Johnsrude I, and Pulvermuller F. Somatotopicrepresentation of action words in human motor premotorcortex. Neuron, 41(2): 301e307, 2004.

Hill EL. Non-specific nature of specific language impairment: Areview of the literature with regard to concomitant motorimpairments. International Journal of Language & CommunicationDisorders, 36(2): 149e171, 2001.

Johnson Frey SH, Maloof FR, Newman-Norlund R, Farrer C, Inati S,and Grafton ST. Actions or hand objects interactions? Humaninferior frontal cortex and action observation. Neuron, 39(6):1053e1058, 2003.

Kemmerer D, Rudrauf D, Manzel K, and Tranel D. Behavioralpatterns and lesion sites associated with impaired processingof lexical and conceptual knowledge of actions. Cortex, 48(7):826e848, 2012.

Kempler D, Andersen ES, and Henderson VW. Linguistic andattentional contributions to anomia in Alzheimer’s disease.Neuropsychiatry, Neuropsychology, and Behavioral Neurology, 8(1):33e37, 1995.

Kroliczak G and Frey SH. A common network in the left cerebralhemisphere represents planning of tool use pantomimes andfamiliar intransitive gestures at the hand-independent level.Cerebral Cortex, 19(10): 2396e2410, 2009.

Levelt WJM, Roelofs A, and Meyer AS. A theory of lexical access inspeech production. Behavioral and Brain Sciences, 22(1): 1e38,1999.

Luzzatti C, Willmes K, and De Bleser R. Aachener Aphasie Test:Versione Italiana. [Aachener Aphasia Test: Italian version].Firenze, Italy: Organizzazioni Speciali, 1986.

Luzzatti C, Raggi R, Zonca G, Pistarini C, Contardi A, andPinna GD. Verbenoun double dissociation in aphasic lexicalimpairments: The role of word frequency and imageability.Brain and Language, 81(1e3): 432e444, 2002.

Mahon BZ and Caramazza A. A critical look at the embodiedcognition hypothesis and a new proposal for groundingconceptual content. Journal of Physiology, Paris, 102(1e3):59e70, 2008.

McNeill D. Hand and Mind. Chicago, IL: University of Chicago Press,1992.

Mozaz M, Rothi LJ, Anderson JM, Crucian GP, and Heilman KM.Postural knowledge of transitive pantomimes and intransitivegestures. Journal of the International Neuropsychology Society, 8(7):958e962, 2002.

Nakamura A, Maess B, Knosche TR, Gunter TC, Bach P, andFriederici AD. Cooperation of different neuronal systemsduring hand sign recognition. NeuroImage, 23(1): 25e34, 2004.

Negri GAL, Rumiati RI, Zadini A, Ukmar M, Mahon BZ, andCaramazza A. What is the role of motor simulation in actionand object recognition? Evidence from apraxia. CognitiveNeuropsychology, 24(8): 795e816, 2007.

Ochipa C, Rothi LJG, and Heilman KM. Ideational apraxia: A deficitin tool selection and use. Annals of Neurology, 25(2): 190e193,1989.




5

c o r t e x x x x ( 2 0 1 2 ) 1e1 1 11

1301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343

1344134513461347134813491350135113521353135413551356135713581359136013611362136313641365136613671368136913701371137213731374137513761377137813791380138113821383138413851386


Oldfield RC. The assessment and analysis of handedness: TheEdinburgh inventory. Neuropsychologia, 9(1): 97e113, 1971.

Oliveri M, Finocchiaro C, Shapiro K, Gangitano M, Caramazza A,and Pascual-Leone A. All talk and no action: A transcranialmagnetic stimulation study of motor cortex activation duringaction word production. Journal of Cognitive Neuroscience, 16(3):347e381, 2004.

Papeo L and Hochmann JR. A cross-talk between brain-damagepatients and infants on action and language. Neuropsychologia,50(7): 1222e1234, 2012.

Papeo L, Corradi-Dell’Acqua C, and Rumiati RI. “She” is not like“I”: The tie between language and action is in our imagination.Journal of Cognitive Neuroscience, 23(12): 3939e3948, 2011.

Papeo L, Negri GAL, Zadini A, and Rumiati RI. Action performanceand action-word understanding: Evidence of doubledissociations in left-damaged patients. CognitiveNeuropsychology, 27(5): 428e461, 2010.

Papeo L, Vallesi A, Isaja A, and Rumiati RI. Effects of TMS ondifferent stages of motor and non-motor verb processing inthe primary motor cortex. PloS ONE, 4: e4508, 2009.

Piaget J. The Construction of Reality in the Child. New York, NY:Ballantine, 1954.

Pulvermuller F. Brain mechanisms linking language and action.Nature Reviews Neuroscience, 6(1): 576e582, 2005.

Rizzolatti G and Arbib MA. Language within our grasp. Trends inNeuroscience, 21(5): 188e194, 1998.

Rizzolatti G and Craighero L. The mirror-neuron system. AnnualReview of Neuroscience, 27(1): 169e192, 2004.

Rumiati RI, Zanini S, Vorano L, and Shallice T. A form ofideational apraxia as a selective deficit of contentionscheduling. Cognitive Neuropsychology, 18(7): 617e642, 2001.

Scorolli C and Borghi AM. Sentence comprehension and action:Effector specific modulation of the motor system. BrainResearch, 1130(1): 119e124, 2006.

Shallice T. From Neuropsychology to Mental Structure. Cambridge,UK: Cambridge University Press, 1988.

Spinnler H and Tognoni G. Standardizzazione e taratura italianadi test neuropsicologici. [Italian standardization andvalidation neuropsychological tests]. Italian Journal ofNeurological Sciences, 8: 1e120, 1987.

Tessari A, Canessa N, Ukmar M, and Rumiati RI.Neuropsychological evidence for a strategic control ofmultiple routes in imitation. Brain, 130(4): 1111e1126, 2007.


Tessari A and Rumiati RI. The strategic control of multiple routesin imitation of actions. Journal of Experimental Psychology:Human Perception and Performance, 30(6): 1107e1116, 2004.

Tettamanti M, Buccino G, Saccuman MC, Gallese V, Danna M,Scifo P, et al. Listening to action related sentences activatedfronto-parietal motor circuits. Journal of Cognitive Neuroscience,17(2): 273e281, 2005.

Thal D, Bates E, Goodman J, and Jahn-Samilo J. Continuity oflanguage abilities: An exploratory study of late- and early-talking toddlers. Developmental Neuropsychology, 13(3): 239e273,1997.

Tomasino B, Werner CJ, Weiss PH, and Fink GR. Stimulusproperties matter more than perspective: An fMRI study ofmental imagery and silent reading of action phrases.NeuroImage, 36(2): 128e141, 2007.

Toraldo A, Cattani B, Zonca G, Saletta P, and Luzzatti C. Readingdisorders in a language with shallow orthography: A multiplesingle-case study in Italian. Aphasiology, 20(9): 823e850, 2006.

Wang L and Goodglass H. Pantomime, praxis, and aphasia. Brainand Language, 42(4): 402e418, 1992.

Warrington EK. Recognition Memory Test. London, UK: NFER-Nelson, 1984.

Warrington EK and James M. The Visual Object and Space PerceptionBattery. Bury St Edmunds, UK: Thames Valley Test Company,1991. Q

Warrington EK and McCarthy RA. Categories of knowledge:Further fractionations and an attempted integration. Brain,110(5): 1273e1296, 1987.

Warrington EK and Shallice T. Category specific semanticimpairment. Brain, 107(3): 829e854, 1984.

Werner H and Kaplan B. Symbol Formation: An Organismic-developmental Approach to the Psychology of Language. New York,NY: Wiley, 1963.

Wu YC and Coulson S. Meaningful gestures: Electrophysiologicalindices of iconic gesture comprehension. Psychophysiology,42(6): 654e667, 2005.

Xu J, Gannon PJ, Emmorey K, Smith JF, and Braun AR. Symbolicgestures and spoken language are processed by a commonneural system. Proceedings of the National Academy of SciencesU.S.A., 106(49): 20664e20669, 2009.

Zwaan RA and Taylor LJ. Seeing, acting, understanding: Motorresonance in language comprehension. Journal of ExperimentalPsychology: General, 135(1): 1e11, 2006.




Lexical and gestural symbols in left-damaged patients

Documents

Transcript of Lexical and gestural symbols in left-damaged patients