argument structure deficits in aphasia

Upload: bengue-guenel

Post on 08-Apr-2018

226 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/6/2019 Argument Structure Deficits in Aphasia

    1/25

    PLEASE SCROLL DOWN FOR ARTICLE

    This article was downloaded by: [TBTAK EKUAL]On: 2 March 2011Access details: Access Details: [subscription number 772815468]Publisher Psychology PressInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-

    41 Mortimer Street, London W1T 3JH, UK

    AphasiologyPublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t713393920

    Argument structure deficits in aphasia: New perspective on models oflexical productionElena Barbieria; Anna Bassob; Mirella Frustacic; Claudio Luzzattiaa University of Milan-Bicocca, Milan, Italy b University of Milan, Italy c Salvini General Hospital, Rho-Passirana, Milan, Italy

    First published on: 09 August 2010

    To cite this Article Barbieri, Elena , Basso, Anna , Frustaci, Mirella and Luzzatti, Claudio(2010) 'Argument structuredeficits in aphasia: New perspective on models of lexical production', Aphasiology, 24: 11, 1400 1423, First publishedon: 09 August 2010 (iFirst)

    To link to this Article: DOI: 10.1080/02687030903580325URL: http://dx.doi.org/10.1080/02687030903580325

    Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

    This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.

    The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,

    actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.

    http://www.informaworld.com/smpp/title~content=t713393920http://dx.doi.org/10.1080/02687030903580325http://www.informaworld.com/terms-and-conditions-of-access.pdfhttp://www.informaworld.com/terms-and-conditions-of-access.pdfhttp://dx.doi.org/10.1080/02687030903580325http://www.informaworld.com/smpp/title~content=t713393920
  • 8/6/2019 Argument Structure Deficits in Aphasia

    2/25

    APHASIOLOGY, 2010, 24 (11), 14001423

    2010 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business

    http://www.psypress.com/aphasiology DOI: 10.1080/02687030903580325

    PAPH0268-70381464-5041APHASIOLOGY, Vol. 1, No. 1, May 2010: pp. 00APHASIOLOGY

    Argument structure deficits in aphasia: New perspective on

    models of lexical production

    Argument StructureDeficit sin AphasiaBARBIERI ETAL.

    Elena Barbieri

    University of Milan-Bicocca, Milan, Italy

    Anna Basso

    University of Milan, Italy

    Mirella FrustaciSalvini General Hospital, Rho-Passirana, Milan, Italy

    Claudio Luzzatti

    University of Milan-Bicocca, Milan, Italy

    Background: The study explores the ability of fluent and non-fluent aphasic patients toproduce verbs using the appropriate argument structure, i.e., a feature specifying thenumber and type of participants in the event described by the verb. According to Chomskys

    Minimalist Program (1995), the lexical entry of a verb contains information about thenumber of arguments and the thematic roles assigned, which is then mapped onto thesentence argument structure, while Bock and Levelts (1994; see also Levelt, Roelofs, &Meyer, 1999) model of lexical production assumes that the information regarding verbargument structure (VAS) is part of the lemma and is accessed before the retrieval of thephonological word form, i.e., the lexeme.Aim: Participants were tested to investigate their ability to map thematic informationonto the corresponding syntactic argument structure.Methods & Procedures: Seven aphasic patients (five suffering from non-fluent aphasiawith agrammatism and two from fluent aphasia) and ten neurologically unimpairedindividuals participated in the study. They were given a picture description task formulatedin two conditions: in the first condition they were asked to provide a free description of

    the image, while in the second condition they had to complete the sentence structureprovided by the examiner. Patients showing deficits in verb production were also testedfor the use of prepositions within prepositional compounds and in a sentence context.Outcomes & Results: Four agrammatic patients and one fluent aphasic patient scored ahigh rate of argument structure errors in the selection of the appropriate verb, whichwas employed with an incorrect argument structure. Furthermore, these patients tendedto substitute rather than to omit prepositions when required to fill the gap in a sentence.

    Address correspondence to: Claudio Luzzatti, Department of Psychology, University of Milano-Bicocca,

    Piazza dellAteneo Nuovo 1; 20126 Milan, Italy. E-mail: [email protected]

    This research was partially supported by a Grant from the Regione Lombardia to E.B. and from theMIUR to C.L. We are grateful to Prof. Germano Rossi for the helpful suggestions given regarding the

    statistical analyses, and to Prof. Alfonso Caramazza for his insightful comments on a previous version of

    the manuscript.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

    http://www.psypress.com/aphasiologymailto:[email protected]://www.psypress.com/aphasiologymailto:[email protected]
  • 8/6/2019 Argument Structure Deficits in Aphasia

    3/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1401

    Conclusions: Our results indicate an impaired access to (VAS) and/or in mapping thethematic role information onto the syntactic argument structure. This deficit, whichwas found in both the agrammatic and the fluent aphasic patients, conflicts with analternative interpretation of verb production errors in terms of omission of the preposi-tion introducing the manner adjunct. Data support the hypothesis of retrieval of the

    verb lexeme without prior access to the corresponding lemma, where information aboutargument structure is stored.

    Keywords: Argument structure deficits; Agrammatism; Lemma and lexeme; Picturedescription; Verb retrieval deficits; Syntactic impairment.

    A verb argument structure (VAS) is a lexical feature specifying the number and type

    of participants in the event that is described by a verb. Verbs always require at least

    one external argument (the subject)1, the number depending on the category to which

    the verb belongs. More specifically, while intransitive verbs usually take one argu-ment, which plays the role of grammatical subject, transitive verbs need at least two

    arguments, i.e., a subject and a direct object respectively. For instance, in the sen-

    tence The woman cuts the cake with a knife, to cut is a transitive verb, and the

    noun phrases (NPs) the woman and the cake are arguments. This sentence also

    contains a prepositional phrase (PP) with the knife, which is an adjunct, i.e., a

    non-obligatory element. Each argument (but not adjuncts) also receives a thematic

    role describing the logical function that an element plays in the sentence and is

    assigned according to the thematic grid of the verb. The latter is part of the verbs

    lexical representation, which also contains the verb argument structure.

    Several studies (Berndt, Haedinges, Mitchum, & Sandson, 1997; Luzzatti et al.,2002; Miceli, Silveri, Nocentini, & Caramazza, 1988; Thompson, Shapiro, Li, &

    Schendel, 1995) have shown that aphasic patients, particularly those suffering from

    agrammatism, may show more severe impairment in retrieving verbs than nouns.

    Agrammatism is one of the main aspects of Brocas aphasia and its main feature is a

    telegraphic output, i.e., sentences are short (S-V or S-V-O) and syntactically

    simplified, with omission of closed class vocabulary, i.e., omission of bound morphemes

    in English, substitution of marked inflected forms with less-marked ones in Italian.2

    Furthermore, agrammatic patients show impaired comprehension of syntactically

    complex sentences where thematic role assignment is non-canonical, as in interroga-

    tive, passive, and object relative clauses. Another morphosyntactic deficit that often

    occurs in fluent aphasia is paragrammatism, which can be defined as a production

    disorder where grammatical morphemes are substituted instead of being omitted.

    Several theories have been proposed to explain morphosyntactic disorders. A

    theoretical account of agrammatic production was advanced by Friedmann and

    co-workers (Friedmann, 2000, 2002; Friedmann & Grodzinsky, 1997) to explain the

    difficulty of producing syntactically complex sentences and the dissociation between

    the production of tense (impaired) and agreement (spared) inflections. According to

    Friedmann, the production deficit in agrammatic aphasia is due to a disruption

    1This is not the case for Italian meteorological verbs such as piovere (piove, [it] rains) and for imper-

    sonal verbs such as sembrare (sembra, [it] seems), which do not take a subject; furthermore, the Italian

    language allows subject pro-drop, i.e., the omission of subject pronouns.2Italian has a very rich morphology in which citation forms are also inflected; therefore, an omission of

    bound morphemes often results in the production of non-lexical roots.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    4/25

    1402 BARBIERI ET AL.

    (pruning) of the syntactic tree, in which the upper nodes are impaired. It is

    assumed that the syntactic tree is pruned at the tense phrase (TP)3 node in severe

    agrammatic patients and at the complementiser phrase node (CP) in milder cases.

    The basic assumption therefore is that the tree is pruned from the impaired node

    upward: thus, patients who have problems with tense inflections and subject pronounsare impaired at the TP node, while if the damage is at the CP level, the deficit affects

    the production of questions and embedded sentences.

    Several studies showed that impaired access to the verb argument structure is one

    of the factors that causes difficulty in retrieving verbs typical of agrammatic production.

    For instance, Thompson and co-workers (Thompson, Lange, Schneider & Shapiro,

    1997) found an influence of the number of arguments taken by the verb: agrammatic

    patients found it easier to produce intransitive verbs, i.e., verbs taking only one exter-

    nal argument (the subject) to which the role of Agent is assigned, rather than transi-

    tive verbs, i.e., verbs requiring two arguments to which they assign the roles of Agent

    and Theme. These verbs were easier to produce than verbs such as to give or to put,which require a three-place-argument structure, with the theta roles being respec-

    tively Agent, Theme, and Goal and Agent, Theme, and Location. Finally, even

    threeplace verbs were more easily produced than verbs such as to say or to know,

    which are usually followed by a whole sentence (for example: the girl knows the cat is

    in the tree). If these data are interpreted in the light of Levelts model of language

    production (Bock & Levelt, 1994; see also Levelt, Roelofs, & Meyer, 1999), the verb

    production deficit which emerges in agrammatism may be localised at the lemma

    level and this, in a serial model, also implies failure in retrieving the lexeme. In the

    case of a real lemma disruption, as discussed by Bastiaanse and van Zonneveld

    (2004), patients could be expected to encounter the same kind of difficulty in compre-hension as well, since there is only one set of lemmas, although several studies (Jonk-

    ers, 1998; Kim & Thompson, 2000; Shapiro & Levine, 1990) reported good verb

    comprehension and a preserved ability in detecting argument structure errors in a

    grammaticality judgment task. Thus Kim and Thompson hold that verb-naming

    impairments emerging in Brocas aphasia and depending on argument structure

    complexity, would arise at a lemma selection level. Bastiaanse and van Zonneveld,

    on the other hand, suggest that the deficit would arise after the lemma has been

    selected, at the grammatical encoding level, i.e., when the creation of a syntactic envir-

    onment takes place (phrases, clauses, and sentences). This interpretation rephrases

    the mapping hypothesis in a syntactic frame (Rochon, Laird, Bose, & Scofield,2005; Schwartz, Linebarger, Saffran & Pate, 1987); this hypothesis conceived the

    agrammatic deficit as an impaired mapping between thematic roles and the syntactic

    argument structure required by a verb. The assumption made by Bastiaanse and van

    Zonneveld also explains why aphasic patients have greater difficulty in retrieving

    unaccusative (e.g., to fall) as opposed to unergative verbs (e.g., to sleep; Luzzatti et

    al., 2002; Thompson, 2003). The latter require only one argument, the subject, which

    is external to the verb phrase (VP) and directly receives the role of Agent; on the con-

    3Following Chomskys Principles and Parameters Theory (Chomsky & Lasnik, 1993), a sentence may

    be represented as a tree with nodes and branches. Each node splits into a head, a complement, and a

    specifier. A precise hierarchy is assumed to govern the syntactic tree, which dictates the order in which the

    nodes occur; for example, the CP node, which hosts complementisers (such as that) and question

    morphemes (such as who and what), projects to the TP node, which is responsible for the tense inflection of

    the verb.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    5/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1403

    trary, unaccusative verbs differ in that their grammatical subjects do not perform but

    undergo the action expressed by the verb. Thus, these verbs assign thematic roles dif-

    ferently from unergative verbs, in that the grammatical subject receives the role of

    Theme in its initial position, that is as verb complement, and moves out of VP, leav-

    ing behind a trace which allows the assignment of the thematic role (see, for instance,Burzio, 1986). Therefore, unaccusative verbs require a more complex syntactic pro-

    cessing than unergative verbs, which are easier to produce for aphasic patients.

    Bock and Levelts (1994) language production model specifies Garretts (1980)

    distinction between functional and positional levels. In their model, information

    about argument structure and thematic role assignment is stored in the lemma which

    carries the grammatical information associated with individual lexical concepts,

    such as grammatical class (i.e., noun or verb) and the number of arguments subcate-

    gorised by the verb (Bock & Levelt, 1994, p. 947). This model considers word

    production as a serial process, where each unit is assumed to be influenced only by

    information represented at the level directly above it (Bock & Levelt, 1994, p. 949).The functional level is conceived as the first step of grammatical encoding, i.e., the

    set of processes generating the skeleton of an utterance. On the contrary, lexemes,

    which represent the morphological and phonological features of words (word form).

    Caramazza (1997) developed an alternative model, known as the independent net-

    work model of lexical access, which aims at overcoming lemmalexeme seriality.

    According to Caramazza the information that in Levelts framework is assumed to

    be stored in the lemma, is divided into two independent networks: semantic features

    are assumed to be part of a lexical-semantic network, while the lexical-syntactic net-

    work contains information on grammatical category, gender, auxiliary type, and

    argument structure. There are also two modality-specific lexeme networks for pho-nological (P-lexeme) and orthographic (O-lexeme) representations. He assumed that

    the lexical-semantic network (i) projects to both the P and the O lexeme, which in

    turn project to the lexical-syntactic network, and (ii) is directly connected to this lat-

    ter. The activation of a lexical-semantic representation spreads simultaneously and

    independently to both the lexical-syntactic and the word-form systems. However,

    while the lexeme-networks can reach the threshold and thus be directly activated by

    the lexical-semantic network, the lexical-syntactic system can only be primed by it,

    and needs to be reinforced by the phonological/orthographic representation in order

    to be activated.

    To summarise, the aphasia literature confirms that the information regardingargument structure is stored at the lexical level (the lemma) and that it may be

    impaired by brain damage; alternatively, aphasic patients may suffer from a deficit in

    mapping the thematic information onto the corresponding sentence argument structure.

    Most of these studies were conducted on agrammatic patients, which resulted in a

    number of hypotheses accounting for this type of language disorder. However, it has

    been seen that fluent patients also suffer from the same type of syntactic damage as

    agrammatic patients (see for instance Luzzatti et al., 2001).

    The objective of the present study is to pursue investigation of argument structure

    deficits in Italian aphasic patients suffering from non-fluent agrammatic or fluent

    aphasia, in order to test contemporary models of lexical production and verify current

    theories of agrammatic production. A specific picture-naming task (Experiment 1) was

    prepared, which tested the patients ability to describe pictures using syntactically

    simple sentences containing a transitive two-argument verb. In Experiment 2 patients

    were given two further tasks to assess their use of prepositions, as impaired processing

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    6/25

    1404 BARBIERI ET AL.

    of this word type usually characterises agrammatism (e.g., Grodzinsky, 1990; Lonzi

    & Luzzatti, 1995); these tasks were used to disentangle this possible disorder from a

    deficit in realising verb arguments (or in mapping thematic roles to the corresponding

    argument structure).

    EXPERIMENT 1

    Aim

    Our aim was to throw light on the use of verbs with an abnormal argument structure,

    resulting in errors like *luomo mangia la forchetta (*the man eats the fork) through

    testing fluent and non-fluent aphasic patients in a picture description paradigm.

    Hence it is possible that the patients encounter difficulties in mapping thematic

    knowledge to the corresponding argument structure. The occurrence of this type of

    error has a theoretical consequence: the production of a sentence of this nature indi-cates that verbs are being retrieved but used with an incorrect argument structure.

    In order to investigate this phenomenon further, we developed a new task in

    which patients were asked to describe 53 pictures of actions using the following

    sentence structure:

    Subject + Verb + Direct object + Manner adjunct

    The task had two conditions: a free condition, where participants described the

    picture without any constraints, and a constrained condition, in which the examiner

    provided the participants with the sentence structure, asking them to complete it by

    filling in the lexical gaps.

    Method and materials

    Participants. The study was carried out with seven Italian aphasic participants

    (five males and two females), recruited at the Aphasia Rehabilitation Unit of Milans

    Ospedale Maggiore Policlinico and at the Rehabilitation Unit of the G. Salvini

    General Hospital at Passirana (Milan) and a control group of ten healthy partici-

    pants (six males and four females). Controls were between 25 and 66 years old,

    and had 8 to 13 years of education.

    Participants with mild-to-moderate fluent or non-fluent aphasia with predominantmorpho-syntactic difficulties (omission or substitution of verbs, articles, prepositions)

    were selected for the study. Patients suffering from visual deficits that might impede

    correct perception of the test material were excluded from the study.

    Case reports. All participants took a standard language examination which included

    the Token Test (De Renzi & Vignolo, 1962) and a standard aphasia battery, such as the

    Esame di Milano II (Ciurli, Marangolo, & Basso, 1996), the Battery for the Analyses

    of Aphasic Disorders (Miceli, Burani, Capasso, & Laudanna, 1994), or the Italian

    version of the Aachen Aphasia Test (AAT; Luzzatti, Willmes, & De Bleser, 1994).

    Five of the aphasic patients participating in the study suffered from non-fluentagrammatic aphasia, and two had a fluent and paragrammatic output, but occasionally

    omitted free and bound morphemes. Table 1 summarises clinical data for each

    patient, and a sample of their spontaneous speech is given in Appendix 1.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    7/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1405

    L.G. was a right-handed 25-year-old barman. His speech output was slow and

    dysprosodic with omission of function words and impaired comprehension of sentences

    with non-canonical word order. Picture naming was only slightly impaired (95%

    correct on nouns; 90% correct on verbs); word-to-picture matching was unimpaired

    (100% correct), while sentence-to-picture matching was moderately impaired (75%

    correct). However, his score on the Token Test was low (9 correct responses out of

    36) due to a severe phonological short-term memory deficit. His language disorderwas classified as a moderate non-fluent aphasia with agrammatism.

    F.C. was a 48-year-old right-handed mechanic, whose language disorder appeared

    in 1997. His spontaneous speech was slow, with long and frequent pauses, no subor-

    dination, and scarce retrieval of verbs, which were often produced in the non-finite

    form. Picture naming was moderately impaired (62% correct on names; 56% on

    verbs) Comprehension was good on single words (99% correct), but impaired on

    sentences (85% correct). His performance on the Token Test was moderately

    impaired (16 correct responses out of 36). F.C.s language disorder was classified as a

    moderate non-fluent aphasia with agrammatism.

    N.C. was a 56-year-old right-handed housewife with 5 years of education. Herspontaneous speech was slow and dysprosodic and often interrupted by long pauses,

    with phonemic paraphasia and speech apraxia. Sentences were very short, usually

    limited to one or two words, with constant omission of verbs and of grammatical free

    and bound morphemes. Picture naming was moderately to severely impaired on

    verbs (47% correct) and mildly to moderately impaired on nouns (80% correct).

    Comprehension was relatively preserved (but she obtained only 20 correct responses

    out of 50 on the Token Test section of the AAT). Her language disorder was classified

    as a moderate non-fluent aphasia with agrammatism.

    L.M. was a 56-year-old male clerical employee. His spontaneous speech was

    severely impaired due to anomia and several automatic elements. He produced veryfew verbs and substituted inflectional endings with less-marked forms; closed-class

    words were almost always omitted. Naming of pictures was severely impaired on

    verbs (34% correct) and only moderately on nouns (72% correct). Comprehension

    was mildly to moderately impaired on the AAT battery and moderately to severely

    TABLE 1

    Main features of the aphasic patients participating in the experiment

    Name Sex Age YOE Aetiology

    Months post

    onset Lesion Diagnosis

    Language

    examination

    L.G. M 25 13 traumatic brain

    injury

    24 left TP and

    right FP

    agrammatism Esame di

    Milano II

    F.C. M 48 13 stroke 118 left FTP and

    left sylvian

    agrammatism BADA

    N.C. F 56 5 stroke 28 left FTP agrammatism AAT

    L.M. M 56 13 stroke 26 left TPO agrammatism AAT

    A.C. F 31 11 stroke 86 left FTP and

    left insula

    agrammatism AAT

    D.S. M 63 5 stroke 17 left FT, left

    caudate

    nucleus

    fluent aphasia AAT

    B.D. M 67 6 stroke 38 left FTP fluent aphasia BADA

    YOE = years of education; T = temporal, F = frontal, P = parietal, O = occipital.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    8/25

    1406 BARBIERI ET AL.

    impaired on the Token Test (13 correct responses out of 50). His language disorder was

    classified as a moderate-to-severe non-fluent aphasia with agrammatism.

    A.C. was a 31-year-old shop assistant with 11 years of education. She had no

    articulatory deficit, but her speech output was impaired with recurring phonemic

    paraphasia, anomia, and frequent use of circumlocutions. Sentences were short andsyntactically simplified, with predominant omission of function words; inflectional

    endings were frequently substituted with less-marked forms. Her performance on the

    naming task was moderately impaired (62% correct on verbs, 70% correct on nouns).

    She also had difficulty in oral comprehension (and scored 17 correct out of 50 on the

    Token Test). Her language disorder was classified as a moderate non-fluent aphasia

    with agrammatism.

    D.S. was a 63-year-old metal worker by trade with 5 years of education, whose

    language impairment emerged in 2005. His speech was fluent, with severe anomia,

    phonemic paraphasia, neologisms, and some perseveration. He produced severe

    morpho-syntactic errors on the figure description task; his sentences were short andoften interrupted but contained elements indicating traces of a complex syntactic

    structure. A language examination showed severe repetition disorders on the AAT,

    impaired picture naming (50% correct on verbs, 54% correct on nouns), and mild-to-

    moderate comprehension disorders, but he obtained 20 out of 50 correct responses at

    the Token Test. His language impairment was classified as a moderate fluent aphasia.

    B.D. was a 67-year-old retired butcher. His language impairment was fluent, with

    conspicuous grammatical/syntactic difficulties. Although he did retain some clear

    traces of a complex sentence structure, he seldom produced complete subordinate

    sentences, and there were several agreement errors (subjectverb or articlenoun

    agreement). Picture naming was very good for nouns (98% correct) and mildlyimpaired on verbs (81% correct). Word and sentence comprehension were mostly

    preserved (100% and 81% correct respectively), while his performance at the Token

    Test was moderately impaired (20 correct responses out of 36). His language disorder

    was classified as a mild-to-moderate fluent aphasia.

    Stimuli and experimental procedure. A picture-naming task was devised in order

    to detect impaired access to the verb argument structure. The task was composed of

    53 images (28 b/w drawings and 25 pictures), which depicted actions requiring a tool.

    All the images were preliminarily shown to a sample of control participants, who hadto describe each event using verbs taking the same argument structure, i.e., a subject,

    an object, and a manner adjunct. The name of the tool had to be morphologically

    unrelated to the name of the verb, so images triggering answers such as luomo sega

    un tronco con la sega (the man saws a trunk with a saw) were not included. Appendix

    2 contains the list of the sentences included in the final version of the target task. The

    task had two conditions.

    Condition 1. Participants were shown the images depicting the actions. The examiner

    read the following instructions: Now I will show you pictures and drawings depicting

    actions. Look at each picture and tell me what happens in it. Tell me who is depicted

    in the picture, what the person is doing and which tool he/she is using. Two exampleswere given before the actual task was set; if the action Theme or tool were omitted, the

    examiner pointed to the omitted element in the picture and prompted: Well, and

    what about this? This support was not given during the test itself.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    9/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1407

    Condition 2. Each item, examples included, was presented in a second run; in the

    second condition the participants were also presented with the sentence structure

    (Figure 1) that they had to complete using two flash cards (Figure 2) provided by the

    examiner.

    Figure 1. Sketched representation of item number 52, luomo affetta il pane con il coltello (the man slices

    the bread with the knife). In Condition 1, participants were shown only the upper part of the figure, and

    were asked to describe the picture by saying who is depicted in the figure, what the person is doing, and

    which tool he/she is using. In Condition 2, participants were also presented with the sentence structure

    (lower part of the figure), were asked to add a verb and complete the sentence using two flash cards pro-

    vided by the examiner (see Figure 2) and to spell out the whole sentence.

    L UOMO

    The man

    Figure 2. Example of the flash card supplied to the participants to fill the gaps of Figure 1. The picture on

    the right represents the direct object and has to be located after the verb. The picture on the left depicts the

    tool by which the action is carried out and should be placed in the final position.

    IL COLTELLO

    the knife

    IL PANE

    the bread

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    10/25

    1408 BARBIERI ET AL.

    Each sentence was composed of five elements: the grammatical subject, the verb,

    the object/Theme, the preposition heading the manner adjunct, followed by the tool

    used to perform the action. Theme and instrument were represented in written and

    picture form on two flash cards given to the participant (Figure 2), who then had to

    fill the gaps in the sentence frame (see lower part of Figure 1) and to spell out the sen-tence. Pictures and names of VP internal arguments/adjuncts were provided in order

    to bypass possible lexical retrieval disorders.

    In condition 2 the examiner said: Now the task will be easier, because I will give

    you some cues. Look at the lower part of the stimulus: first you have to say which

    person carries out the action (pointing to the first element on the left), then (indicat-

    ing the second element) you must retrieve the verb and finally (pointing to the gaps in

    the third and fifth position) produce the two elements of the action which are

    depicted on the flash cards. The two flash cards were stacked one upon the other to

    avoid the patient thinking that the order given by the examiner would correspond to

    that of the target. The examiner opened the first example with the words A man . . .and the participant had to retrieve the appropriate verb by placing the two flash

    cards in the right order and by uttering the corresponding content words and the

    fitting preposition (fourth element).

    All patients were examined in both conditions. There was no time limit; however,

    if a patient was unable to generate a sentence within 34 minutes, s/he was invited to

    pass to the following item. If one of the main elements of a sentence (direct object or

    tool) was omitted, the examiner asked the participant to produce it. Task instructions

    were repeated whenever needed. The patients entire production was audio-recorded

    and transcribed including pauses, self-corrections, and the examiners interventions.

    The responses produced by the participants were divided into six differentcategories:

    1. Correct production: sentences containing an Agent, an appropriate transitive

    verb, a Theme, and a manner adjunct were scored as correct. Errors in the

    articlename or nameverb agreement, or the production of verbs in a non-

    finite form were not counted as errors. We also classified as correct responses

    that were generated after the examiner elicited the production of omitted ele-

    ments, and sentences containing lexical substitutions but resulting in argument

    structures that were appropriate for the retrieved verbs (for instance, the man

    folds a piece of paper with his hands instead of the man models a plane withpaper). In condition 1, sentences were computed as correct when the gram-

    matical subject, the direct object, or the tool of the action were omitted (the

    woman reads with a magnifying glass for the woman reads a book with a

    magnifying glass). Finally, responses in which the elements of interest were

    distributed into two clauses (for instance the boy opens a bottle by using a

    corkscrew for the boy opens a bottle with a corkscrew) were classified as

    correct.

    2. Sentences containing passe-partout verbs (to do, to use, to put, and to take) were

    considered incorrect. An answer was classified in this class if the passe-partout

    verb was the only one produced, as in *the woman puts a piece on the pot withsome glue (target sentence the woman repairs a pot with some glue).

    3. Verb-argument structure errors. In most cases, the action tool became the Theme

    of the action as in the sentence *the girl sews a thread for the button (target

    sentence a woman sews a button with a thread). Errors were also classified in this

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    11/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1409

    class if they occurred in a sentence containing one or more lexical substitutions of

    the subject, theme, or adjunct (e.g., *the woman eats a fork with an egg,

    produced instead of the woman whisks an egg with a fork).

    4. Substitution/omission of prepositions. This category grouped sentences where the

    preposition with was missing or substituted (e.g., *the woman cuts the paperfor the scissors instead of the woman cuts the paper with the scissors).

    5. We classified as other errors all the remaining abnormal responses and, in

    particular, all cases in which it was not possible to make a clear-cut decision

    regarding the type of error: e.g., *luomo importa la sua salute (*the man cares his

    health).

    6. Non-structured responses / no response.

    Statistics. Data were analysed with chi-square statistics to assess the goodness-of-fit

    to a theoretical model, and particularly to investigate the occurrence of verb-argument

    structure errors. The behaviour of the non-neurological participants in the first two

    conditions of the task constituted the theoretical model of reference. We considered the

    mean of the errors made by control participants in the two conditions (in percentage).

    Responses were divided into two error categories: verb-argument structure errors

    (VAS errors) and other errors, including categories (2), (4), (5), and (6). The rate

    of correct answers was not considered, as it would have induced an artificial increase

    ofc2 values. Observed values source from the aphasic speakers suffering from mild-to-

    moderate disorders, while expected values refer to neurologically intact participants.

    Two-by-two contingency tables were prepared using the following procedure: first

    we computed the mean number of verb-argument structure errors and of other typesof errors in the control sample (10 participants) for the two conditions (106 responses

    for each participant). Mean values were then transformed into percentages. Expected

    frequencies were computed for each patient: i.e., the number of errors made by each

    patient was divided by the total error rate of the control group and then multiplied

    by the expected error rate in each category. Finally, we computed c2 by applying the

    general formula:

    where OF and EF indicate observed and expected frequencies respectively.

    Results

    Control participants. The control participants did not have any difficulty in

    performing the task (see Table 2). They responded correctly to most items in the two

    conditions, and the percentage of hits increased in the constrained condition. Theirerrors were mainly the use of passe-partout verbs (as to put or to do); only two

    responses contained a verb argument error and no omissions or substitutions of

    prepositions were generated. The total rate of error was 3.2%: 0.1% for the verb-

    argument structure errors category and 3.1% for other types of error. These

    ( )OF-EF

    EF

    2

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    12/25

    1410 BARBIERI ET AL.

    percentages were considered as the theoretical model of reference in computing the

    c2 values.

    Aphasic patients. Contrary to the pattern of performance that emerged in the

    control group, some of the aphasic patients did not improve their performance in the

    constrained condition (see Table 3). In particular, participants L.M., N.C., and F.C.

    produced a greater number of errors in the second condition than in the first one.

    When invited to complete a sentence, they often produced verb argument errors and

    substituted or omitted prepositions, while they tended to omit the action tool in the freedescription condition. They also produced more passe-partout verbs and unstruc-

    tured responses. Table 3 reports the various error types produced by the patients.

    In the performance of the agrammatic patient F.C., VAS errors represented 30%

    and 24.5% respectively of the responses in the two conditions (see Table 3). When

    considering the rate of VAS errors with respect to the entire number of errors, it con-

    stituted 52% and 27% in the two conditions respectively. The patient produced a

    higher proportion of VAS errors than the control participants (see Table 4).

    The performances of L.G. and A.C. were also characterised by several VAS

    errors. For L.G., errors mainly occurred in the free description condition (15% of all

    responses), while in the second condition he produced a majority of passe-partoutforms (45%) and errors in the use of prepositions. A.C. produced the same rate of

    correct sentences in the two conditions (66% of responses in the first and 68% in the

    second); her errors were mainly substitutions of the target verb form with passe-par-

    tout forms and the production of sentences with anomalous argument structure

    (respectively 17% and 41% of all errors). L.M. made approximately the same rate of

    VAS errors in the two conditions (9.5% and 15% respectively), and his errors were

    mainly constituted by the use of passe-partout verbs. N.C. and D.S., on the other

    hand, made only a few VAS errors, and errors mostly involved the substitution of

    verbs with passe-partout forms, the retrieval of prepositions (which were often omit-

    ted) and the production of unstructured sentences (N.C. in particular). Finally, the

    performance of the fluent aphasic patient B.D. did not reveal a clear pattern of dam-

    age, since he obtained the highest percentage of correct responses and his few errors

    were distributed among the various categories without a clear prevalence of sentences

    containing anomalous VAS (see Table 4). Regarding the use of prepositions, L.G. did

    TABLE 2

    Results obtained by the control participants in the two experimental conditions

    Score (range: 053)

    Condition 1 Condition 2

    Type of answer M % SD M % SD

    Correct response 50.3 94.9 2.0 52.3 98.7 1.1

    Omission/Substitution of preposition 0 0

    Passe-partout verb 2.6 4.9 1.8 0.6 1.1 0.8

    Verb-argument structure error 0 0.1 0.2 0.3

    Other types of error 0.1 0.2 0.3 0

    Non-structured/no response 0 0

    For each condition mean (M), percentage (%) and standard deviation (SD) are given.

    Condition 1 = unconstrained picture description. Condition 2 = constrained condition.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    13/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1411

    TABLE

    3

    Resultsobtainedbythes

    evenaphasicpatients

    Corre

    ctresponse

    Omission/Substitution

    ofprepositio

    n

    Passe-partoutverb

    Verbargument

    structureerror*

    Othertypesoferror

    Non-structured

    /noresponse

    Patient

    N

    %

    N

    %

    N

    %

    N

    %

    N

    %

    N

    %

    CONDITION1

    L.G.

    22

    41.5

    4

    7.5

    7

    13

    8

    15

    8

    15

    4

    7.5

    F.C.

    22

    41.5

    8

    15

    3

    6

    16

    30

    0

    0

    4

    7.5

    N.C.

    18

    34

    7

    13

    2

    4

    0

    0

    0

    0

    26

    49

    L.M.

    11

    21

    3

    5.5

    26

    49

    5

    9.5

    5

    9.5

    3

    5.5

    A.C.

    35

    66

    1

    2

    12

    23

    3

    6

    2

    4

    0

    0

    D.S.

    29

    55

    4

    7.5

    16

    30

    0

    0

    1

    2

    3

    5.5

    B.D.

    42

    79

    1

    2

    5

    9

    2

    4

    2

    4

    1

    2

    CONDITION2

    L.G.

    20

    38

    4

    7.5

    24

    45

    3

    5.5

    2

    4

    0

    0

    F.C.

    4

    7.5

    20

    38

    7

    13

    13

    24.5

    2

    4

    7

    13

    N.C.

    6

    11

    19

    36

    0

    0

    3

    6

    1

    2

    24

    45

    L.M.

    1

    2

    3

    5.5

    38

    71.5

    8

    15

    2

    4

    1

    2

    A.C.

    36

    68

    0

    0

    10

    19

    7

    13

    0

    0

    0

    0

    D.S.

    31

    58

    9

    17

    9

    17

    1

    2

    2

    4

    1

    2

    B.D.

    44

    83

    1

    2

    1

    2

    3

    6

    4

    8

    0

    0

    Foreachcategory,number(N)andrate(%)ofcor

    rectresponsesarereportedseparatelyforthetwoexperimentalconditions.

    *Someofthee

    rrorsthatwereincludedinthisca

    tegorycouldactuallybeexplained

    aserrorsinselectingtheprepositionintroducingthemanneradjunct.Thiscould

    accountfortheper

    formanceofsomepatients:L.G.=

    7outof11errors;F.C.=7outof39errors;N.C.=1outof3errors;L

    .M.=2outof13errors;B.D.=1o

    utof5errors.

    A.C.andD.S.didnotproduceanyambiguouserror.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    14/25

    1412 BARBIERI ET AL.

    not omit prepositions in any of the sentences in which VAS errors emerged. F.C.

    introduced a prepositional adjunct in 19 out of 29 sentences containing VAS errors,

    while A.C.s VAS errors (10) were always based on the use of a direct object for the

    tool and a prepositional adjunct for the Theme of the action, e.g., *donna chiude le

    chiave con la. . .la porta (*woman closes the keys with the. . .the door). L.M. pro-

    duced prepositions in 8 out of 13 sentences with an error in the VAS. Finally, B.D.

    made a total of five VAS errors and never omitted prepositions.

    Discussion

    Of the seven aphasic patients participating in the study, five scored a statistically

    significant rate of errors involving the VAS. Four (L.G., F.C., L.M., and A.C.) suf-

    fered from non-fluent aphasia with agrammatism, whereas one patient (B.D.) was

    classified as having Wernickes aphasia. Patients made a relevant number of errors in

    mapping thematic roles to the corresponding sentence argument structure in both

    conditions. These errors indicate that patients may be able to retrieve the correct verb

    but be unable to use it with the appropriate VAS. In detail, patients seemed to adopt a

    default VAS, in which the role of Theme is assigned to the instrument, as is the case

    with the verb to use. This result is particularly interesting since it is at odds with a

    strict application of Levelts model of lexical selection (Bock & Levelt, 1994; see also

    Levelt et al., 1999), which maintains a strictly sequential lemma lexeme access.However a sequential lemmalexeme access cannot account for the performance of a

    patient who is able to retrieve the verb lexeme but uses it with an incorrect argument

    structure (i.e., without accessing the information contained in the lemma). An altern-

    ative explanation is that agrammatic patients usually omit function words, therefore

    their errors may be explained as an omission of the preposition with introducing

    the manner adjunct, rather than an impaired selection of the appropriate argument

    structure; for instance, the production of a sentence like * la donna taglia la forbice

    (*the woman cuts the scissors) could be the result of a correct retrieval of the target

    sentence structure la donna taglia con la forbice (the woman cuts with the scissors)

    and the omission of the preposition with. However, there are numerous cases that

    do not fit this interpretation (61 out of 78 errors made by patients throughout the

    TABLE 4

    Observed frequencies and 2-values referring to the patients performance on the picture

    description task

    Verb argument

    structure errors Other errors Total

    Controls 0.10% 3.10% 3.20% 2 p

    L.G. 11 53 64 41.80 < .001

    F.C. 29 51 80 290.00 < .001

    N.C. 3 79 82 0.062 ns

    L.M. 13 81 94 36.30 < .001

    A.C. 10 25 35 74.30 < .001

    D.S. 1 45 46 0.104 ns

    B.D. 5 15 20 33.30 < .001

    Errors made in Condition 1 and Condition 2 were considered. Controls = mean % of the controlsample production.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    15/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1413

    picture description task, i.e., 72% of the whole corpus): e.g., L.M. produced *il mura-

    tore costruisce il mattone per il muro (*the mason builds the brick for the wall), L.G.

    said *la donna taglia la forbice con la carta (*the woman cuts scissors with the paper),

    while F.C. stated *la signora pulisce il fazzoletto per il naso (*the woman cleans the

    handkerchief for the nose). These sentences were all produced by non-fluent aphasicpatients and showed that they were able to retrieve prepositions but placed the tools

    in the direct object position, and expressed the Theme as a prepositional phrase. Fur-

    thermore, a patient suffering from Wernickes aphasia produced errors in the verb

    argument structure, which is clearly in contrast with the preposition omission expla-

    nation, since fluent aphasic patients generally do not omit prepositions.

    These results therefore cannot be interpreted in terms of omission of preposition

    introducing the manner adjunct, while a defective retrieval of verb argument structure

    appears to be a more likely explanation.

    However, in order to further exclude the possibility that the abnormal production

    described so far is caused by an impaired retrieval of prepositions, a second experi-ment was introduced to test their correct use in patients with presumed impaired

    VAS retrieval.

    EXPERIMENT 2

    Aim

    The aim of this experiment was to evaluate the use of prepositions in five patients

    who made a relevant number of VAS errors in the picture description task. In par-

    ticular we wanted to exclude that the results obtained in the first experiment derivedfrom an inability to retrieve prepositions (a condition which usually characterises

    agrammatic production), rather than from a failure to retrieve the correct VAS. In

    order to achieve this, we administered two different tasks, tapping the use of the

    prepositions that are encoded at lexical level (Luzzatti & De Bleser, 1996) or in a sen-

    tence context (Lonzi & Luzzatti, 1995). The first task was included as a control con-

    dition, to differentiate the impaired use of prepositions at sentence level from a

    deficit already appearing within the NP. The second task specifically taps the selec-

    tion of prepositions in a condition that is similar to that required in Experiment 1,

    since they depend on the VAS.

    Method and materials

    The participants were asked to take a test tapping the retrieval of prepositions. The

    first task checked their ability to generate prepositions in noun phrases such as pasta

    al forno (baked pasta [lit. pasta at the oven]), statua di marmo (marble statue [lit.

    statue of marble]) or arma da fuoco (firearm [lit. arm for fire]). The test was composed

    of 60 nouns modified by a prepositional phrase (for example mulino a vento [wind-

    mill]). The examiner pronounced two nouns separated by an audible pause reinforced

    by a click. The patient had to orally produce the prepositional compound while

    adding the linking preposition and, if necessary, the clitic article. The 60 preposi-

    tional compounds were mixed with 18 noun-noun (N-N) and 9 noun-adjective com-

    pounds (for example,ferrovia, railway), in which neither a preposition nor an article

    had to be inserted (Luzzatti & De Bleser, 1996)

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    16/25

    1414 BARBIERI ET AL.

    The second task (Lonzi & Luzzatti, 1995) aimed at verifying the patients ability

    to use prepositions in a sentence context, as for example Giovanni andr dal dottore

    per una visita (Giovanni will go to the doctor for a medical examination). The task

    was composed of 100 sentences which were read aloud by the patient and/or by the

    examiner, and the patient had to fill the blank with: (i) an article, as in Mario compraun regalo (Mario buys a present); (ii) a preposition, as in devi dare questo libro a

    Giovanni([you] must give this book to Giovanni);4 (iii) both a preposition and an art-

    icle as in ho visto una bottiglia nel frigo ([I] saw a bottle in the fridge); (iv) the task also

    included 16 fillers, in which there was nothing to add, as for instance la mamma con-

    vincer . . . Luisa (mother will convince Luisa).

    Prepositions in this task could be strictly subcategorised by the verb, i.e., introducing

    arguments, such as questo libro appartiene a Luigi(this book belongs to Luigi), or not

    strictly subcategorised, i.e., introducing adjuncts, such asprese il telegramma con pre-

    occupazione ([he] took the telegram with concern). Prepositions could also be part of

    a by-phrase in passive sentences, as in la lettera sar consegnata dal postino (the letterwill be delivered by the postman).

    In both tasks the participants had to decide whether the clause required a preposition an article to fill the blank or whether nothing had to be added. The examiner started

    the first task with three examples. L.M. could not take the second task because of his

    short-term memory impairment and his severe reading difficulties.

    The patients responses were classified into four possible types of error: inappro-

    priate addition, inappropriate omission, substitution of article/preposition, and yes,

    but I dont know. In the last category the patients understood that the blank

    required an article/preposition, but were unable to produce the appropriate element.

    Results

    Use of prepositions within noun phrases (prepositional compounds). Table 5 sum-

    marises the results of Experiment 2. Patients L.G. and F.C. easily identified

    noun-noun compounds (25 and 26 out of 27 respectively), could detect the preposi-

    tional compounds, and generally inserted the correct preposition: their errors were

    more often substitutions than omissions. L.M., on the contrary, frequently omitted

    prepositions (51 out of 60). A.C. produced very few correct responses and made a

    high number of omissions and substitutions. B.D. produced a majority of correct

    insertions (62 out of 87), and used the correct preposition in 58% of the cases; hiserrors were almost equally distributed between omissions and substitutions. He

    omitted the preposition in 14 items only, while in 46 cases his responses contained a

    preposition that was mostly correct.

    Use of prepositions in a sentence context. The three agrammatic participants (L.G.,

    F.C., and A.C.) were often able to supply prepositions (see Table 6) when required,

    but often not the correct ones: omissions, however, were rare (9 for F.C., 10 for A.C.,

    and 11 for L.G.). B.D., on the other hand, often managed to retrieve the correct

    4Contrary to the English verb to give, which can take a double argument structure, either transi-

    tive (+ complement) or ditransitive (Mary gives the book to John and Mary gives John the book), in

    Italian the verb dare takes only one argument structure, always requiring a prepositional phrase

    introduced by a (to).

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    17/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1415

    TABLE

    5

    Patientsretrievalofprepositionwithinprepositionalcompounds(Experim

    ent2)

    Correctresponses

    Abnormalinsertionof

    Omission

    of

    Substitutiono

    f

    Yesbutdontknow

    Patient

    Art

    Prep

    Both

    Art

    Prep

    Both

    Art

    Prep

    Both

    Correct

    Incorrect

    Total

    L.G.

    54

    5

    1

    1

    8

    1

    17

    87

    F.C.

    63

    1

    4

    2

    14

    3

    87

    L.M.

    26

    4

    1

    2

    49

    1

    4

    87

    A.C.

    5

    5

    9

    24

    7

    27

    10

    87

    B.D.

    62

    6

    8

    11

    87

    Art=insertion

    /omission/substitutionofthearticleonly.Prep=insertion/omission/substitutionoftheprepositiononly.Both=insertion/omission/substitutionofboth

    articleandpreposition.Yes,butdontknow=answer

    sinwhichthepatientcorrectlyreco

    gnisesthenecessityofinsertingap

    reposition(=correct).Incorrectreferstoitems

    thatdidnotrequireanyadditions.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    18/25

    1416 BARBIERI ET AL.

    preposition (72 items out of 84). His errors were predominantly substitutions (20),

    with very few omissions (3).

    The patients performances in the two tasks were compared by a paired t-test, in

    which the percentage of errors for each type (omissions, substitutions, additions, yesbut dont know) was compared across the two tasks. No difference was found

    between the two tasks (t = 0.267, ns).

    Discussion

    The results indicate that at least four of the five patients were often able to retrieve the

    correct prepositions, and only a few omissions were made in the task assessing the abil-

    ity to insert prepositions into sentences. No differences were found in the patients abil-

    ity to use prepositions in the two tasks. These results indicate that our patients had an

    almost preserved ability to employ prepositions and, in particular, were able to identifywhen a preposition had to be inserted in a sentence depending on the verb provided, and

    thus on the VAS, as in the sentences in Experiment 1. These results would appear to be

    in contrast with the interpretation of the errors made by aphasic patients in Experiment

    1 in terms of a specific impairment in retrieving prepositions.

    TABLE 6

    Retrieval of prepositions in a sentence context

    Subcateg. Not subcateg. Partially subcateg. Passive-da No prep. Total

    L.G.correct 18 7 5 3 13 46

    substituted 13 15 2 5 * 35

    omitted 6 2 1 2 * 11

    added * * * * 3 3

    yes, but dont know 3 2 0 0 0 5

    total 40 26 8 10 16 100

    F.C.

    correct 9 10 3 6 15 43

    substituted 18 6 3 1 * 28

    omitted 5 1 2 1 * 9

    added * * * * 1 1

    yes, but dont know 8 9 0 2 0 19total 40 26 8 10 16 100

    A.C.

    correct 17 7 5 0 9 38

    substituted 15 16 1 9 * 41

    omitted 5 2 2 1 * 10

    added * * * * 7 7

    yes, but dont know 3 1 0 0 0 4

    total 40 26 8 10 16 100

    B.D.

    correct 26 22 6 5 13 72

    substituted 10 3 2 5 * 20

    omitted 2 1 0 0 * 3

    added * * * * 2 2

    yes, but dont know 2 0 0 0 1 3

    total 40 26 8 10 16 100

    Results from L.M.s performance are not reported as he was unable to complete the task.

    Subcateg = subcategorised by the verb; Prep = preposition; Passive-da = passive-by.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    19/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1417

    GENERAL DISCUSSION

    The aim of the present study was to assess the ability of aphasic patients to employ

    verbs in an appropriate syntactic frame, i.e., using an appropriate argument structure.

    Argument structure specifies the number and type of participants in the event

    described by the verb, and several studies (Bonakdarpour, Thompson, & Fix, 2007;

    Kim & Thompson, 2000; Luzzatti et al., 2002; Thompson, 2003; Thompson et al.,

    1997) have proved the influence of this feature on verb processing in aphasic

    patients. According to models of lexical production (Levelt et al., 1999), information

    about verb argument structure is stored in the lemma, defined as a package of

    semantic and syntactic information concerning, for instance, grammatical gender of

    nouns and VAS. Clinical observation indicates that some fluent and non-fluent

    aphasic patients implement verbs in a sentence frame using an inappropriate argu-

    ment structure.

    In Experiment 1 we used a picture description task to investigate verb productionin sentences and found that five out of seven aphasic patients (four suffering from

    agrammatism and one from Wernickes aphasia) used verbs in an abnormal syntactic

    frame. These errors emerged both when patients were asked to describe the picture

    freely and when they had to complete the sentence frame by placing flash cards in the

    right position and inserting the correct preposition.

    Experiment 2 aimed at assessing the use of prepositions by patients who made a

    relevant number of verb argument structure errors, in order to exclude an alternative

    explanation of this deficit. Since most of these patients were suffering from non-fluent

    aphasia and agrammatism, some of their errors (for instance *la donna taglia la

    forbice [*the woman cuts the scissors]) could have been explained by the omission ofthe preposition introducing the manner adjunct. However, the results indicate that

    most patients tend to substitute prepositions rather than to omit them, even if

    required to insert them in a sentence context.

    To sum up, this study shows that fluent and non-fluent aphasic patients may retrieve

    the correct verbs but then use them with an improper VAS. In particular, they tend to

    retrieve a default argument structure such as that of the verb to use; indeed, although a

    patient (e.g., F.C.) made an error when he attempted to describe an image depicting a

    woman sewing a button with a needle with the words *la signora cuce lago per il bot-

    tone (*the lady sews the needle for the button), had he substituted the verb cuce

    (sews) with the verb usa (uses) the sentence would have been acceptable.Since most patients who made VAS errors did not omit prepositions either in prep-

    ositional compounds or in a sentence context (see Experiment 2), it can be claimed that

    these errors are not caused by difficulty in the use of prepositions, but by a deficit

    affecting the mapping of thematic roles onto the corresponding sentence argument

    structure. In addition, our results are in line with those obtained by Biran and Fried-

    mann (2008), who tested aphasic patients ability in processing verb argument

    structure by means of a verb completion, a grammatical judgement, and a sentence

    production task.

    As stated in the introduction, the verb production deficit in Brocas aphasia is

    explained either as an impaired lemma selection (Kim & Thompson, 2000), or as adeficit at the grammatical encoding level, i.e., during the creation of the syntactic

    environment (Bastiaanse & van Zonneveld, 2004). Even if our study does not directly

    address this issue, damage at the lemma level may be assumed in our patients. Specif-

    ically, they might encounter difficulties in accessing lemma information, i.e., in map-

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    20/25

    1418 BARBIERI ET AL.

    ping stored lemma thematic knowledge to the corresponding argument structure, or

    the lemma knowledge itself may be damaged. This latter case is in contrast with the

    strictly serial lemmalexeme access posited by Bock and Levelts model of lexical

    production (1994; see also Levelt et al., 1999). These authors assume that word pro-

    duction is a five-step process, in which the syntactic features included in the lemmaare always accessed before the phonological word form, i.e., the lexeme. Therefore,

    when producing a verb, information about its argument structure and thematic role

    assignment should be accessed before retrieving its phonological word form. How-

    ever, if a patient is able to retrieve the correct lexical label of a verb but employs it in

    an incorrect argument structure, it can be assumed that s/he retrieves the lexeme

    without having accessed the lemma. Given its strictly sequential lemma-lexeme

    order, Levelts model does not fully account for this type of error. The independent

    network modelof lexical access proposed by Caramazza (1997) appears to be most

    coherent with our data, in that the lexical-syntactic network is activated by both the

    lexical-semantic network and the phonological and orthographical representations(P-lexeme and O-lexeme). In addition, information concerning verb argument struc-

    ture is assumed to be fully activated only after the retrieval of the phonological/

    orthographic representation. Thus, in this theoretical framework, the damage in our

    patients could be located at the connection between the P-lexeme and the lexical-syn-

    tactic network: indeed, if this connection is disrupted, it is impossible to access all the

    information stored in the lexical-syntactic network, including the verb argument

    structure. Further investigation is needed in order to shed light on the locus of the

    deficit in these patients.

    Manuscript received 18 July 2009Manuscript accepted 21 December 2009

    First published online 9 August 2010

    REFERENCES

    Bastiaanse, R., & van Zonneveld, R. (2004). Brocas aphasia, verbs and the mental lexicon. Brain and

    Language,90, 198202.

    Berndt, R.S., Haedinges, A., Mitchum, C. C., & Sandson, J. (1997). Verb retrieval in aphasia. 2. Relation-

    ship to sentence processing. Brain and Language,56, 107137.

    Biran, M., & Friedmann, N. (2008). Syntactic information in the lexicon: Argument structure and grammatical

    gender in aphasia. Paper presented at the 9th Science of Aphasia Conference, Chalkidiki (Greece), Sept.2025.

    Bock, K., & Levelt, W. J. M. (1994). Language production: grammatical encoding. In M. A. Gernsbacher

    (Eds.), Handbook of psycholinguistics. San Diego, CA: Academy Press.

    Bonakdarpour, B., Thompson, C. K., & Fix, S. C. (2007). Neural signatures of verb argument structure in

    agrammatic aphasic and age-matched individuals. Brain and Language,103, 8485.

    Burzio, L. (1986). Italian syntax. Dordrecht, The Netherlands: Reidel.

    Caramazza, A. (1997). How many levels of processing are there in lexical access? Cognitive Neuropsychology,

    14, 177208.

    Chomsky, N. (1995). The Minimalist Program. Cambridge, MA: MIT Press.

    Chomsky, N., & Lasnik, H. (1991). Principles and parameters theory. In J. Jacobs, A. von Stechow, W.

    Sternefeld, & T. Venneman (Eds.), Syntax: An international handbook of contemporary research.

    Berlin: de Gruyter.Ciurli, P., Marangolo, P., & Basso, A. (1996). Esame del Linguaggio-II. Firenze, Italy: Organizzazioni

    Speciali.

    De Renzi, E., & Vignolo, L.A. (1962). The Token Test: A sensitive test to detect receptive disturbances in

    aphasics. Brain,85, 665678.

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    21/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1419

    Friedmann, N. (2000). Moving verbs in agrammatic production. In R. Bastiaanse & Y. Grodzinsky

    (Eds.), Grammatical disorders in aphasia: A neurolinguistic perspective (pp. 152170). London: Whurr.

    Friedmann, N. (2002). Question production in agrammatism: The tree pruning hypothesis. Brain and

    Language,80, 160187.

    Friedmann, N., & Grodzinsky, Y. (1997). Tense and agreement in agrammatic production: Pruning the

    syntactic tree. Brain and Language,56, 397425.Garrett, M. F. (1980). Levels of processing in sentence production. In B. Butterworth (Ed.), Language

    production (Vol. 1). New York: Academic Press.

    Jonkers, R. (1998). Comprehension and production of verbs in aphasic speakers. Groningen: Grodil.

    Kim, M., & Thompson, C. K. (2000). Patterns of comprehension and production of nouns and verbs in

    agrammatism: Implications for lexical organization. Brain and Language,74, 125.

    Levelt, W. J. M., Roelofs, A., & Meyer, A. (1999). A theory of lexical access in speech production. Behav-

    ioural and Brain Sciences,22, 175.

    Lonzi, L., & Luzzatti, C. (1995). Omissions of prepositions in agrammatism and the Universal Grammar

    constraint of recoverability. Brain and Language,51, 129132.

    Luzzatti, C., & De Bleser, R. (1996). Morphological processing in Italian agrammatic speakers: Eight

    experiments in lexical morphology. Brain and Language,54, 2674.

    Luzzatti, C., Raggi, R., Zonca, G., Pistarini, C., Contardi, A., & Pinna, G. (2002). Verbnoun doubledissociation in aphasic lexical impairments: The role of word frequency and imageability. Brain and

    Language,81, 432444.

    Luzzatti, C., Toraldo, A., Guasti, M. T., Ghirardi, G., Lorenzi, L., & Guarnaschelli, C. (2001). Compre-

    hension of reversibile active and passive sentences in agrammatism. Aphasiology,15, 419441.

    Luzzatti, C., Willmes, K., & De Bleser, R. (1994). Aachener Aphasie Test Versione Italiana. Firenze,

    Italy: Organizzazioni Speciali.

    Miceli, G., Burani, C., Capasso, R., & Laudanna, A. (1994). B.A.D.A. Batteria per lanalisi dei deficit

    afasici(2nd ed.). Roma: CEPSAG.

    Miceli, G., Silveri, M. C., Nocentini, U., & Caramazza, A. (1988). Patterns of dissociation in comprehen-

    sion and production of nouns and verbs. Aphasiology,2, 351358.

    Rochon, E., Laird, L., Bose, A., & Scofield, J. (2005). Mapping therapy for sentence production impair-

    ments in nonfluent aphasia. Neuropsychological Rehabilitation, 15, 136.

    Schwartz, M. F., Linebarger, M., Saffran, E. M., & Pate, D. (1987). Syntactic transparency and sentence

    interpretation in aphasia. Language and Cognitive Processes,2, 85113.

    Shapiro, L. P., & Levine, B. A. (1990). Verb processing during sentence comprehension in aphasia. Brain

    and Language,38, 2147.

    Thompson, C. K. (2003). Unaccusative verb production in agrammatic aphasia: The argument structure

    complexity hypothesis. Journal of Neurolinguistics,16, 151167.

    Thompson, C. K., Lange, K. L., Schneider, S. L., & Shapiro, L. P. (1997). Agrammatic and non-brain

    damaged subjects verb and verb argument structure production. Aphasiology,11, 473490.

    Thompson, C. K., Shapiro, L. P., Li, L., & Schendel, L. (1995). Analysis of verbs and verb argument

    structure: A method for quantification of aphasic language production. In P. Lemme (Ed.), Clinical

    aphasiology. Austin, TX: PE.

    APPENDIX 1

    Spontaneous speech samples

    The names of places and persons were changed in order to protect patients privacy.

    Patient L.G.

    Tell me about your day: what do you do?

    Oggi . . . nel la . . abbiamo no a . . (. . .) . ho fatto colazione . ho preso il caffelatte e . . fioc-

    chi e . . basta . Poi . . . i . . la Chiara . fisioterapia . . . . abbiamo abbiamo . . fatto un .1 2 3 4 5 6 7 8 9 . . e . . basta . Poi . la Teresa . . ho guardato . . i compiti . . e . . . . basta .

    Poi Vanessa . . logopedia . . . . leggere . scrivere . poi . . basta . . La sera . . guarda no .

    ho mangiato e . . . finito . . e . . . . . buonanotte (. . .)

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    22/25

    1420 BARBIERI ET AL.

    [Today . . . in the . . had1stPPl no . . . (. . .) . (I) had1stPSg breakfast . (I) had1stPSg coffee-

    with-milk and . . flakes and . . stop . Then . . . . . Chiara . physiotherapy . . . . (we)

    have1stPl (we) have1stPl . . done a . 1 2 3 4 5 6 7 8 9 . . and . . stop . Then . Teresa . . (I)

    have1stSg looked at. . the homework . . and . . . . stop . Then Vanessa . . speech ther-

    apy . . . . readInf . writeInf . then . . stop . . The evening . . watches no . (I) have1stSgeaten and . . . finished . . and . . goodnight (. . .)]

    Patient F.C.

    Tell me about your day: what do you do?

    Alla mattina . . a casa . . preparo la colazione . . . . . e . . i ragazzi . . . vanno a scuola

    e la . . Serena . a scuola lavoro . Io . . . ehm . . la mattina faccio i letti . . e . . . poi vado

    in bicicletta . . . . . . in bicicletta per . . . la carrozzeria . . e parlare del pi e del meno . . . gli

    amici per discorsi . . . (. . .) A casa mangio e . il telegiornale . Parlare del pi e del meno

    per tavola apparecchiata per tre . . la Elisa e il Davide (. . .)[In the morning . . at home . . (I) prepare1stSg the breakfast . . . . . and . . the kids . . .

    go3rdPl to school and . . Serena . to school work . I . . . uh . . the morning (I)

    do1stSg the beds . . and . . . then (I) go by bike . . . by bike . . . for . . . the garage . .

    and chatInf (lit. talkInf of-the-more-and-the-less) . . . the friends for speeches . . .

    (. . .) At home (I) eat and . the news . ChatInf(lit. talkInfof-the-more-and-the-less)

    but table setPstPrtFemSing for three . Elisa and Davide (. . .)]

    Patient N.C.

    Describe your family.

    Eh . . . . . . (a)spetta . . . . ecco eh . . . Luigi . . Marta . . . e . . Simona (. . .) Marta . . . eh . .

    . case . . ehm . . case . . . case (. . .) . . . eh . . . no . .. disegni . . . . . . clicca (. . .). E suo mar-

    ito? Pensione. E di che cosa si occupava? Lavorare . . . pi . . . pi . . . lavorare . . . . computer.

    [Uh . . . . . . wait . . . . well uh . . . Luigi . . Marta . . . uh . . Simona (. . .) Marta . . . eh . . .

    houses . . uh . . houses . . . houses (. . .) . . . uh . . . no . . . drawings . . . . . . clicks (. . .). And

    your husband? Retired. And what did he do? WorkInf. . . pi . .. pi . . workInf. . . computer. .

    Patient L.M.

    Tell me what happened to you at the beginning of your disease.

    E . . per adesso quando o . . o . . via e . . lui . . lui . . e . . ferma . Questo qui qualche cosa

    qualche cosa . . e . . questo questo tutti i giorni . E . . e . . gamba, la gamba non . . tutti i

    giorni tutti i giorni (. . .). Capisco o o . . Ciao come stai? Come stai? Tutti questo .

    Parlare parlare parlare per ogni giorno questo non riesco (. . .). E. . posso par . . e . .

    la baldorante . . e Debora scende perch chiedo chiedo, come si chiama, qualcosa perch

    chiu non cera e . . portato e . . portato portato portato e . .clinica, clinica (. . .)

    [And . . for now when or . . or . . away and . . he . . he . . and . . stops . This one something

    something . . and . . this this, every day . And . . and . . leg the leg not . . every day every

    day (. . .) (I) understand or or . . Hello how are you? How are you?All this . Talk InftalkInfbut every day I do not manage (. . .). And . . (I) can ta . . and . . the [neologism] . .

    and Debora comes down because (I) ask (I) ask, whats its name, something because

    there was not and . . taken and . . taken taken taken and . . clinic, clinic.]

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    23/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1421

    Patient A.C.

    Tell me what happened to you at the beginning of your disease.

    (. . .) Era una sera . . eh ha ho guardato la tiv . . no? In camera sul letto . e dopo un po,

    niente, . . cio . . aspetta . . poi va . . . ritorna in bagno no . . sempre un po che (. . .) . Alle

    luna e mezza . la mamma fina(l)mente ritorna in casa (. . .) e la mamma fa: cos cosho

    fatto?. E . e . io parlare cio . . . insomma . . . allora faccio schivi (scrivi) un attimo la penna.

    . .Aloro (Allora), ho letto, no, scri, un foglio, no tivo (scrivo) niente, cio pi-ri-tu (. . .).

    [(. . .) It was3rdSg an evening . . uh has (I) have1stSg watched the TV . . uh? In my bedroom,

    on the bed . and after a while, nothing, . . that is . . wait . . then goes, goes back to (the)

    bathroom, uh . . . always a while that (. . .) . At DETFemPl DETFemSg half past one . mum

    finally [phonemic paraphasia] comes back home (. . .) and mum says: whats that, what

    have1stSg I done?. And . and I. talkInfthat is . .well . well do1stSg write2ndSg a moment the

    pen. . .So, I read, no, wri, a paper, (I) write nothing, that is pi-ri-tu (. . .) ]

    Patient D.S.

    Tell me what happened to you at the beginning of your disease.

    (. . .) Eh cera un giorno sedici (. . .) sedici o o-ooo-obbre si eh . porto mia moglie a Novara

    no? Laaaa mia figlia, no? E dopo cera . c la tua figlia no? Io va . (. . .) C, c Maria?

    S s c:. e vai gi . su . (. . .) Dopo per (l)a ma(c)china, no? Vaio (Vado) a casa no?

    (. . .) E dopo cos io cos, no? (. . .)

    [(. . .) Uh there was a day sixteen (. . .) sixteen October yes uh . . I take 1stSg my wife to

    Novara, uh? There, my daughter, uh? And then there was . there is your daughter, uh? I

    goes (. . .) Is there, is there Maria? Yes, yes (she) is there; and go2ndSg down . up . (. . .)Then for the car, isnt it? Go1stSg home, uh? (. . .), and then so, I so, isnt it? (. . .)]

    Patient B.D.

    Describe a day in your life: what did you do yesterday?

    Alle sette . . . mi sono svegliato perch . . . il il nipote . . . porto dalle scuola con la

    macchina da Verdello a Dalmine verso le otto meno un quarto. Dopo faccio la colazione

    . . . e . . . prendo la macchina (. . .) . Verso le undici e mezza . . . vado la Gazzetta . . .

    pre preparo . . . il tavola e mangio (. . .) . Pomeriggio . . . guardo la televisione, bevo il

    caff . . . e dopo . . . prepara la cena . . . leggo la televisione un film (. . .)

    [At seven oclock . . . (I) woke upRefl+Aux+PstPrt because . . . the the grandson . . . (I)

    take1stSg from theFemPl school with the car from Verdello to Dalmine about a quarter

    to eight. Then (I) have1stSg breakfast . . . and . . . take1stSg the car (. . .) . About half

    past eleven. . . (I) go1stSg the Gazzetta (newspaper) . . . (I) prepare1stSg . . . theMscSgtable and eat1stSg. . Afternoon . . . (I) watch1stSg the television, drink1stSg coffee . . . and

    then . . . prepares the dinner; . . . (I) read1stSg the television a movie (. . .)]

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    24/25

    1422 BARBIERI ET AL.

    APPENDIX 2List of target sentences

    Agent Verb Theme Instrument

    Ex Luomo misura il quadro con il metro The man measures the picture with the metre

    Ex Luomo pesa il bambino con la bilancia The man weighs the baby with the scales

    1 Luomo pianta il chiodo con il martello The man embeds the nail with the hammer

    2 Luomo dipinge il muro con il

    pennello

    The man paints the wall with the brush

    3 Luomo scava la buca con la pala The man digs a hole with the spade

    4 La ragazza scrive la lettera con la penna The girl writes the letter with the pen

    5 La ragazza lega il ragazzo con la corda The girl ties up the boy with the rope

    6 La donna accende il televisore

    con il telecomando

    The woman switches on the TV with the remote

    control

    7 Luomo ascolta il CD con le cuffie The man listens to the CD with theheadphones

    8 La donna attacca il bottone con lago The woman sews the button with the needle

    9 La bambina bagna il pap con la canna The girl sprays her father with the garden

    hose

    10 Il cameriere stappa la bottiglia con il

    cavatappi

    The waiter opens the bottle with the

    corkscrew

    11 La donna lava i denti con lo spazzolino The woman brushes her teeth with the

    toothbrush

    12 La donna pulisce il pavimento con lo

    straccio

    The woman cleans the floor with the cloth

    13 La nonna lava la padella con la spugna The grand

    mother

    washes the pan with the sponge

    14 La donna benda il ragazzo con il

    fazzoletto

    The woman bandages the boy with the

    handkerchief

    15 La donna ripara il vaso con la colla The woman repairs the pot with the glue

    16 La donna taglia i capelli con la forbice The woman cuts her hair with the scissors

    17 Il bambino disegna la macchinina con

    la matita

    The child draws the toy car with the pencil

    18 La donna pela la patata con il coltello The woman peels the potato with the knife

    19 Luomo taglia le unghie con la forbice The man cuts his nails with the scissors

    20 Linfermiera pulisce la vetrina con lo

    strofinaccio

    The nurse cleans the display

    cabinet

    with the cloth

    21 La donna raccoglie la carta con la

    paletta

    The woman sweeps up the paper with the dustpan

    22 La donna trita laglio con la mezzaluna The woman dices garlic with the knife

    23 La donna asciuga i capelli con il phon The woman dries her hair with the

    hairdryer

    24 Luomo riempie la piscina con la canna The man fills the

    paddling

    pool

    with the garden

    hose

    25 La donna asciuga il piatto con lo

    strofinaccio

    The woman wipes the dish with the

    dishtowel

    26 La donna legge il libro con la lente The woman reads the book with the

    magnifying glass

    27 Il bambino tira il sasso con la fionda The child throws the stone with the sling28 La donna apre la porta con la chiave The woman opens the door with the key

    29 Il ladro colpisce i poliziotti con il

    bastone

    The thief hits the

    policemen

    with the stick

    (Continued)

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011

  • 8/6/2019 Argument Structure Deficits in Aphasia

    25/25

    ARGUMENT STRUCTURE DEFICITS IN APHASIA 1423

    APPENDIX 2

    (Continued)

    Agent Verb Theme Instrument

    30 La donna accende la candela con ilfiammifero

    The woman lights the candle with the match

    31 La donna versa lacqua con la brocca The woman pours water with (from) the

    jug

    32 Luomo gonfia la bicicletta

    con la pompa

    The man pumps up the bicycle

    (wheel)

    with the bicycle

    pump

    33 Il pittore dipinge la tela con il pennello The painter paints the picture with the brush

    34 Luomo mangia gli spaghetti con la

    forchetta

    The man eats the

    spaghetti

    with the fork

    35 La donna lucida gli stivali con la

    spazzola

    The woman polishes her boots with the brush

    36 Il dentista toglie il dente con la pinza The dentist pulls out the tooth with the pincers

    37 La donna taglia la carta con la forbice The woman cuts the paper with the scissors

    38 La donna sbatte luovo con la forchetta The woman whisks the egg with the fork

    39 La donna attacca la figurina con la

    colla

    The woman pastes the picture

    card

    with the glue

    40 Il pastore tira la pecora con la fune The shepherd pulls the sheep with the rope

    41 Luomo bagna i fiori con linnaffiatoio The man waters the

    flowers

    with the

    watering-can

    42 Il giardiniere raccoglie le foglie con il

    rastrello

    The gardener gathers the leaves with the rake

    43 Luomo aggiusta il rubinetto con la

    chiave inglese

    The man repairs the faucet with the monkey-

    wrench44 La donna beve il succo con la cannuccia The woman drinks the juice with the straw

    45 Luomo lava la macchina con la spugna The man cleans the car with the sponge

    46 Luomo costruisce laeroplanino con la

    carta

    The man builds the model

    plane

    with paper

    47 La donna assaggia la minestra con il

    cucchiaio

    The woman tastes the soup with the spoon

    48 Luomo rompe la tazza con il martello The man breaks the cup with the hammer

    49 Il muratore costruisce il muro con i

    mattoni

    The mason builds the wall with bricks

    50 Luomo scopisce la statua con martello

    e scalpello

    The man carves the statue with hammer and

    chisel

    51 La donna soffia il naso con il fazzoletto The woman blows her nose with thehandkerchief

    52 Luomo affetta il pane con il coltello The man slices the bread with the knife

    53 La donna sottolinea il libro con

    levidenziatore

    The woman highlights the book with the marker

    DownloadedBy:[TBTAKEK

    UAL]At:13:182March2011