TIME REFERENCE IN PATIENTS WITH NON-FLUENT AGRAMMATIC APHASIA: PRODUCTION AND COMPREHENSION.

University of Groningen

FACULTY OF ARTS

Master Neurolinguistics

TIME REFERENCE IN PATIENTS WITH NON-FLUENT

AGRAMMATIC APHASIA: PRODUCTION AND

COMPREHENSION.

Master’s Thesis:

Supervisor:

ALICE TORTORA

Dr.

ROEL JONKERS

2

Table of Contents

Acknowledgements

4

List of tables

5

List of figures

6

Abstract

7

Introduction

9

1.1 Time reference and discourse linking 9

1.2 Time reference in agrammatic aphasia 11

1.3 Time reference in Italian 15

1.4 The current study 17

2. Method

19

2.1 Participants 19

2.2 Materials 20

2.2.1 Spontaneous speech samples 20

2.2.2 TART – Production test 20

2.2.3 TART – Comprehension test 23

2.3 Procedure 23

2.4 Scoring and analysis 24

3. Results

25

3.1 Control subjects 25

3.2 Agrammatic patients 25

3.2.1 Quantitative analysis – TART production test 25

3.2.2 Qualitative analysis – TART production test 26

3.2.3 TART comprehension test 28

4. Conclusions and discussion

30

References

33

Appendices

36

3

4

Acknowledgements

I left for Italy on the 8th of December, with the hope of reaching my objectives. Without the love and support of my parents Renzo and Agnese Tortora, my sister Claudia, and Johan, it would have been impossible to bring this project to a good ending. I would like to thank my family for always believing in me.

A very important person during this last semester, has been my aunt Patrizia Tortora. As soon as she heard about my plans of coming to Italy to carry out this research for my Master’s thesis, she immediately offered to stay with her. Her intelligence, love and humor have been a great support, and the time we spent together, laughing, talking and – not to forget – brainstorming, in the famous ‘cucina della zia’, have rendered this adventure unforgettable. I want to thank her for her wisdom, love for life and her ability to listen and think along. I also want to thank my cousins Riccardo and Giulia for their loveand their determination to make me leave the house and the computer every now and then.

I would like to thank Dr. Roel Jonkers, who has guided me during my Bachelor’s thesis and finally during my Master’s thesis as well. During the past few years, I have had the pleasure of attending courses in Aphasiology lectured by Dr. Roel Jonkers. His way of teaching and his passion for this field of study enthused me and thanks to him I discovered my passion for Neurolinguistics. Thank you for your availability and your valuable insights.

Thanks to Dott.ssa Mariangela Taricco, head of the Dipartimento di Medicina Fisica e

Riabilitazione of Sant’Orsola-Malpighi hospital, who offered her collaboration and allowed me to test

patients from her department, I was able to carry out this research. A big thank you to Viviana, who was as determent as I was, to reach a minimum of 10 patients for the experimental group and who made all the phone calls and set up the appointments with the participants. I would also like to thank the speech therapists Silvia, Carla and Loredana, for their availability.

5

List of Tables

Table 1. Classification of regular Italian verbs 15

Table 2. Sample of the Italian conjugation 17

Table 3. Background information control subjects 19

Table 4. Background information agrammatic subjects 20

Table 5. Examples of TART production test items 21

6

List of Figures

Figure 1. Example of production test item 22

Figure 2. Example of comprehension test item 23

Figure 3. Correct scores in percentages (%) for all six elicited tense types in the TART 26 production test for the agrammatic subjects.

Figure 4. Correct scores in percentages (%) for each timeframe on the TART 28

7

Abstract

Reference to the past through verb morphology has been found to be particularly difficult for individuals with non-fluent agrammatic aphasia. Bastiaanse, Bamyaci, Hsu, Lee, Yarbay Duman and Thompson (2011) formulated the PADILIH, to account for this specific past time reference problem. The PADILIH predicts a specific past time reference deficit, as past time reference is discourse linked, while non-past time reference is processed through narrow syntax. Discourse linking needs more processing resources and it has been shown that discourse linking is more difficult for agrammatic individuals (Bastiaanse, Dragoy, Avrutin, Iskra & Bos, 2013).

In the present study, the Test for Assessing Reference of Time (TART; Bastiaanse, Jonkers & Thompson, 2008) was adapted for Italian. The aim of the present study was to investigate time reference through verb morphology, using the Italian version of the TART.

A total of 11 non-fluent agrammatic speakers and 10 non-brain-damaged individuals were tested using the Italian adaptation of the TART, which consists of a sentence completion task with photographs and a picture-matching task. Spontaneous speech samples were also collected, through a semi-structured interview and a picture description task. All responses on the TART tasks were tallied and analyzed, both quantitavely and qualitatively.

While the non-brain-damaged subjects performed at ceiling, the agrammatic individuals results showed a selective past time reference deficit in both production and comprehension. Simple past was found to be more impaired than present perfect. However, future time reference was found to be more severely impaired than present time reference. An error analysis revealed that the simple past was substituted significantly more often with non-past tense forms than with other past tense forms.

9

1. Introduction

1.1 Time reference and discourse linking

Time reference is a semantic feature of an event and can be expressed through using a verb

phrase. In many languages time reference is accomplished through temporal adverbs, such as

yesterday, now and tomorrow, through prepositional phrases such as a moment ago and next year

and through verb morphology.

A finite verb is a verb form that shows agreement with the subject and is inflected for tense (past, present or future) and number (singular or plural). Tense can be expressed through morphological inflection resulting in a finite verb. Tense contains information about the temporal relation between the time frame of the event and the time of the evaluation configured by the context. Three different time intervals have been discerned and are also referred to as event time,

speech time and reference time (Reichenbach, 1947). Temporal relations are further specified by aspect, providing information about the situational boundaries, like the beginning of the situation

and the end point. Aspect provides the listener with information about whether an action is ongoing (imperfect) or finished (perfect). Aspect is expressed through verb morphology in many languages.

There is a lot of cross-linguistic variation in the different ways to express tense and aspect. Tense and aspect can be inflected on a single lexical verb, called a simple verb form or by using a

periphrastic verb form, which is a verb form consisting of an auxiliary with a lexical verb. In some

languages, such as Chinese, verbs are not inflected for tense nor aspect and, therefore, time reference is not accomplished through verb inflection. Instead, time reference is expressed by

aspectual adverbs in languages without verb inflection, such as Chinese.

Different syntactic mechanisms are involved in sentence processing. Avrutin (2000) made a distinction between discourse linked processes and non-discourse linked processes. According to Avrutin (2006) agreement is a purely morphosyntactic mechanism that can be processed by narrow

syntax. Narrow syntax is a computational system that operates at the level of the linguistic

information that is given within a sentence. Agreement is non-referential and establishes local binding relations, thus rendering extra-sentential or discourse information irrelevant. Agreement is therefore not discourse linked. Tense, however, goes beyond narrow syntax processing, because it requires an integration of syntactic information and discourse-related knowledge. A connection needs to be made between the different timeframes and the information structure of the sentence, hence, according to Avrutin (2000, 2006) tense requires discourse linking.

10 example given in (1a). To process an anaphoric sentence like (1a), where the reflexive pronoun ‘himself’ refers to ‘the boy’ in the same sentence, the information within the sentence is sufficient and no discourse linking is needed.

(1) a. The boyi washes himselfi

b. The boyi washes himj

According to Avrutin (2000, 2006) discourse linking is needed to process sentences with a referential element such as the pronoun ‘him’ in the example given in (1b), where the pronoun refers to an extra-sentential entity.

Agreement and tense differ in the same way. Agreement is a locally bound relation that is established within the sentence, as agreement occurs between the subject and the verb. Extra-sentential information is irrelevant. The processing of tense, however, needs discourse information. An example of a sentence where agreement is processed with solely the syntactic information within the sentence and tense is processed using discourse information, is given in (2) (taken from Avrutin, 2000).

(2) ‘In 1917, Lenin comes to power.’

The sentence in (2) is syntactically well-formed, but there is a mismatch between the tense morphology and the timeframe the sentence refers to. To establish agreement between the subject and the verb, only the information structure of the sentence is needed. Discourse information is needed to process the tense and timeframe mismatch in (2). In this case the discourse information consists of the knowledge, that in historical narration present tense is often used to refer to past events.

11 discourse linking is needed. Past time reference lacks this conjuncture between these time intervals and, hence, past tense needs to be discourse-linked.

1.2 Time reference in agrammatic aphasia

There has been extensive research on the language production and comprehension problems of patients with agrammatic aphasia. Data are available from various comprehension and production tasks, as well as from the analysis of spontaneous speech samples.

The spontaneous speech of patients with agrammatic aphasia is non-fluent and is characterized by frequent omissions of function words (Menn & Obler, 1990). Inflectional morphology is frequently omitted or substituted (Penke, Janssen en Krause, 1999). The reduced speech fluency results in a shorter Mean Length of Utterance (MLU: Rossi & Bastiaanse, 2008). Bastiaanse and Jonkers (1998) found that, in contrast to previous findings, the number of verbs produced by the agrammatic patients in their study was normal. However, the diversity of the lexical verbs produced by the agrammatic patients was lower than the control subjects. The agrammatic patients in their study showed a high production of verbs with one or no internal argument and a lower proportion of finite verbs in their spontaneous speech production. Bastiaanse, Hugen, Kos and Van Zonneveld (2002) investigated the production of verbs in patients with agrammatic aphasia and observed that the patients produced a lower proportion of finite verbs than the control subjects that participated in their study. In line with these findings, Bastiaanse (2008) concluded from her research that the production of finite verbs is more difficult than the production of nonfinite verbs. Furthermore, Bastiaanse concluded that past tense was more difficult than present tense, irrespective of finiteness of the verb. A comparison between the finite and nonfinite past verb forms and the finite and nonfinite present verb forms resulted in the conclusion, that reference to the past in general is more problematic than present time reference. A selective past time reference deficit in agrammatic aphasia had previously been observed for Hebrew and Arabic (Friedmann & Grodzinsky, 1997), for Greek (Stavrakaki & Kouvava, 2003) and for German (Burchert, Swoboda-Moll & De Bleser, 2005). In addition, the perfective aspect has been found to be more severely impaired as opposed to the imperfect aspect of verbs (Stavrakaki & Kouvava, 2003; Nanousi, Masterson, Druks & Atkinson, 2006).

12 a time reference deficit, rather than a tense problem. The disentanglement of tense and time reference has been further elaborated by Bos, Dragoy, Stowe and Bastiaanse (2013).

Bos et al. presented neurophysiological evidence for the distinction between tense and time reference. They found that the brain responses of their non-brain-damaged participants on time reference violations by simple and periphrastic verb forms were similar, irrespective of tense. In line with the findings from Yarbay Duman and Bastiaanse (2009) Bos et al. (2013) argued that the past time reference deficit seen in agrammatic patients is irrespective of tense.

To account for the specific past time reference deficit in agrammatic aphasia, Bastiaanse, Bamyaci, Hsu, Lee, Yarbay Duman and Thompson (2011) formulated the PAst DIscourse LInking Hypothesis (PADILIH), based on the theories of Avrutin (2000, 2006) and Zagona (2003). Avrutin argued that agrammatic speakers have insufficient processing resources to enable an interaction between narrow syntax, which establishes inter-sentential relations, and discourse-linking, which is used to process extra-sentential information. Narrow syntax needs less processing resources compared to discourse-linking. Avrutin stated that agreement can be processed through narrow syntax, while tense is discourse-linked. Therefore, agrammatic speakers will have more difficulties in tense processing, because there are insufficient resources to process discourse-linking. However, Zagona (2003) suggested a difference between processing past time reference and present time reference. Zagona argued that present time reference should be considered as a type of local binding relation, due to the simultaneity between event time and evaluation time, while in past time reference these time intervals do not coincide. According to Zagona (2003), past time reference processing therefore needs discourse linking, whereas present tense may be processed through narrow syntax.

Bastiaanse et al. (2011) rephrased Avrutin’s and Zagona’s theories, by stating that past time reference through verb inflection needs discourse linking. Therefore, verb forms referring to a past timeframe, whether through simple verb forms or through periphrastic verb forms, are more complex than those referring to a present of future timeframe. This reformulation resulted in the PADILIH, which predicts a selective past time reference deficit in patients with non-fluent agrammatic aphasia. In line with the PADILIH, Dragoy, Stowe, Bos and Bastiaanse (2012) presented neurophysiological evidence that past tense processing needs discourse linking and that present tense processing is established through local binding relations. Discourse linking in particular has proven to be impaired in agrammatic patients (Bastiaanse, Dragoy, Avrutin, Iskra & Bos, 2013).

13 grammatical morphology associated with time reference and includes a sentence completion task with photographs and a picture-matching task. The TART has been adapted for a large number of languages and has been used in several cross-linguistic studies.

Bastiaanse et al. (2011) examined the production and comprehension of time reference through grammatical morphology in English, Turkish and Chinese agrammatic patients using the TART. English is a language like Dutch, which has a combination of simple verb forms and periphrastic verb forms to express time reference. Turkish is an agglutinative language that expresses time reference predominantly through simple verb forms. Chinese uses free grammatical morphemes to refer to a certain timeframe. The comprehension data from Bastiaanse et al. (2011) showed the a specific past time reference deficit for all three languages, as the production data only showed this specific past time reference deficit for Turkish and English. In Chinese, time reference is established through the use of optional aspectual adverbs. In the Chinese version of the TART, the aspectual adverb along with a temporal adverb were mentioned in the first sentence. In the target sentence, the temporal adverb was repeated, which is why the Chinese agrammatic speakers frequently omitted the aspectual adverb is all three timeframes.

In order to circumvent this problem, the TART was adapted for Indonesian (Anjarningsih, 2012), a language in which tense and aspect are not inflected through verb morphology and the use of aspectual adverbs is optional. In Indonesian, aspectual adverbs are only used when the event time cannot be deduced from the discourse. This would imply that aspectual adverbs are referential for every timeframe, hence, all aspectual adverbs in Indonesian must be processed through discourse linking. Therefore, Anjarningsih predicted that all three timeframes should be equally impaired in agrammatic Indonesian speakers, as discourse linking seems to be problematic for agrammatic individuals in general, according to the PADILIH. In fact, no significant differences in agrammatic performance were found between the three timeframes in Indonesian and hardly any omission errors were made.

14 Dragoy and Bastiaanse (2013) examined time reference in Russian agrammatic individuals using the TART. The results from their study confirmed a specific past time reference deficit in patients with agrammatic aphasia. Interestingly, reference to the non-past was better preserved only for imperfective verbs. For perfective verbs, the agrammatic patients performed better in reference to the past than in reference to the non-past. Dragoy and Bastiaanse found an interaction between time reference and aspect. They suggested a prototypical and therefore less resource-consuming use of imperfective and perfective aspect. As imperfective aspect is prototypically associated with non-past time reference and perfective aspect is prototypically associated with non-past time reference, perfectiveness does not necessarily have a negative effect on the performance of Russian agrammatic speakers.

Time reference in Spanish and Catalan agrammatic speakers was studied by Martínez-Ferreiro and Bastiaanse (2013) using the Spanish and the Catalan versions of the TART. In accordance with the PADILIH, past time reference was found to be more severely impaired than non-past time reference in both production and comprehension. In line with previous findings, Martínez-Ferreiro and Bastiaanse concluded that perfective aspect was more difficult than imperfective aspect (Stavrakaki & Kouvava, 2003; Nanousi et al., 2006; Yarbay Duman & Bastiaanse, 2009). However, a dissociation in performance was found between present and future time reference, at least in the comprehension data. The agrammatic patients in this study performed worse in the future tense condition than in the present tense condition. Nanousi et al. (2006) reported similar results for Greek. This dissociation was not predicted by the PADILIH, as it is suggested that future tense and present tense are similarly processed through discourse linking.

Rofes, Bastiaanse and Martínez-Ferreiro (2014) extended the PADILIH to the conditional tense and future tense in Catalan testing again non-fluent agrammatic patients. A comparison was made between periphrastic conditional tense, which refers to an irrealis past event (e.g. ‘if the man had had time, he would have …’), simple conditional tense, which refers to a irrealis future event (e.g. ‘if the man had time, he would …’), and future tense. The periphrastic conditional tense was found to be more difficult than simple conditional tense and future tense. Rofes et al. concluded that a specific past time reference deficit is also present when referring to irrealis event times.

15 speakers. Within the past timeframe, the simple past tense was more impaired than the present perfect tense. Further error analysis showed that the agrammatic subjects produced more infinitive verb forms and other verb forms without time reference characteristics. Furthermore, if the agrammatic aphasic subjects remained in the correct timeframe, they preferred to use the periphrastic verb form to refer to the past instead of the simple past.

Time reference involves an interaction between tense, aspect and discourse. Past time reference seems to be more difficult for agrammatic speakers, as it needs to be processed through discourse linking, whereas reference to a present or future timeframe is processed through narrow syntax.

1.3 Time reference in Italian

Italian is a Romance (Indo-European) language with a rich system of inflectional morphology. In the case of regular verb morphology, inflectional morphemes expressing tense, mood, and number-person agreement are attached to the verb stem (Dardano & Trifone, 1995,1997; Mosca, 2011).

Table 1.: Classification of regular Italian verbs with examples of verb inflection

Verb ending Verb 3SG-PRES.

3SG-PAST-IMP-MOOD:conj. 1st Conjugation -are lodare egli loda che lui lodasse 2ndConjugation -ere temere egli teme che lui temesse 3rd Conjugation -ire partire egli parte che lui partisse

Table 1.: A summary of the classification of regular Italian verbs is given, showing the verb ending of the infinitive. Examples of infinitive verb forms for each of the three conjugations are given. Examples of verb inflection are

given, showing how person (3rd person) number (SG for singular), tense (PRES. for present tense or PAST for past

tense) aspect (IMP for imperfective aspect) and mood (MOOD:conj. for conjunctive mood) are inflected on the verb.

16 conjugation). A representation of the three regular verb conjugations is given in Table 1, including examples of how person, number, tense, aspect and mood are inflected on regular verb forms.

A morphologically uniform paradigm has either derived inflectional forms or underived inflectional forms, exclusively (Jaeggli & Safir 1989). Italian is a stem-based morphology language where all inflectional rules are applied on the verb stem and has therefore only derived inflectional forms. The inflection paradigm of such a language does not contain word forms without inflectional suffixes, therefore the stem does not occur as a free morpheme (Aronoff, 1994). In the case of a stem-based morphology language, omission of the inflectional suffix would result in a non-word (Grodzinsky, 1990).

The neutral word order Italian is Subject-Verb-Object (SVO). However, Italian permits a large variation in word order, due to a rich system of agreement markers. As a result, all possible sentence structures can and do occur in Italian under particular pragmatic circumstances (Bates, 1976). An example of the variation in sentence order in Italian are SOV sentence structures occuring under certain pragmatic conditions (Velupillai, 2012). These SOV structures occur under the influence of

pro-drop. Pro-drop, being a linguistic feature where a pronoun is pre-cliticized to the verb, is only

possible in languages where the paradigm is morphologically uniform, like Italian (Jaeggli & Safir, 1989). The rich system of agreement markers in Italian permits deviations from the neutral word order. An example of pre-cliticization in Italian is given in (2).

(2) Lo studente ha visto il professore. [The student has seen the professor]

S V O

Lo studente l’ha visto.

[The student him has seen] pre-cliticization S O V

17 Table 2.: Sample of the Italian conjugation

Tense Time reference

Infinitive Scrivere

Simple Present Scrive Present Present

Simple past scriveva Past Past

Present Perfect Ha scritto Present Past

Past Absolute (Passato Remoto) Scrisse Past Past

Simple Future Scriverà Futuro Futuro

Table 2.: Sample of the Italian conjugation. Note that while in Simple past and Past Absolute tense is inflected on the verbal root, only in Present Perfect tense a discrepancy may be observed between the tense form and the timeframe it refers to.

In Italian grammar the simple past is used to express the duration or the repetition of an action in a past timeframe. Simple past tense has no clear temporal boundaries and therefore the distance between the timeframe expressed through simple past tense and the speech point is not evident. Present perfect tense expresses an action with a beginning and ending point in the past. The action is completely concluded in a past timeframe, but is closely connected to the present. Past absolute tense expresses a completely concluded action in the past and indicates a timeframe with the greatest distance from the present compared to the other two past tenses used in this study.

1.4 The current study

The aim of the present study is to investigate time reference in Italian using the Italian adaptation of the Test for Assessing Reference of Time (TART; Bastiaanse, Jonkers & Thompson, 2008), which includes a sentence completion task with photographs and a picture-matching task. The TART was developed to examine the production and comprehension of grammatical morphology associated with time reference and has been used for cross-linguistic research.

The PADILIH predicts a specific past time reference deficit, irrespective of the tense of the verb. Following these predictions, Italian agrammatic aphasic speakers will show a specific impairment in past time reference through both simple and periphrastic verb forms, whereas reference to the non-past will be relatively spared, in both production and comprehension.

18 to refer to the past and imperfect aspect is prototypically used to refer to the non-past, and in particular to the present (Dahl, 1985). Prototypical use of aspect needs limited processing resources, hence, imperfective aspect will be more difficult than perfective aspect when referring to the past. When referring to the past, Italian agrammatic speakers will prefer the present perfect, as opposed to the simple past.

Reference to the present can be processed through narrow syntax, whereas past time reference is processed through discourse linking, which will be particularly problematic for agrammatic speakers. It is however unclear which mechanisms are involved in processing future time reference. Abuom and Bastiaanse (2013) suggested that future time reference is neither discourse linked nor locally bound, as future tense concerns an irrealis event time. Future tense is frequently merged with present tense and labeled as non-past and subsequently compared to past time reference for both production and comprehension. Abuom and Bastiaanse (2013) did not find any significant differences in agrammatic performance between future tense and present tense, in neither production nor comprehension. Martínez-Ferreiro and Bastiaanse (2013), however, did report significant worse performance in future time reference compared to present time reference, but did not offer a clear explanation for this difference. In the current study, the imperfective future tense or the simple future is tested. As pointed out above, imperfective aspect is prototypically associated with non-past time reference, hence, no differences are expected in agrammatic performance between present and future time reference, in neither production nor comprehension.

19

2. Methods

2.1 Participants

A group of ten non-brain damaged native Italian speakers participated in the experiment. These control subjects were all between 48 and 62 years of age, right-handed and had no history of speech or language impairment. Relevant background information on the control subjects is given in Table 3.

Table 3.: Background information control subjects

Subject Gender Age Education Handedness

C1 F 61 13 R C2 F 53 13 R C3 F 59 23 R C4 F 57 15 R C5 F 49 16 R C6 M 48 15 R C7 F 57 16 R C8 F 54 15 R C9 M 51 11 R C10 F 53 17 R Means 54;2 15;5

Table 3.: Gender is indicated with M (male) or F (female); education is given in years; handedness is indicated with R (right-handed) or L (left-(right-handed); mean age and mean education are given in years;months.

11 Italian agrammatic aphasic speakers participated in this study. The aphasic individuals, six males and five females, ranged between 35 and 81 years of age with a mean age of 70;6 and were recruited from the Dipartimento di Medicina Fisica e Riabilitazione of Sant’Orsola-Malpighi Hospital in Bologna, Italy. All patients were right handed and had no severe visual or hearing impairments. The agrammatic individuals were classified as having Broca’s aphasia using the Italian version of the

Aachener Aphasia Test (AAT; Luzzatti, Willmes & De Bleser, 1994). Patient A8 was not classified as

20 Table 4.: Background information agrammatic subjects

Subject Gender Age Education Handed- ness

Etiolology HS TPO Aphasia Classification

A1 M 78 5 R iCVA L 3 Broca’s aphasia

A2 M 77 18 R iCVA L 11 Broca’s aphasia

A3 M 73 5 R hCVA L 11 Broca’s aphasia

A4 M 77 8 R hCVA L 12 Broca’s aphasia

A5 F 81 4 R iCVA L 3 Broca’s aphasia

A6 F 70 5 R iCVA L 36 Broca’s aphasia

A7 F 76 5 R iCVA L 34 Broca’s aphasia

A8 M 75 5 R iCVA L 38 Non-fluent aphasia

A9 M 54 5 R hCVA L 58 Broca’s aphasia

A10 V 35 14 R hCVA L 71 Broca’s aphasia

A11 V 79 5 R iCVA L 51 Broca’s aphasia

MEANS 70;6 7;2 30

Table 4.: Gender is indicated with M (male) or F (female); education is given in years; handedness is indicated with R (right-handed) or L (left-(right-handed); the type of Cerebral Vascular Accident (CVA) is indicated with iCVA, (ischemic CVA) or hCVA (hemorrhagic CVA); localization of lesion is indicated with L (left hemisphere); Time Post Onset (TPO) at the time of testing is indicated in months post onset; Aphasia classification according to the Italian version of the Aachener Aphasia Test (AAT; Luzzatti et al., 1994). Mean age and mean education are given in years;months; mean TPO is rounded to months.

2.2 Materials

2.2.1 Spontaneous speech samples

A spontaneous speech sample was obtained by means of a semi-structured interview, which included the questions from the Italian version of the Aachener Aphasia Test (AAT; Luzzatti et al., 1994). In order to obtain a spontaneous narrative speech sample, all subjects were also presented with the ‘Flood Rescue’- picture (Rubin & Newton, 2001). When presented with this picture, the participants were asked three questions, namely, What happened?, What is happening now? and

What will happen next? These questions enabled the elicitation of different timeframes in the

spontaneous speech. An example of a spontaneous speech sample is given in Appendix A.

2.2.2 TART production test

21 research. 11 pairs of transitive verbs were included, of which five verbs with irregular tense inflection. The production test contains 126 items, including six practice items. Each verb pair selects the same object, for example to fill/to empty the box; to iron/to fold a shirt. A complete list of the verbs used in the current study is given in Appendix B.

To elicit verb forms referring to the present, past or future, the patients were presented with colored photographs depicting either an ongoing action, a completed action or the intention to commence an action, respectively.

Table 5.: Examples of TART production test items

Tense Example

Infinitive (Infinito)

Qui la ragazza vuole cucire la maglia. [Here the girl wants to sew the sweater]

Simple present (Presente)

Adesso la ragazza cuce la maglia. [Now the girl sews the sweater]

Simple past (Imperfetto)

Poco fa la ragazza cuciva la maglia. [Short ago the girl sewed the sweater]

Present perfect

(adjacent past)

(Passato prossimo)

Ieri la ragazza ha cucito la maglia. [Yesterday the girl has sewn the sweater]

Past absolute

(distant past)

(Passato remoto)

Anni fa la ragazza cucì la maglia.

[Years ago the girl <to sew- 3sg PAST ABS.> the sweater]

Simple future (Futuro semplice)

Dopo la ragazza cucirà la maglia.

[Later the girl <to sew-3sg FUT> the sweater]

22 The production test follows a sentence completion paradigm were the intended verb forms are primed. To elicit different timeframes, different temporal adverbs were used, such as yesterday,

now and soon. The patients were presented with two photographs depicting different actions, but in

the same timeframe. In order to circumvent possible word finding problems, the infinitive verb form corresponding to the depicted action was printed above each picture. The experimenter read aloud a sentence corresponding to the photograph on the left and the first part of the second sentence that the participants were expected to complete using the photograph on the right. To elicit the same verb forms, the partial sentence corresponding to the second photograph consisted of the same temporal adverb and subject the experimenter produced in the first sentence. An example of a test item is given in (3).

(3) Adesso il ragazzo versa il latte. Adesso il ragazzo ... Subject: ‘beve il latte’ [Now the man pours the milk] [Now the man … (drinks the milk)]

Five different temporal adverbs were used to elicit five different tenses, namely the simple present,

simple past, past perfect, past absolute and the simple future. The elicitation of infinitive verb forms

was established through the use of the modal periphrasis ‘want to + infinitive’. Instead of a temporal adverb, the locative adverb ‘here’ was used, to ensure uniformity in sentence length and structure.

Fig. 1: The pair of photographs with the verbs to pour and to drink, used for eliciting the target

23

2.2.3 TART comprehension test

The comprehension test consists of a sentence-to-picture matching task. The same verb pairs from the production test were used for the comprehension test. Each verb was used in past, present and future tense. The comprehension task includes 60 items with 20 items for each timeframe. For each target sentence, the participants were presented with two photographs. Both photographs depicted the same action, but in two contrasting timeframes. Only present-past and future-past contrasts were tested, to avoid possible ambiguities. The experimenter spoke aloud a sentence that contained the target tense inflection, after which the participants were requested to indicate the photograph that corresponded to the sentence. An example of a test item is given in figure 2.

Fig.2: Example of a comprehension test item; two photographs depicting the same action in contrasting

timeframes. The target sentence was The man will glue the paper.

2.3 Procedure

All participants were tested individually. The test sessions took place in a speech therapy room at Sant’Orsola-Malpighi Hospital in Bologna, Italy. The participants were asked to accommodate themselves in front of a laptop computer screen (Toshiba, 15 inch). All test sessions were recorded with a digital voice recorder (Sony). Before conducting the TART, a spontaneous speech sample was obtained and recorded. Spontaneous speech was elicited by means of a semi-structured interview and a narrative task. The TART was administered subsequently.

The TART was administered in the same order in all subjects: all participants were presented with the comprehension test only after completing the production test. All control subjects completed the test in one single session, which lasted between 30 and 35 minutes, including a break if necessary. The aphasic subjects, completed the test in one or two sessions. Each session lasted between 45 and 60 minutes, including a break every 15 minutes.

24 feedback was given to the participants. No further feedback was given during the test procedure and the test items were only repeated if the subjects requested this explicitly.

2.4 Scoring and analysis

Responses on the TART production test were scored both quantitatively as qualitatively and were scored as correct when the target tense form was produced. In the production test the neutral SVO structure was elicited. However, pre-cliticization is a common linguistic phenomenon in Italian, which results in the production of SOV-structures. These structures have therefore not been counted as errors in the production test as long as the target verb forms were produced. Omissions of substitutions of the object in the TART production test, were not scored as errors. All errors produced by the subjects were categorized and tallied. Responses on the TART comprehension test were scored as correct if the subjects indicated the target picture. Subjects were allowed to correct themselves, in which case the second response was scored.

25

3. Results

3.1 Control subjects

All control subjects scored at ceiling on both the TART production test and on the TART comprehension test. The number of correct responses on the TART production test for the control subjects ranged between 107 (89,2%) and 120 (100%), with a mean overall correct score of 118 (98,3%). The control subjects’ number of correct responses on the TART comprehension test ranged between 56 (93,3%) to 60 items (100%), with a mean overall correct score of 59,2 (98,7%). The number of correct responses for every elicited tense in the TART production test, is given in Appendix C for all control subjects. An overview of the control subjects’ scores on the TART comprehension test is given in Appendix D, with the correct scores given for all three timeframes elicited in the TART comprehension test.

3.2 Agrammatic patients

3.2.1 Quantitative analysis – TART production test

All correct scores and errors on the TART production test per agrammatic subject are given in Appendix E. The overall correct scores for every elicited timeframe in the TART production test are given in Appendix G.

An independent samples t-test indicated that the agrammatic subjects (M = 66.82, SD = 15.600) performed significantly worse on the TART production test than the control subjects (M = 118.00, SD = 4.082), t(19) = -10.045, p = .022. Regularity effects of Italian past tense were examined. An independent samples t-test showed no difference in agrammatic performance between regular (M = 56.58, SD = 13.551) and irregular verbs (M = 55.15, SD = 12.231), t(20) = .261, p = .797.

26 tense (M = 10.64, SD = 4.388, p = .041). The correct scores for each tense condition are graphically represented in figure 3.

Fig. 3: Overall correct scores for the six elicited tenses in the TART production test, given in percentages for the agrammatic subjects.

A comparison was made between the three timeframes that were elicited in TART-production. The past timeframe consisted of simple past, present perfect and past absolute tense and was compared with the present and future timeframe. A significant difference was found between the three timeframes at the p <.05 level (F(2,52) = 4.735, p = .013). Post hoc analysis of the three timeframes using the Tukey HSD test only indicated a significantly higher correct score in the present timeframe (M = 13.36, SD = 4.007) compared to the past timeframe (M = 8.21, SD = 5.337, p = .011).

Correct scores on past imperfective aspect (simple past) and past perfective aspect (present perfect and past absolute) were compared. Present perfect and past absolute were grouped together and compared to simple past tense. However, an independent samples t-test did not show a significant difference between imperfective (M = 9.55, SD = 5.956) and perfective aspect (M = 7.55, SD = 5.012), t(31) = 1.015, p = .318.

3.2.2 Qualitative analysis – TART production test

The errors that occurred in the TART production test were tallied and categorized. A summary of all errors produced by the agrammatic individuals in the TART production test is given in

27 Appendix I. A summary of the substitution errors given in numbers and percentages of the total number of errors is given in table 6 for the agrammatic group.

A total number of 582 errors were counted in the TART production test, which consisted exclusively of substitution errors. 51,9% (302 errors) of the errors consisted of substitutions with infinitive verb forms. In 17,2% (100 errors) of the cases the elicited tense was substituted with simple present tense. 9,3% (54 errors) of the substitutions with simple present tense occurred when past tense was elicited.

Table 6.: Substitution errors

Target verb form

Non-past Past

INF SPr FU IMP PP PRe _Total

Substitutions INF - 59 (10,1%) 48 (8,3%) 66 (11,3%) 63 (10,8%) 66 (11,3%) 302 (51,9%) Non-past SPr 14 (2,4%) - 32 (5,5%) 22 (3,8%) 10 (1,7%) 22 (3,8%) 100 (17,2%) FU 1 (0,2%) 7 (1,2%) - 9 (1,5%) 4 (0,7%) 3 (0,5%) 24 (4,1%) total 15 (2,6%) 7 (1,2%) 32 (5,5%) 31 (5,3%) 14 (2,4%) 25 (4,3%) 124 IMP 4 (0,7%) 3 (0,5%) 9 (1,5%) - 22 (3,8%) 55 (9,5%) 93 (16%) Past PP 1 (0,2%) 2 (0,3%) 4 (0,7%) 10 (1,7%) - 15 (2,6%) 32 (5,5%) PRe 0 2 (0,3%) 9 (1,5%) 8 (1,4%) 10 (1,7%) - 29 (5%) total 5 (0,9%) 7 (1,2%) 22 (3,8%) 18 (3,1%) 32 (5,5%) 70 (12%) 154 Other 0 0 0 1 (0,2%) 1 (0,2%) 0 2 (0,3%) Total 20 73 102 116 110 161 582 Total % 3,4% 12,5% 17,5% 20,0% 18,9% 27,7% 100%

Table 6.: Substitution errors given in number and percentages of the total number of errors produced by the agrammatic group. INF = infinitive, SPr = simple present, FU = future, IMP = simple past, PP = present perfect and PRe = past absolute tense.

28 (1,7%) with present perfect tense and 8 times (1,4%) with past absolute tense. Future tense was substituted 32 times (5,5%) with present tense and 22 times (3,8%) with a past tense form.

The substitution errors were also analyzed in terms of past and non-past substitution. The substitution errors made in past time reference were analyzed, by merging infinitive, simple present and future tense substitution errors and labeling them as past substitution. The number of non-past substitution errors were analyzed for the three elicited non-past timeframes. Simple non-past tense was substituted more often with a non-past tense form (31 times) than with a past tense form (18 times), whereas both present perfect and past absolute tense were substituted more often with another past tense form, rather than with a non-past tense form. Substitution with a non-past tense form occurred 14 times in present perfect tense and 25 times in past absolute tense. Substitution with a past tense form occurred 32 times in the present perfect tense condition and 70 times in the past absolute tense condition.

3.2.3 TART comprehension test

The individual scores in the TART-comprehension are given in Appendix F. The overall scores per subject with the appurtenant percentages for all three elicited timeframes in TART-comprehension are summarized in Appendix H.

Fig. 4: Overall correct scores in percentages on the TART comprehension test. Significant differences between the three

elicited timeframes in TART-comprehension for agrammatic comprehension are indicated with an asterisk (*).

An independent samples t-test indicated that the agrammatic subjects (M = 45.18, SD = 6.838) performed significantly worse on the TART comprehension test than the control subjects (M =

0,0 20,0 40,0 60,0 80,0 100,0

PAST PRESENT FUTURE

%

Agrammatic subjects

29 59.20, SD = 1.398), t(19) = -6.349, p = .001. Regularity effects of Italian past tense were examined. An independent samples t-test showed no difference in error-percentages between regular (M = 26.336, SD = 10.8386) and irregular verbs (M = 19.991, SD = 14.3046), t(20) = 1.173, p = .255.

Significant differences were found between past, present and future tense for the agrammatic subjects at the p <.05 level (F(2,30) = 4.913, p = .014). Post hoc analysis using the Tukey HSD test indicated that the agrammatic individuals made significantly fewer errors in the present tense condition (M = 17,17, SD = 1.250) than in the future (M = 14.18, SD = 2.714, p = .042) and past tense condition (M = 13.82, SD = 3.737, p = .020). Figure 4 is a graphical representation of overall correct scores on all three timeframes for the agrammatic group.

30

4 Conclusions and discussion

The aim of the present study was to adapt the Test for Assessing Reference of Time (TART; Bastiaanse et al., 2008) for Italian and to investigate the production and comprehension of grammatical morphology associated with time reference in Italian patients with non-fluent agrammatic aphasia.

The agrammatic subjects performed significantly worse on both the TART production test and the TART comprehension test than the control group, who performed at ceiling. No regularity effects of Italian past tense were found for the verbs used in the Italian version of the TART in neither the production nor the comprehension test.

In line with the PADILIH (Bastiaanse et al., 2011), Italian patients with non-fluent agrammatic aphasia show a specific past time reference deficit in both production and comprehension. Past time reference was found to be more severely impaired than non-past time reference, however, this comparison only reached statistical significance in the TART production test. When past, present and future time reference were compared, instead of past and non-past time reference, the differences were significant in both the TART production test and the TART comprehension test and the specific past time reference problem became apparent. In the TART comprehension test, the agrammatic subjects performed significantly better in the present tense condition than in the future and past tense condition.

A closer examination of the different tense forms that were elicited in the TART production test, showed that the agrammatic subjects performed significantly worse in the past absolute, simple past, present perfect and future tense condition than in the baseline condition, in which an infinitive verb form was elicited. Agrammatic performance was also found to be significantly worse in the past absolute tense condition than in the simple present and future tense conditions, respectively. The agrammatic subjects made more errors in simple past tense than in present perfect tense, although these results did not reach statistical significance.

31 present and the future tense condition, at least for the TART comprehension test. In line with previous findings (Nanousi et al., 2006; Bastiaanse et al., 2011; Martínez-Ferreiro & Bastiaanse, 2013) reference to the future was found the more impaired than reference to the present. These findings indicate that different processing mechanisms are involved in present and future time reference, and therefore, a purely past and non-past categorization is inadequate.

Future tense is a form of irrealis event time reference, such that it has not an actual event time, because the event has not yet happened. In present tense the speech, event and reference time coincide. Therefore, present time reference and future time reference are substantially different from one another. Narrow syntax does not suffice for processing future time reference, as future events cannot be locally bound, because the actual (irrealis) event occurs outside of the linguistic information range.

It is plausible to presume that the underlying mechanisms are of the same nature, because in both past time reference and future time reference, the speaker needs to idealize the notion of shifting his perspective away from the present, whether it is backwards in time, like when referring to a past timeframe, or forwards in time, as in future time reference. Arguably, a form of discourse linking is needed to process future time reference, as the linguistic information needs to be linked to (irrealis) extra-sentential information. Still, past time reference is more difficult than future time reference for agrammatic individuals, even though the same mechanisms seem to be involved. The differences between discourse linking in future time reference and discourse linking in past time reference need to be further examined.

Grodzinsky (1990) argued that in a stem-based morphology language, where all inflectional rules are applied to the verb stem, omissions of inflectional suffixes would result in the production of non-words. Agrammatic aphasic speakers of a stem-based morphology language will, therefore, only produce substitution errors, as they will avoid producing non-words. In line with this notion and with previous findings from Martínez-Ferreiro and Bastiaanse (2013), the agrammatic patients in the current study exclusively produced substitution errors.

In line with Dahl (1985), Dragoy and Bastiaanse (2013) observed a prototypical use of perfective categories for past time reference in Russian agrammatism, while imperfective categories are used for non-past time reference. When referring to the past, imperfective aspect was found to be more severely impaired than perfective aspect, whereas perfective aspect was more impaired than imperfective aspect in non-past time reference. Dragoy and Bastiaanse therefore suggested, that there is an interaction between time reference and aspect.

32 imperfective simple past, although this difference did not reach statistical significance. A more profound error analysis revealed that the imperfective simple past was substituted more often with non-past tense forms, rather than other past tense forms, whereas present perfect and past absolute tense were substituted more often with other past tense forms than with non-past tense forms.

The error analysis indicates that the Italian paradigm seems to have a prototypical categorization of perfective and imperfective categories. This categorization results in the association of perfective aspect with past time reference and imperfective aspect with non-past time reference, as is the tendency across many languages (Dahl, 1985). This explains why simple past was more difficult than present perfect tense and it explains the observed pattern in substitution errors, even though not all results reached statistical significance. Nonetheless, in line with Dragoy and Bastiaanse (2013), the error analysis does indicate an interaction between time reference and aspect in Italian language.

The findings from the current study are, however, partially contrastive to the findings from Martínez-Ferreiro and Bastiaanse (2013). Imperfective aspect was found to be better preserved than perfective aspect in Spanish and Catalan agrammatic production. However, present and future tense forms were grouped together and labeled as imperfective aspect and simple past and present perfect were grouped together and labeled as perfective aspect. This does not seem to be an accurate comparison, as simple past is considered as being imperfective.

33

References

Abuom, T.O. & Bastiaanse, R. (2010). Test for Assessing Reference of Time (TART): Bilingual

Swahili-English version. Groningen: University of Groningen.

Abuom, T.O. & Bastiaanse, R. (2013). Production and comprehension of reference of time in Swahili-English bilingual agrammatic speakers. Aphasiology, 27(2), 157-177.

Anjarningsih, H.Y. (2012). Time reference in standard Indonesian agrammatic aphasia. Groningen: Grodil.

Aronoff, M. (1994). Morphology by itself: Stems and Inflectional Classes. Massachusetts: MIT-Press. Avrutin, S. (2000). Comprehension of discourse-linked and non-discourse-linked questions by

children and Broca’s aphasics. In Y. Grodzinsky, L. Shapiro, & D. Swinney (Eds.), Language

and brain: Representation and processing (pp. 295-313). San Diego: Academic Press.

Avrutin, S. (2006). Weak syntax. In K. Amunts & Y. Grodzinsky (Eds.), Broca’s region (pp. 49-62). New York: Oxford Press.

Bastiaanse, R. (2008). Production of verb in base position by Dutch agrammatic speakers: Inflection versus finiteness. Journal of Neurolinguistics, 21, 104-119.

Bastiaanse, R., Bamyaci, E., Hsu, C., Lee, J., Yarbay Duman, T. & Thompson, C.K. (2011). Time reference in agrammatic aphasia: A cross-linguistic study. Journal of Neurolinguistics, 24, 652-673.

Bastiaanse, R., Dragoy, O., Avrutin, S., Iskra, E. & Bos, L. (2013). Discourse-linking in agrammatic and fluent aphasia. Procedia – Social and Behavioral Sciences, 94, 179-180.

Bastiaanse, R., Hugen, J., Kos, M. & Zonneveld, van, R. (2002). Lexical, morphological and syntactic aspects of verb production in agrammatic aphasics. Brain and Language, 80, 142-159.

Bastiaanse, R. & Jonkers, R. (1998). Verb retrieval in action naming and spontaneous speech in agrammatic and anomic aphasia. Aphasiology, 12(11), 951-969.

Bastiaanse, R., Jonkers, R. & Thompson, C.K. (2008). Test for Assessment of Reference of Time (TART). University of Groningen, Groningen.

Bates, E. (1976). Language and Context: The Acquisition Pragmatics. New York: Academic Press. Bos, L.S. & Bastiaanse, R. (2014). Time reference decoupled from tense in agrammatic and fluent

aphasia. Aphasiology. http://dx.doi.org/10.1080/02687038.2014.886322

Bos, L.S., Dragoy, O., Stowe, L.A., Bastiaanse, R. (2013). Time reference teased apart from tense: Thinking beyond the present. Journal of Neurolinguistics, 26, 283-397.

34 Dardano, M. & Trifone, P. (1995). Grammatica Italiana con nozioni di linguistica. Bologna: Zanichelli

Editore S.p.A., derde editie.

Dardano, M. & Trifone, P. (1997). La nuova grammatica della lingua italiana. Bologna: Zanichelli Editore S.p.A.

Dragoy, O. & Bastiaanse, R. (2013). Aspects of time: Time reference and aspect production in Russian aphasic speakers. Journal of Neurolinguistics, 26, 113-128.

Dragoy, O., Stowe, L.A., Bos, L.S. & Bastiaanse, R. (2012). From time to time: Processing time reference violations in Dutch. Journal of Memory and Language, 66, 307-325.

Friedmann, N., & Grodzinsky, Y. (1997). Tense and agreement in agrammatic production: Pruning the syntactic tree. Brain and Language, 56, 397–425.

Grodzinsky, Y. (1990). Theoretical perspectives on language deficits. Cambridge, MA: MIT-Press. Jaeggli, O. & Safir, K. (1989).The Null Subject Parameter. Dordrecht: Kluwer.

Luzzatti, C., Wilmes, K. & De Bleser, R. (1994). Aachener Aphasia Test (AAT) – Versione Italiana. Firenze: Organizzazioni Speciali.

Martínez-Ferreiro, S. & Bastiaanse, R. (2013). Time reference in Spanish and Catalan non-fluent aphasia. Lingua, 137, 88-105.

Menn, , L. & Obler, L.K. (1990). Cross-language data and theories of agrammatism. In L. Menn & L.K. Obler (Eds.), Agrammatic aphasia: A cross-language narrative sourcebook (pp. 1369-1389). Amsterdam/Philadelphia: John Benjamins Publishing Company.

Mosca, N. (2011). Grammatica essenziale: Italiano. Novara: De Agostini Libri S.p.A.

Nanousi, V., Masterson, J., Druks, J., & Atkinson, M. (2006). Interpretable vs. Uninterpretable features: evidence from six Greek-speaking agrammatic patients. Journal of Neurolinguistics, 19, 209-238.

Penke, M., Janssen, U. & Krause, M. (1999). The Representation of Inflectional Morphology: Evidence from Broca’s Aphasia. Brain and Language, 68, 225-232.

Reichenbach, H. (1947). Elements of Symbolic Logic. New York: Free Press.

Rofes, A., Bastiaanse, R. & Martínez-Ferreiro, S. (2014). Conditional and future tense impairment in non-fluent aphasia. Aphasiology, 28(1), 99-115.

Rossi, E. & Bastiaanse, R. (2008). Spontaneous speech in Italian agrammatic aphasia: A focus on verb production. Aphasiology, 22(4), 347-362.

Rubin, C. & Newton, E. (2001). Capture the moment: the Pulitzer Prize photographs. New York: W.W. Norton.

Stavrakaki, S. & Kouvava, S. (2003). Functional categories in agrammatism: Evidence from Greek.

35 Velupillai, V. (2012). An Introduction to Linguistic Theory. Philadelphia: John Benjamins Publishing

Company.

Yarbay Duman, T. & Bastiaanse, R. (2009). Time reference through verb inflection in Turkish agrammatic aphasia. Brain and Language, 108, 30-39.

36

Appendices

Appendix A – Spontaneous speech sample of agrammatic subject A1

Io mi sono … eh… ero… tutti… i gior- … era… tutto… a portare… mio nipote a scuola… da mio figlio vado a prendere… e lo porto a scuola… io [xx] … alle sette e mezza… mi [xx]

Vado giù … apre … la porta … e c’era qualcosa che non … peró … [xx] … abbiamo quat [xx] … uno, con un braccio che non lo sento [xx] peró … vado giù … apre la porta … per aprire … io … abbiamo tre [xx] … al cortile … à la mia macchina … tiro la chiave de tasca … quando arrivo [xx] … sono caduto … sono stato molto fortunato … perchè c’è un signore che [xx] … più o meno la stessa partenza … lui … sua moglie l’ha vista e [xx] ha chiamato mia moglie … e arrivata subito e il … quando ero ripreso, io mi riccordo benissimo … tutto quel … fino alla chiave, dove è … e loro hanno ditto [xx] … sicurissimo … e loro hanno detto … guardi, che lei, è stato fortunate, perchè vuol dire che la testa è [xx] ancora buona … non ha [xx] niente … e invece dopo ho detto, si sono … [xx] … e adesso mi hanno fatto questa, perche era quella che ha fatto [xx] … e poi … però … con testa vado bene … però … [xx] … il tempo … prima non dicevo niente … adesso … [xx] … adesso un mese … qualcosa si … adesso … dopo, ci vuole … però è andato benissimo … perchè il tempo stato velocissimo.

Appendix B – Overview of verbs used in TART-production and TART-comprehension

Italiaans Nederlandse vertaling

cucire la maglia / lavorare a maglia de trui breien / het breiwerk naaien

disegnare un quadrato / dipingere* un quadrato een vierkant tekenen / een vierkant schilderen

temperare la matita / spezzare la matita het potlood slijpen / het potlood breken

mangiare la mela / sbucciare la mela de appel eten / de appel schillen

bere* il latte / versare il latte melk drinken / melk inschenken

spazzare il pavimento / lavare il pavimento de vloer vegen / de vloer dweilen

stirare la maglia / piegare la maglia de trui strijken / de trui vouwen

strappare la carta / incollare la carta het papier scheuren / het papier plakken

tirare il carretto / spingere* il carretto de kar trekken / de kar duwen

scrivere* una lettera / leggere* una lettera een brief schrijven / een brief lezen

svuotare la scatola / riempire la scatola de doos legen / de doos vullen

37 Appendix C –Control subjects’ correct scores on TART-production

INF SPr IMP PP PRe FU TOTAAL CORRECT

C1 20 20 20 20 20 20 120 C2 20 20 20 20 20 20 120 C3 20 20 20 20 20 20 120 C4 20 20 18 20 20 20 118 C5 20 20 20 20 20 20 120 C6 20 19 18 19 20 20 116 C7 20 20 20 20 20 20 120 C8 20 20 15 19 13 20 107 C9 20 20 19 20 20 20 119 C10 20 20 20 20 20 20 120 MEAN 20 19,9 19 19,8 19,3 20 118

INF SPr IMP PP PRe FU TOTAAL % CORRECT

C1 100 100 100 100 100 100 100 C2 100 100 100 100 100 100 100 C3 100 100 100 100 100 100 100 C4 100 100 90 100 100 100 98,3 C5 100 100 100 100 100 100 100 C6 100 95 90 95 100 100 96,7 C7 100 100 100 100 100 100 100 C8 100 100 75 95 65 100 89,2 C9 100 100 95 100 100 100 99,2 C10 100 100 100 100 100 100 100 MEAN 100 99,5 95 99 96,5 100 98,34

Appendix C: Control subjects’ correct scores on all elicited verb forms (INF for infinitive, SPr for simple present, IMP

38 Appendix D - Control subjects’ correct scores on the TART comprehension test

PAST PRESENT FUTURE TOTAL

C1 20 20 20 60 C2 20 20 20 60 C3 20 20 20 60 C4 20 20 20 60 C5 19 20 19 58 C6 20 20 20 60 C7 20 20 20 60 C8 20 20 16 56 C9 20 20 20 60 C10 19 20 19 58 MEAN 19,8 20 19,4 59,2

PAST PRESENT FUTURE TOTAL %

C1 100 100 100 100 C2 100 100 100 100 C3 100 100 100 100 C4 100 100 100 100 C5 95 100 95 96,7 C6 100 100 100 100 C7 100 100 100 100 C8 100 100 80 93,3 C9 100 100 100 100 C10 95 100 95 96,7 MEAN 99 100 97 98,67

Appendix D: Control subjects’ correct scores on all elicited timeframes (past, present and future) and overall correct

Appendix E – TART Production test: Individual scores aphasic subjects

Item no. A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11

42 97 1 1 0 0 0 0 0 0 0 1 0 98 1 1 1 1 1 1 1 1 1 1 1 99 1 0 0 0 0 0 0 0 0 1 1 100 0 0 0 0 0 1 0 1 0 0 0 101 1 1 1 1 1 1 1 1 1 1 1 102 0 0 1 0 0 0 0 0 0 1 1 103 0 1 1 1 1 1 1 1 1 1 1 104 1 1 0 1 0 1 0 1 0 1 0 105 1 0 0 0 0 0 1 0 0 1 0 106 1 0 0 0 0 0 0 0 0 1 1 107 0 1 0 0 0 1 1 1 0 1 1 108 1 0 0 1 1 1 1 1 0 0 1 109 0 0 0 0 0 0 0 0 0 1 0 110 1 1 0 1 1 0 0 0 0 1 1 111 1 1 0 0 0 1 0 1 0 1 0 112 1 1 0 0 0 1 0 1 0 0 1 113 1 1 1 1 0 1 0 1 0 1 0 114 1 1 0 0 0 1 1 0 0 0 0 115 0 1 0 1 0 0 0 0 0 1 1 116 1 1 0 0 0 0 0 0 0 0 1 117 1 0 0 0 0 0 1 0 0 1 0 118 0 0 0 0 0 0 1 1 0 0 0 119 0 0 1 0 1 1 0 1 0 1 1 120 1 1 0 1 0 1 1 1 0 1 1 TOT.CORR. 80 76 34 68 51 69 68 78 64 58 92 TOT.CORR.% 66,7 63,3 28,3 56,7 42,5 57,5 56,7 65 53,3 48,3 76,7

Appendix E: Overview of correct and incorrect responses on TART-production for all agrammatic individuals. Correct responses were scored as 1, incorrect responses were scored

as 0. The total number of correct responses is given for each agrammatic individual with the appurtenant percentage. The maximum score was 120, with 20 points per tense condition.

43 Appendix F – TART Comprehension test: Individual scores aphasic subjects

Item no. A1 A2 A3 A4 A5 A6 A7 A8 A9 A10 A11

45

Appendix F: Overview of correct and incorrect responses on the TART comprehension test for all agrammatic individuals. Correct responses were scored as 1, incorrect responses were

scored as 0. The total number of correct responses is given for each agrammatic individual with the appurtenant percentage. The maximum score was 60, with 20 points per tense condition.

Appendix G – Agrammatic subjects’ correct scores on TART-production

BA SPr IMP PP PRe FU TOTAL

A1 18 15 16 10 10 11 80 A2 19 17 13 11 6 10 76 A3 19 6 1 4 0 4 34 A4 18 14 5 13 3 15 68 A5 19 6 8 9 0 9 51 A6 18 17 7 8 4 12 66 A7 17 13 10 11 11 6 68 A8 17 18 15 17 0 11 78 A9 17 14 8 10 8 7 64 A10 20 13 2 1 10 12 58 A11 18 14 20 16 4 20 92 MEAN 18,2 13,4 9,5 10,0 5,1 10,6 66,8

BA SPr IMP PP PRe FU TOTAL %

A1 90 75 80 50 50 55 66,7 A2 95 85 65 55 30 50 63,3 A3 95 30 5 20 0 20 28,3 A4 90 70 25 65 15 75 56,7 A5 95 30 40 45 0 45 42,5 A6 90 85 35 40 20 60 57,5 A7 85 65 50 55 55 30 56,7 A8 85 90 75 85 0 55 65 A9 85 70 40 50 40 35 53,3 A10 100 65 10 5 50 60 48,3 A11 90 70 100 80 20 100 76,7 MEAN % 90,9 66,8 47,7 50,0 25,5 53,2 55,9

Appendix G: Agrammatic subjects’ correct scores on all elicited verb forms (INF for infinitive, SPr for simple present,