Ambiguous idiom comprehension in Alzheimer\'s disease

This article was downloaded by: [Universita' di Trento] On: 14 April 2009 Access details: Access Details: [subscription number 778410541] Publisher Psychology Press Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Clinical and Experimental Neuropsychology Publication details, including instructions for authors and subscription information: http://www.informaworld.com/smpp/title~content=t713657736

Ambiguous idiom comprehension in Alzheimer's disease Cecilia Rassiga a; Federica Lucchelli b; Franca Crippa a; Costanza Papagno a a University of Milan-Bicocca, Milan, Italy b Passirana di Rho, Alzheimer's Disease Centre, Milan, Italy Online Publication Date: 01 May 2009

To cite this Article Rassiga, Cecilia, Lucchelli, Federica, Crippa, Franca and Papagno, Costanza(2009)'Ambiguous idiom

comprehension in Alzheimer's disease',Journal of Clinical and Experimental Neuropsychology,31:4,402 — 411 To link to this Article: DOI: 10.1080/13803390802220019 URL: http://dx.doi.org/10.1080/13803390802220019

PLEASE SCROLL DOWN FOR ARTICLE 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 or systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses should 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 directly or indirectly in connection with or arising out of the use of this material.

JOURNAL OF CLINICAL AND EXPERIMENTAL NEUROPSYCHOLOGY 2009, 31 (4), 402–411

Ambiguous idiom comprehension in Alzheimer’s disease

NCEN

IDIOM COMPREHENSION IN AD

Cecilia Rassiga,1 Federica Lucchelli,2 Franca Crippa,1 and Costanza Papagno1 1

Downloaded By: [Universita' di Trento] At: 07:41 14 April 2009

2

University of Milan-Bicocca, Milan, Italy Alzheimer’s Disease Centre, Passirana di Rho, Milan, Italy

Ambiguous idiom comprehension was examined in 15 patients with mild probable Alzheimer’s disease (AD) by means of two tasks: a string-to-picture matching task and a string-to-word matching task. In the first, patients had to choose among four pictures, while in the second they chose among four words. For both tasks the alternatives were the picture/word corresponding to the figurative meaning, a semantic associate (picture/word) to the last word of the idiom, and two unrelated alternatives, which were, in the case of words, an unrelated foil preserving the semantic class and a literal continuation foil (a word that can follow the verb in that sentence), while in the case of pictures the first was substituted by an unrealistic foil. The patients were also submitted to three language, one visuo-perceptual, and two executive tasks. Idiom comprehension was poor, particularly when the string-topicture matching task was used, and correlated with executive tests. We confirm that linguistic and extralinguistic factors must be taken into account to explain ambiguous idiom interpretation, and we underline the role of the testing modality in the case of pathological populations. Keywords: Figurative language; Idiom; Central executive; Stroop; Alzheimer’s disease.

INTRODUCTION A peculiar aspect of linguistic communication is the use of figurative expressions, which differ from the literal meaning of their component words; these expressions are, for example, proverbs, hyperboles, metaphors, and idioms. Two main aspects concerning figurative language have been studied. While linguists have faced the problem related to the mechanisms involved in the interpretation of figurative language, neuropsychologists have investigated its neural correlates. Both agree that idioms, metaphors, proverbs, and so on do not constitute a homogeneous group. Idioms, in particular, which are one of the most common forms of figurative language (Gibbs, 1999), are typically described as frozen phrases whose meanings are stipulated directly in a mental lexicon and cannot be derived from an analysis of the words’

typical meanings. However, the majority of idioms, in particular verbal idioms, are syntactically and semantically processed and, accordingly, can undergo syntactic and semantic variations. Therefore, they vary as to their semantic transparency/ opacity—that is, the ease by which the motivation for their structure can be recovered (Nunberg, Sag, & Wasow, 1994): For instance, “to skate on thin ice” evokes a potentially dangerous situation and therefore can be defined as transparent. Idioms vary as to their decomposability—that is, the extent to which the idiomatic interpretation can be mapped onto single constituents (Gibbs, Nayak, & Cutting, 1989) and also the degree to which they can be syntactically transformed and still retain their idiomatic meaning (Gibbs & Gonzales, 1985). Finally, many idioms (known as “ambiguous”) can also be assigned a literal plausible interpretation (e.g., “to break the ice”).

This study was supported in part by a FAR (Fondi di Ateneo per la Ricerca) grant and by PRIN (Progetti Ricerca di Interesse Nazionale) grants to C.P. Address correspondence to Federica Lucchelli, Alzheimer’s Disease Centre, Via Luigi Settembrini 1, 20017 Passirana di Rho, Milan, Italy (E-mail: [email protected]).

© 2008 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business http://www.psypress.com/jcen DOI: 10.1080/13803390802220019

Downloaded By: [Universita' di Trento] At: 07:41 14 April 2009

IDIOM COMPREHENSION IN AD

Although neuropsychological studies on figurative language comprehension in Alzheimer’s disease (AD) patients concern mainly metaphors, nonetheless some information is also available for idioms. Kempler, Van Lancker, and Read (1988) examined 29 patients with probable AD ranging from mild (Mini-Mental State Examination, MMSE, score 28) to severe (MMSE score 2), who were submitted to tests of word, idiom, proverb, and novel phrase comprehension. AD patients were impaired in interpreting abstract meanings as compared to controls: When faced with alternative interpretations of familiar phrases, they chose concrete responses, suggesting, in the authors’ opinion, that they were using lexical (single word), referential meaning to interpret the phrases. In a recent study (Papagno, 2001), it was shown that the decline of figurative language in AD patients, as assessed by means of an oral explanation task of metaphors and idioms, both ambiguous and nonambiguous, decomposable and nondecomposable, frozen and unfrozen, was not an early symptom and occurred independently from literal language impairment. There were very few patients with an impairment of figurative language but normal propositional language. In a further study (Papagno, Lucchelli, Muggia, & Rizzo, 2003), testing a specific class of idioms—namely, nonambiguous opaque idioms—by means of a sentence-to-picture matching paradigm, AD patients with a mild degree of cognitive decline (MMSE 16–22) proved to be significantly impaired with respect to a comparison group, when they had to choose between two pictures, one representing the figurative interpretation and the other the literal one. However, the performance improved when the foil was substituted with a picture representing a situation described by a sentence that included only the last word of the idiom (for example, when the idiomatic expression was “to have a green thumb,” the alternative was a man hurting his thumb). When an oral explanation was required, patients still produced several literal interpretations, and their performance correlated with that on executive tasks. The literal interpretation could have produced a Stroop effect, interfering with the correct response, no matter whether the literal interpretation was implausible. Therefore, it could be hypothesized that rejecting the literal interpretation of an ambiguous idiom (for which the literal interpretation is plausible) should be even more difficult and should require a greater involvement of the central executive. The possibility of different processing and different neural correlates for ambiguous and nonambiguous idioms is supported by a functional magnetic resonance imaging (fMRI) study in which an additional

403

right temporal lobe activation was found in the case of ambiguous items in a relatedness decision task, as compared to nonambiguous items (Zempleni, Haverkort, Renken, & Stowe, 2007). However, this contrasts with a further fMRI study in which no difference between ambiguous and nonambiguous idioms was reported in cortical activation during a congruence judgment task (Romero Lauro, Tettamanti, Cappa, & Papagno, 2008). In addition, when ambiguous and nonambiguous idioms were tested in aphasic patients (Papagno & Caporali, 2007) a double dissociation was found, suggesting that these two types of idioms can be processed differently. The aim of the present study was to investigate the comprehension of a specific type of idiom— namely, ambiguous literally plausible idioms in AD. This type of expression has never been tested in these patients. As mentioned, in the case of ambiguous expressions, a literal interpretation is plausible and would represent a correct response; consequently this alternative was not included. In addition, since the testing modality has proved to be a significant factor in vascular patients (Papagno & Caporali, 2007; Tompkins, Boada, & McGarry, 1992), two testing modalities were used: a sentence-to-picture and a word-to-picture matching task. The rationale was that picture matching can underestimate the patients’ comprehension of idioms (Papagno & Caporali, 2007), and converging evidence, using different tasks, could strengthen the results. Indeed, the idiomatic meaning does not have a univocal representation as literal sentences have. Patients with impaired “flexibility” would reject pictures that do not match with their own mental representation of that idiom. However, if they recognize the sentence as a figurative expression, but they ignore its exact meaning, they should reject the semantic associate foil, too, since this is related to the literal/concrete interpretation: Therefore, a similar number of unrelated errors should be expected. It has been suggested that adequate executive functions are necessary to inhibit alternative interpretations: In fact, it was found that a suppression of the literal interpretation of familiar nonambiguous idioms was required in order to derive idiomatic meaning (Papagno et al., 2003). On the other side, lexical-semantic deficits could impair a sentence-to-word matching task simply because a word is not available and not because the patient ignores the idiomatic expression. It should be emphasized that our aim was not merely to provide evidence of a significant impairment of idiom comprehension in AD patients. Rather, in view of the neuropsychological profile of this pathological population, we wanted to

404

RASSIGA ET AL.

investigate which components (linguistic and extralinguistic) are involved in this task in general. Given these considerations, we examined: • whether ambiguous idiom comprehension is impaired; • which type of error is more frequent; • whether a correlation exists between ambiguous idiom and literal language performance; • which role executive functions play. METHOD

Materials Patients were submitted to (a) two executive tasks: the Stroop test (Caffarra, Vezzadini, Dieci, Zonato, & Venneri, 2002) and the Pencil-and-Paper Dual Task (Baddeley, Della Sala, Papagno, & Spinnler, 1997); (b) three language tasks: the Token Test (De Renzi & Faglioni, 1978), verbal fluency on semantic cue (a modified version of Novelli et al., 1986), and word auditory comprehension (Laiacona, Barbarotto, Trivelli, & Capitani, 1993), and (c) two visuo-perceptual tasks: the Street’s completion test (Spinnler & Tognoni, 1987) and the Line Orientation Test (Benton, Hamsher, Varney, & Spreen, 1983).

Downloaded By: [Universita' di Trento] At: 07:41 14 April 2009

Participants A total of 15 patients (10 women and 5 men, mean age 75.5 ± 4.8 years; educational level 6.1 ± 2.3 years) with a diagnosis of probable AD (McKhann et al., 1984) were selected on the basis of their performance on a test of screening for dementia (Milan Overall Dementia Assessment, MODA, Brazzelli, Capitani, Della Sala, Spinnler, & Zuffi, 1994). Only patients with a score over 63/100, which corresponds to a mild cognitive impairment (MMSE > 17), were included in the study. Other criteria of inclusion were: absence of previous neurological diseases, absence of lesions on the computed tomography (CT) or MRI scan, a minimum of 3 years of education, and right-handedness. Demographical and clinical data of the 15 AD patients are reported in Table 1. They all gave informed consent to participate in the study. A group of 15 neurologically unimpaired healthy participants, matched one by one to the patients for age, educational level, and gender, was also tested.

Experimental task A total of 28 verbal ambiguous idioms were selected (see Appendix). They were ambiguous in that the string had both a figurative and a literal meaning. In order to control for the semantic transparency of the idiom string, a rating study was run, in which 35 language-unimpaired healthy participants were presented with the idiom list and were asked to rate how easily the idiomatic meaning could be derived from the meaning of the words forming each string. A score ranging from 0 to 3 was assigned, where 0 meant “not at all” and 3 “perfectly” (mean 1.19, range 0.49–1.96). We also controlled for the plausibility of the literal meaning of the idiom string. The same 35 languageunimpaired participants were asked to rate the extent to which the literal meaning of the string was plausible. A score ranging from 0 (implausible) to 3 (absolutely plausible) was assigned to each

TABLE 1 Demographical data of the 15 AD patients and matched controls

Patient

Gender

Age (years)

Education (years)

Control

Gender

Age (years)

Education (years)

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

F F F F F F M F M M M F F F M

72 78 74 79 71 82 72 77 69 73 70 79 86 73 77

8 8 5 8 12 4 8 5 5 5 5 5 5 5 5

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

F F F F F F M F M M M F F F M

72 78 74 79 71 82 72 77 69 71 70 79 86 71 77

8 8 5 8 12 5 8 5 5 5 5 5 5 5 5

Note. AD = Alzheimer’s disease. M = male; F = female.

Downloaded By: [Universita' di Trento] At: 07:41 14 April 2009

IDIOM COMPREHENSION IN AD

string (mean 1.98, range 0.13–3). Finally, they were asked to rate the familiarity of each idiom string with a score of 0 (the idiom was unknown), 1 (known but not sure about the meaning), 2 (familiar with a well-known meaning), or 3 (highly familiar); their familiarity ratings ranged from 1.15 to 2.94 with a mean of 2.52. Two tests were used. The first was a sentence-toword matching task in which each idiomatic expression was paired with four words matched in terms of length (mean number of characters 6.0, range 4– 11) and written frequency (mean 10.4, range 2–26): The target word corresponded to the idiomatic interpretation of the string (henceforth idiomatic target; e.g., WINE, for “alzare il gomito,” “to raise the elbow,” i.e., to drink too much); one foil was semantically associated with the last constituent word of the idiom string (in the previous example, LEG); and two words were unrelated foils (TREE, BOX). Specifically, the first type of unrelated foil was either an abstract or a concrete word depending on the nature of the idiomatic target: The unrelated foil was abstract if the idiomatic target was abstract and concrete if the idiomatic target was concrete. A word was considered concrete based on the availability of the word referent to sensory experience. The second type of unrelated foil was a word that could plausibly complete the verb in the verb phrase (BOX). Each idiom was presented in a syntactically simple sentence formed by a subject followed by the idiom string (e.g., “He has raised the elbow”). The participants’ task was to point to the word that matched the idiomatic meaning of the sentence: The examiner read the sentence followed by the four alternatives that were written on cards and remained in front of the patient until a decision was taken. For example, the examiner read the sentence “He has raised the elbow,” and then the four alternatives (“wine, leg, box, tree”) were presented. Participants were informed that they would be read sentences commonly used in everyday conversation that had a nonliteral interpretation. An example was provided to clarify that the string had to be interpreted figuratively. We tested the strength of the relationship between (a) the semantically associate foil and literal meaning of the string; and (b) the idiomatic target and the idiomatic meaning of the idiom string. To this aim, 25 healthy participants were asked to rate the specific relatedness of (a) and (b) on two 5-point rating scales ranging from 0 (not at all related) to 5 (perfectly related). Idiomatic targets were rated as significantly more related to the idiom strings than semantically associated targets were to the literal meaning of the strings, 4.57 (range 3.68–4.92, SD 0.28) versus 1.61 (range 1–3.28, SD 0.67),

405

t(44) = 19.49, p < .00001, a result that might help the choice of an idiomatic target. For the picture-matching task four line-drawings were prepared: one corresponding to the figurative interpretation (for example in the case of to lift the elbow the drawing of a drunk person was presented); a semantically associated foil, which was described by the verb of the idiom followed by a semantic associate (a dancer lifting his leg); a picture in which the verb was followed by a word that could normally follow that verb (a person lifting a box); and a picture in which the verb was followed by a word that made the sentence implausible (a man lifting a tree on his finger). The above example of the task is presented in Figure 1. The tasks were performed in different sessions: In the first session, AD patients received the general neuropsychological tests, while in the second session they performed the idiom tasks. The order of presentation of the two idiomatic tasks was counterbalanced across participants.

RESULTS The results of the 15 patients for the verbal, executive, and visual tests are shown in Table 2, while the results for the two idiom comprehension tasks are reported in Table 3 and Table 4. Word comprehension and verbal fluency were minimally impaired. As to the visuo-perceptual tasks, performance on the Street completion test was normal in all patients, while the Line Orientation Judgment was mildly impaired in 3 patients. In the sentence-to-picture matching task the mean number of correct responses was 9.07 (32.35%, SD 4.80), while for the 15 healthy participants the mean was equal to 15.93 (56.90%, SD 6.14). Errors were predominantly semantic foils for both the patients (64.43% of all errors) and the comparison group (68.50%). Unrelated errors of Type 1 (a picture described by a sentence in which the last word of the idiom was substituted with a word that could plausibly follow the verb) did not differ from Type 2 errors (totally implausible) in patients, Type 1 = 16.90%, Type 2 = 18.66%, F(1, 14) = 0.42, p = .53, nor in comparison subjects, Type 1 = 16.57%, Type 2 = 14.92%; F(1, 14) = 0.15, p = .71. In the sentence-to-word matching task the mean number of correct responses for the AD patients was 17.73 (63.33%, SD 4.39), while in the case of the comparison group it was 21.93 (78.35%, SD 4.23). Also in this case, errors were mainly semantic foils for patients (67.53% of all errors) and comparison subjects (59.34%). Unrelated errors were 22.73% of Type 1(abstract or concrete foils depending on the

Downloaded By: [Universita' di Trento] At: 07:41 14 April 2009

406

RASSIGA ET AL.

Figure 1. Example of the sentence-to-picture matching task.

TABLE 2 Alzheimer’s disease patients’ neuropsychological assessment Stroop

Patient

MODA

Time (sec)

Errors

Cut-off (range) 89.1 (0–100) < 36.92 < 4.24 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

85.3 73.2 72.4 76.8 79.7 87.5 89.2 90.2 88.2 78.6 82.3 79.1 67.2 83.9 77.0

55.50 107.25 100.75 52.75 97.75 0.00 36.50 44.25 44.00 85.75 101.75 9.75 0.00 44.25 41.75

11.50 1.25 8.25 0.25 3.25 0.00 0.50 0.00 1.50 13.25 7.25 19.00 12.75 1.25 0.00

Dual task (%) 70.0 95.0 65.0 82.3 81.8 67.0 100.0 86.3 77.2 74.1 70.4 58.3 83.3 87.7 97.1 92.0

Token test

Word Semantic Street’s comprehension fluency completion test

26.50 (0–36) 72 (0–80) 29.75 26.75 29.75 27.25 29.25 29.00 33.50 28.00 30.50 28.75 23.25 32.75 28.50 33.25 33.00

78 69 73 74 80 76 76 80 77 77 77 79 66 80 70

9.5 5.6 6.6 4.9 7.6 10.7 7.1 7.2 10.3 9.1 9.0 7.1 6.7 4.6 6.7 6.1

Benton line orientation

2.25 (0–14)

17 (0–30)

7.50 5.00 5.75 4.50 4.50 6.75 7.25 8.25 7.50 4.75 4.75 5.75 4.75 7.75 2.25

24 16 22 25 23 27 27 30 17 19 18 14 16 22 27

Note. MODA = Milan Overall Dementia Assessment. Word comp. = Word comprehension.

target being abstract or concrete) and 9.74% of Type 2 (a word that could follow the verb, but had no relation both with the literal and the idiomatic meaning) in patients. The difference was significant, F(1, 14) = 9.12, p < .01, Cohen’s f = 0.81, while they did not differ in the comparison group, Type 1 = 20.88%, Type 2 = 19.78%; F(1, 14) = 0.028, p = .87. Age proved to be the only significant confounding factor in modeling scores, F(2, 26) = 7.49, p = .003,

unlike gender, F(2, 23) = 1.39, p = .27, and years of education, F(2, 23) = 0.132, p = .88. A multivariate analysis of covariance (MANCOVA) model was therefore tested, with the responses on the two matching tasks as dependent variables, group as factor, and age as covariate. Results showed that the difference in number of correct responses of both the sentence-to-word and the sentence-to-picture matching tasks between AD and comparison

IDIOM COMPREHENSION IN AD

407

TABLE 3 Sentence-to-word matching task

Downloaded By: [Universita' di Trento] At: 07:41 14 April 2009

Patient

Ta

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

18 14 13 22 19 23 13 23 12 16 12 16 21 25 19

Mean SD

17.73 4.39

SF 6 8 11 5 8 2 8 3 9 7 13 8 6 3 7 6.93 2.96

D1

D2

Control

1 4 3 0 1 2 6 2 5 3 2 3 1 0 2

3 2 1 1 0 1 1 0 2 2 1 1 0 0 0

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

2.33 1.71

1.00 0.92

Ta 15 22 26 22 28 15 19 18 27 21 22 25 25 26 18 21.93 4.23

SF 10 4 1 2 0 4 6 8 0 2 3 1 2 2 9 3.60 3.22

D1

D2

1 1 1 3 0 6 1 1 0 2 0 2 1 0 0

2 1 0 1 0 3 2 1 1 3 3 0 0 0 1

1.26 1.57

1.20 1.14

Note. Performance of AD (Alzheimer’s disease) patients and matched controls. Ta: target. SF: semantic foil. D1 (errors Type 1): foil preserving the semantic class. D2 (errors Type 2): literal continuation foil. TABLE 4 Sentence-to-picture matching task Patient 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. Mean SD

Ta 11 2 6 9 18 7 5 20 8 10 12 6 6 10 6 9.07 4.80

SF

D1

D2

Control

14 17 13 9 5 14 16 4 13 11 12 16 14 10 15

0 4 3 5 3 2 4 3 3 2 2 4 5 5 3

3 5 6 5 2 5 3 1 4 5 2 2 3 3 4

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

12.20 3.84

3.20 1.37

3.53 1.46

Ta 11 15 14 20 23 4 8 16 25 17 21 22 13 21 9 15.93 6.14

SF 12 8 12 6 3 15 13 10 2 5 5 3 10 6 14 8.26 4.32

D1

D2

2 3 0 2 2 3 5 1 0 4 1 2 4 0 1

3 2 2 0 0 6 2 1 1 2 1 1 1 1 4

2.00 1.55

1.80 1.57

Note. Performance of AD (Alzheimer’s disease) patients and matched controls. Ta: target. SF: semantic foil. D1 (errors Type 1): foil preserving the semantic class. D2 (errors Type 2): literal continuation foil.

subjects was highly significant, F(2, 26) = 5.87, p = .008, with a substantial effect size for each task (Cohen’s f = 0.66 and f = 0.84, respectively). With respect to the sentence-to-word matching task, the 95% confidence interval for the difference between groups was equal to (0.98; 7.49), while for the sentence-to-picture task it was equal to (2.8; 10.6). The performance on the sentence-to-word matching task correlated with that on the dual task (r = .569, p = .027), but neither with the severity of cognitive decline as assessed by the MODA (r = .060,

p = .833) nor with the Stroop (r = –.24, p = .38). However, the performance on the Stroop correlated with the number of semantically related errors (i.e., the longer time to perform the Stroop was related to more semantic errors with idioms; r = .59, p = .021). In light of such a correlation structure, a causal model was adopted in order to measure how much each of the different independent abilities accounted for the AD performance. Dual task proved to be the only predictor for the sentenceto-word matching task (b = 20.14, p = .027).

Downloaded By: [Universita' di Trento] At: 07:41 14 April 2009

408

RASSIGA ET AL.

The performance on the sentence-to-picture matching task correlated with verbal fluency (r = .705, p = .003), as well as with word comprehension (r = .671, p = .006). However, the correlation with verbal fluency was not significant (r = .50, p = .07) once the effect of word comprehension was partialed out. Conversely, no significant correlation was found with word comprehension, after removing the effect of verbal fluency (r = .42, p = .13). All these results held even when variables were transformed in order to stabilize their variance. It is, for instance, the case of count variables such as errors or correct answers, transformed via a square root function. Item analysis was performed by a repeated measures analysis of variance (ANOVA). Results showed a significant difference between the word-to-sentence and the picture-to-sentence matching tasks, both for the target and for the semantic associate foil, in patients and comparison subjects. With regard to the correct responses in the word-to-sentence matching task, the mean was significantly lower for patients (9.50) than for comparison subjects (11.75), F(1, 27) = 35.03, p < .001. Also in the case of the word-to-picture matching task, the mean number of correct responses was lower for patients (4.86) than for comparison subjects (8.54), with a significant difference, F(1, 27) = 55.56, p < .001. As far as the semantic associate foil was concerned, the mean of semantic errors was significantly higher for patients, both in the word-to-sentence (3.71 vs. 1.96), F(1, 27) = 18.194, p < .001, and in the sentence-to-picture matching task (6.54 vs. 4.43), F(1, 27) = 15.64, p < .001. Moreover, performance in the sentence-to-word matching task was significantly better than that in the sentence-to-picture matching task for both groups, t(14) = 6.363, p = .000 and t(14) = 5.81, p = .000, for patients and comparison group, respectively. However, the correlation of the sentenceto-word with the sentence-to-picture matching task is dramatically different between the two groups: While in the case of the comparison group it proves to be significant (r = .76, p = .001), this was not the case for AD patients (r = .35, p = .20). Familiarity did not correlate with the number of correct responses in AD and in the comparison group, while the number of correct responses correlated with the goodness of the pictorial representation of the idioms (r = .53, p = .003).

DISCUSSION The aim of the present study was to assess the comprehension of ambiguous idioms in AD patients, by means of two different tasks. Adding to previous findings (Papagno et al., 2003), it appears that AD

patients are impaired even with the easier category of ambiguous idioms and even when the task is made easier than usual. Indeed, ambiguous idioms proved to be a much easier category with respect to nonambiguous idioms in aphasic patients (see Papagno & Caporali, 2007). However, a significantly better performance was found in the sentenceto-word matching task. The type of task, therefore, had a significant effect on the participants’ performance. The result is consistent with that obtained with aphasic patients (Papagno & Caporali, 2007). It appears that the overt representation of the literal meaning prompted the semantic associate response, suggesting that participants were unable to suppress the literal interpretation when its explicit pictorial representation was available. A possible interpretation could be that these tasks require the engagement of extensive extralinguistic resources. Even if the semantic performance of patients was still in the normal range (as checked by means of word comprehension and verbal fluency), and the task was easier (two versions of the same task, one of which is much easier), nonetheless AD patients were impaired. However, if the results meant only that an extensive engagement of extralinguistic resources is required, a correlation with the degree of cognitive decline should be found, which was not the case. Therefore, it seems clear that semantic deficits alone and executive deficits alone cannot explain the impairment of patients, but the two factors should be combined. This means that, in general, people need both semantic resources and extralinguistic abilities in order to understand idioms, even with the easier category of ambiguous idioms and with an easier task. Neuroimaging studies, showing both a temporal and frontal activation in idiom processing, support this possibility (Romero Lauro et al., 2008; Zempleni et al., 2007). Visuo-perceptual deficits could not be considered responsible for the poor performance in the picture-matching task, since the performance on these tests was overall in the normal range and did not correlate with the idiom task. In the picture-matching task, a high variability emerged among AD patients but also among the elderly healthy participants. Some performed very poorly, while others’ performance was almost errorless. Unfortunately, the comparison group was not submitted to executive tests, but it can be argued that, since age proved to significantly affect performance, and executive functions deteriorate with age (see Phillips, MacPherson, & Della Sala, 2002, for a review), this variability would reflect different executive performance even in normal

Downloaded By: [Universita' di Trento] At: 07:41 14 April 2009

IDIOM COMPREHENSION IN AD

subjects. We can hypothesize that elderly people with good executive functions perform well, while elderly with less efficient executive functions perform at a lower level. Although the hypothesis requires further support, it is well known that age has a significant effect on neurologically unimpaired subjects in many psychometric tasks. This also holds true for idiom comprehension: In a sample of 201 healthy 20- to 80-year-old individuals it was found that performance significantly decreased with aging (Papagno et al., 2007). In the present study, a relation between executive tasks and idiom comprehension was found in the case of the sentence-to-word matching task. The correlation between the sentence-to-word matching task and the dual task in the case of ambiguous idioms has been found also in aphasic patients (Papagno & Caporali, 2007). As far as errors are concerned, the most frequently produced type was the semantically associated foil. This suggests that the literal interpretation remains active while the sentence is being processed and cannot be inhibited, even when the interference of the literal interpretation is minimized—that is, when the target word is semantically associated with the last constituent word and not with the global literal meaning of the string. Although semantically based errors were evident in both idiom tasks, it should be noted that all the patients, with the exception of only one participant, were not significantly impaired at a word comprehension level as well at a sentential level. The wrong choice (literal/semantic associate) might be the outcome of an impairment at several levels. First, the possibility that an associative mechanism might have contributed to this result exists, especially because of the correlation between semantic errors and Stroop performance, suggesting that an automatic association is not inhibited by the executive system, producing the semantically related answer. The AD patients’ degraded semantics can contribute to this effect (Chertkow & Bub, 1990). The choice of AD patients seems to be influenced by a lexical effect, and apparently no processing of the sentential meaning took place. An alternative interpretation could be a tendency to give concrete responses to individual words resulting from the premature choice of an interpretation, as found in proverbs by Burgess and Chiarello (1996) and by Iakimova, Passerieux, and Hardy-Baylé (2006) in schizophrenic patients, whose common response mode in a task of ambiguous idiom comprehension was the selection of responses relating to the concrete meaning of a single word in the idiom. However, even if we cannot rule out the possibility that an associative mechanism may have contributed

409

to our results, especially because of the choice of foils that were semantically associated with the last word of the idiomatic string, previous evidence (Papagno et al., 2003) suggests that the impairment goes far beyond a single-word level. In addition, unrelated errors of Type 1 (a word that could be abstract or concrete depending on the target semantic class) were significantly more frequent than Type 2 errors in patients, suggesting that, at least in some cases, the patients processed the sentence and rejected the literal meaning, but the correct idiomatic one was not available. Finally, sentential processing was not impaired in AD patients, and there is no reason to believe that AD patients (but also the comparison subjects, who also produced semantic errors) did not process the sentences. A last comment concerns the lack of correlation between the two tasks in the AD patients, in contrast with the significant correlation found in the comparison subjects, and also in aphasic patients using nonambiguous idioms. A possible interpretation could be that while aphasic patients with a focal vascular lesion are expected to have deficits confined to language, and the language impairment plays the same role in each test, AD patients are impaired in different cognitive domains, and, since tests are multifactorial, this can result in an asymmetric, though pathological, performance in the two idiom tasks. All in all, this study shows that idiom comprehension is a complex task, involving extralinguistic (executive) functions. This is in line with a transcranial magnetic stimulation (TMS) chronometric study (Fogliata et al., 2007), in which a selective impairment in accuracy for idioms was found when repetitive transcranial magnetic stimulation (rTMS) was applied to the prefrontal and temporal cortex 80 ms after picture presentation, but with rTMS continuing to affect performance after a longer time of 120 ms. Original manuscript received 20 January 2008 Revised manuscript accepted 21 May 2008 First published online 11 July 2008

REFERENCES Baddeley, A. D., Della Sala, S., Papagno, C., & Spinnler, H. (1997). Dual-task performance in dysexecutive and nondysexecutive patients with a frontal lesion. Neuropsychology, 11, 187–194. Benton, A. L., Hamsher, K., Varney, N. R., & Spreen, O. (1983). Contribution to neuropsychological assessment. New York: Oxford University Press. Brazzelli, M., Capitani, E., Della Sala, S., Spinnler, H., & Zuffi, M. (1994). A neuropsychological instrument adding to the description of patients with suspected

Downloaded By: [Universita' di Trento] At: 07:41 14 April 2009

410

RASSIGA ET AL.

cortical dementia: The Milan Overall Dementia Assessment. Journal of Neurology, Neurosurgery and Psychiatry, 57, 1510–1517. Burgess, C., & Chiarello, C. (1996). Neurocognitive mechanisms underlying metaphor comprehension and other figurative language. Metaphor and Symbolic Activity, 11, 67–84. Caffarra, P., Vezzadini, G., Dieci, F., Zonato, F., & Venneri, A. (2002). Una versione abbreviata del test di Stroop: Dati normativi nella popolazione italiana [A shortened version of the Stroop test: Normative data for the Italian population]. Nuova Rivista di Neurologia, 12, 111–115. Cherktow, H., & Bub, D. (1990). Semantic memory loss in Alzheimer-type dementia. In M. Schwartz (Ed.), Modular deficits in Alzheimer-type dementia (pp. 207–244). Cambridge, MA: MIT Press. De Renzi, E., & Faglioni, P. (1978). Normative data and screening power of a shortened version of the token test. Cortex, 14, 41–49. Fogliata, A., Rizzo, S., Reati, F., Miniussi, C., Oliveri, M., & Papagno, C. (2007). The time course of idiom processing. Neuropsychologia, 45, 3215–3222. Gibbs, R. (1999). Figurative language. In R. Wilson & F. Keil (Eds.), The MIT encyclopedia of the cognitive sciences (pp. 314–315). Cambridge, MA: MIT Press. Gibbs, R. W., & Gonzales, G. P. (1985). Syntactic frozenness in processing and remembering idioms. Cognition, 20, 243–259. Gibbs, R. W., Nayak, N. P., & Cutting, C. (1989). How to kick the bucket and not decompose: Analyzability and idiom processing. Journal of Memory and Language, 28, 576–593. Iakimova, G., Passerieux, C., & Hardy-Baylé, M. C. (2006). Interpretation of ambiguous idiomatic statements in schizophrenics and depressive patients. Evidence for common and differential cognitive patterns. Psychopathology, 39, 277–285. Kempler, D., Van Lancker, D., & Read, S. (1988). Proverb and idiom comprehension in Alzheimer disease. Alzheimer Disease and Associated Disorders, 2, 38–49. Laiacona, M., Barbarotto, R., Trivelli, C., & Capitani, E. (1993). Dissociazioni semantiche intercategoriali: Descrizione di una batteria standardizzata e dati normativi [Intercategory dissociations: Description and normative data of a standardized battery]. Archivio di Psicologia, Neurologia e Psichiatria, 2, 209–248. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., & Stadlan, E. M. (1984). Clinical diagnosis of Alzheimer’s disease: Report of the

NINCDS-ADRDA Work Group under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology, 34, 939–944. Novelli, G., Papagno, C., Capitani, E., Laiacona, M., Vallar, G., & Cappa, S. F. (1986). Tre test clinici di ricerca e produzione lessicale: Taratura su soggetti normali [Three clinical tests of lexical search and production: Standardization in normal subjects]. Archivio di Neurologia, Psicologia e Psichiatria, 47, 477–506. Nunberg, G., Sag, I. A., & Wasow, T. (1994). Idioms. Language, 70, 491–538. Papagno, C. (2001). Comprehension of metaphors and idioms in patients with Alzheimer’s disease: A longitudinal study. Brain, 124, 1450–1460. Papagno, C., & Caporali, A. (2007). Testing idiom comprehension in aphasic patients: The effects of task and idiom type. Brain and Language, 100, 208–220. Papagno, C., Caporali, A., Rizzo, S., Reati, F., Morreale, L., & Venneri, A. (2007). La comprensione di espressioni idiomatiche non ambigue: Taratura di un test di associazione frase-figura [Comprehension of non ambiguous idioms: Standardization of a sentence-to-picture matching test]. Giornale Italiano di Psicologia, 34, 1019–1039. Papagno, C., Lucchelli, F., Muggia, S., & Rizzo, S. (2003). Idiom comprehension in Alzheimer’s disease: The role of the central executive. Brain, 126, 2419–2430. Phillips, L. H., MacPherson, S. E. S., & Della Sala, S. (2002). Age, cognition and emotion: The role of anatomical segregation in the frontal lobes. In J. Grafman (Ed.), Handbook of neuropsychology (2nd ed., Vol. 7, pp. 73–97). Amsterdam: Elsevier. Romero Lauro, L. J., Tettamanti, M., Cappa, S. F., & Papagno, C. (2008). Idiom comprehension: A prefrontal task? Cerebral Cortex, 18, 162–170. Spinnler, H., & Tognoni, G. (1987). Standardizzazione e taratura italiana di test neuropsicologici [Italian standardization and normative data of neuropsychological tests]. The Italian Journal of Neurological Sciences, 8(Suppl. 6). Tompkins, C. A., Boada, R., & McGarry, K. (1992). The access and processing of familiar idioms by brain-damaged and normally aging adults. Journal of Speech and Hearing Research, 35, 626–637. Zempleni, M. Z., Haverkort, M., Renken, R., & Stowe, L. A. (2007). Evidence for bilateral involvement in idiom comprehension: An fMRI study. NeuroImage, 34, 1280–1291.

IDIOM COMPREHENSION IN AD

APPENDIX

Downloaded By: [Universita' di Trento] At: 07:41 14 April 2009

List of idiomatic expressions Plausibility of the literal meaning (range = 0–3) Expression

Plausibility

Alzare il gomito (to lift the elbow, to drink too much) Vuotare il sacco (to empty the sack, to confess something) Sputare il rospo (to spit the toad, to reveal a secret) Vedere le stelle (to see the stars, to experience a pain) Essere al verde (to be at the green, to be broke) Prendere la porta (to take the door, to go away) Levare le tende (to remove the tents, to leave) Perdere la testa (to lose the head, to get mad) Avere il pollice verde (to have a green thumb) Prendere un granchio (to take a crab, to make an error) Tirare la cinghia (to pull the belt, to be poor) Perdere il filo (to lose the thread, to get lost) Cambiare pagina (to change page, to change your own life) Mettersi le mani nei capelli (to put the hands in the hair, to be in despair) Tagliare la corda (to cut the rope, to escape) Scendere a rotta di collo (to go down breaking the neck, to behave riskily) Avere le mani bucate (to have holes in the hands, to spend too much) Saltare la mosca al naso (to jump the fly on the nose, to get angry) Mettere la pulce nell’orecchio (to put the flea in the ear, to suggest something) Avere carta bianca (to have white paper, to be free of taking decisions) Alzare i tacchi (to lift the heels, to go away abruptly) Attaccare un bottone (to attach a button, to speak too much) Stringere i denti (to tighten the teeth, to strongly pursue a goal) Gettare la spugna (to throw the sponge, to throw in the towel) Rompere il ghiaccio (to break the ice) Mangiarsi il fegato (to eat the liver, to eat one’s heart out) Chiudere un occhio (to shut an eye, with one’s eye shut) Perdere la bussola (to lose the compass, to lose one’s bearings)

2.75 2.78 0.75 3.0 0.13 1.75 2.44 0.38 0.70 2.70 2.90 1.90 2.75 2.85 2.85 0.35 0.35 0.60 0.75 2.10 2.45 2.95 2.55 2.60 2.80 2.0 3.0 2.50

411