Gender Differences in Emotional Prosody Processing — an Fmri Study —

Psychologia, 2004, 47, 113–124

GENDER DIFFERENCES IN EMOTIONAL PROSODY PROCESSING — AN FMRI STUDY —

Satoshi IMAIZUMI1), Midori HOMMA1), Yoshiaki OZAWA1),

1)Hiroshima Prefectural College of Health Sciences, Japan,

Masaharu MARUISHI2) and Hiroyuki MURANAKA2)

2)Hiroshima Prefectural Rehabilitation Center, Japan

To clarify how the brain understands the speaker’s mind for verbal acts, we analyzed fMRI images obtained from 24 subjects when they judged linguistic meanings or emotional manners of spoken phrases. The target phrases had linguistically positive or negative meanings and were uttered warmheartedly or coldheartedly by a woman speaker. Significant interaction effects of meaning and manner were observed on the acoustic characteristics of utterances, such as F0 range, and also on the perceptual behavior evaluated by response time and judgment correctness. When compared to the female subjects, the male subjects showed significantly stronger activation only in the right frontomedian cortex, which can be hypothesized to analyze and understand speaker’s hidden but true intentions from speaking acts. These results suggest that emotion modulates linguistic processes not only in speech production but also in speech perception, and such modulations differ between the genders at least in perceptual processes.

To interpret the communicative intent of others, how emotionally something is said may be as important as what is linguistically said. Emotional prosody is an important non- linguistic component of verbal acts conveying speaker’s emotional intent, and is acoustically characterized by pitch height and range, the temporal structure and other supra-segmental features of speech (Ladd, 1996). Phrases with positive linguistic meanings may convey negative intent when uttered with coldhearted emotion. For instance, “It’s wonderful” uttered with evident sarcasm may convey the information “It’s not wonderful at all.” Phrases with negative linguistic meanings may convey positive intent when uttered with warmhearted emotion. A verbal act “What a stupid guy you are” may be interpreted as “What a nice guy you are” when uttered with a friendly feeling evoking strong sympathy. Listeners have to correctly extract the speaker’s emotional intent separately from the linguistic meanings of the phrases, and integrate them to understand hidden but true intentions of the speaker. Interaction between linguistic meanings and emotional prosody in the brain has recently started to be analyzed (Mitchell, Elliot, Barry, Cruttenden, & Woodruff, 2003; Schirmer & Kotz, 2003). But it is not yet clear how linguistic and

Correspondence concerning this article should be addressed to Satoshi Imaizumi, Department of Communication Sciences and Disorders, Hiroshima Prefectural College of Health Sciences, 1-1 Gakuen, Mihara, Hiroshima 723-0053, Japan (e-mail: [email protected]).

113 114 IMAIZUMI, HOMMA, OZAWA, MARUISHI, & MURANAKA emotional information is processed in the brain to read the speaker’s mind, and if there are any sex differences in the neural resources responsible for mind reading with regard to verbal acts. The primary purpose of this study is to help clarify how the mind reading system for speaking acts is organized in the brain. The secondary purpose of this study is to investigate how the functional organization of the male brain for mind reading differs from the female brain. As the following brief review of the literature shows, neither behavioral nor brain imaging studies have yet provided a clear picture of the functional organization of the brain for language and prosody processing, and their gender differences. Some recent data indicate that both hemispheres engage in the perception of emotional prosody (Kotz, Meyer, Alter, Besson, von Cramon, & Friederici, 2003), while some other studies suggest that both emotional and linguistic prosody (Bryan, 1989; Dykstra, Gandour & Stark, 1995), or only emotional prosody (Blonder, Bowers, & Heilman,1991; Borod, 1993; Starkstein, Federoff, Price, Leiguarda, & Robinson, 1994), or only linguistic prosody (Bradvik et al., 1991; Weintraub, Mesulam, & Kramer, 1981) are processed in the right hemisphere. Some results imply that sentence-level linguistic prosody is processed in the left hemisphere (Emmorey, 1987; Van Lancker, 1980). Alternatively, it has been proposed that the lateralization of prosodic processing may vary as a function of the acoustic parameters of prosody, such as fundamental frequency, intensity or duration under study (Zatorre, Belin, & Penhune, 2002). Some others suggest that the lateralization of prosodic and phonetic processing may vary depending on the tasks imposed and subjective attention (Imaizumi, Mori, Kiritani, Hosoi, & Tonoike, 1998; Mitchell et al., 2003), and the left hemisphere activity may vary depending on the magnitude of task demands (Kotz et al., 2003). A much debated question is whether sex differences exist in the functional organization of the brain for language and prosody (Baxter et al., 2003; Frost et al., 1999; Koelsch, Maess, Grossmann, & Friederici, 2003; Kotz et al., 2003; Schirmer, Kotz, & Friederici, 2002; Schirmer & Kotz, 2003; Shaywitz et al., 1995; Vikingstad, George, Johnson, & Cao, 2000). A long-held hypothesis posits that language functions are more likely to be highly lateralized in males and to be represented in both cerebral hemispheres in females. This hypothesis was developed based on behavioral results that aphasia was more common in males with left hemisphere damage (48% in males vs. 13% in females), while language deficits were observed after both right and left hemisphere damage in females (Hampson & Kimura, 1992; McGlone, 1977). However, attempts to prove this hypothesis have been inconclusive. Some data indicate that language primarily lateralizes to left in males, and approximately half have left lateralization and the other half have bilateral representation in females (Vikingstad et al., 2000), while other results suggest that men and women show very similar, strongly left lateralized activation patterns (Frost et al., 1999). Using spoken phrases with positive or negative meanings uttered warmheartedly or coldheartedly by a female Japanese speaker, we analyzed brain activities when subjects judged linguistic meaning and emotional manner of the phrases. Based on the fMRI GENDER DIFFERENCES IN EMOTIONAL PROSODY PROCESSING 115 analyses, we discussed possible neural mechanisms for mind reading for verbal acts and their gender differences.

METHOD

All experiments in this study were conducted in accordance with Declaration of Human Rights, Helsinki 1975 and the research ethics regulations of the authors’ affiliate. Written informed consent was obtained from each subject after explaining the purpose and the outline of the method for this research and the advantages and disadvantages expected for the subjects.

Speech Material: Eighty frequently-used phrases with positive linguistic meanings, such as “I love you,” and 80 phrases with negative linguistic meanings, such as “I hate you,” were uttered warmheartedly or coldheartedly by 2 female and 3 male Japanese speakers. The warmhearted utterances were made with strong pleasure, while the coldhearted ones were made with strong hatred. These utterances were digitized and analyzed to extract prosodic characteristics such as F0 contour, and then used for listening tests in preliminary experiments. In a preliminary listening experiment, 24 right-handed healthy Japanese adults judged whether the linguistic meaning of a randomly presented speech sample is positive or negative by pressing one of two buttons as fast and correctly as possible. In another preliminary experiment, the same listening subjects judged whether the emotional manner of speaking is warmhearted or coldhearted for a randomly presented speech sample by pressing one of two buttons as fast and correctly as possible. The male subjects showed significantly lower correctness scores with longer response time (RT) than the female subjects in these preliminary listening experiments. Based on the judgment correctness scores in these preliminary experiments, 40 linguistically positive and 40 linguistically negative phases uttered in two emotional modes were selected so as to minimize the sex-dependent differences in the correctness scores from speech samples uttered by a female speaker who gave the highest correctness scores. The selected samples were used in fMRI measurement. This selection was done to minimize the effects of task difficulties due to speaker’s verbal skill on the brain imaging results, and eventually speaker-dependent effects could not be addressed in this study.

Acoustic Analyses: Fundamental frequency at various characteristic timing points and the total length of utterances were measured and analyzed by ANOVA with two factors of Manner (Warmhearted vs. Coldhearted) as a within- item measure and Meaning (Positive vs. Negative) as a between- item measure.

Listening Tasks: The listening subjects were 24 right-handed healthy Japanese adults (12 males and 12 females aged 24.7 on average). Two listening tasks were imposed, the language task and the emotion task. For the language task, the listening subjects were instructed to judge whether the linguistic meaning of a presented speech sample is positive or negative by pressing one of two buttons as fast and correctly as possible. For the emotion task, they were instructed to judge whether the emotional prosody of a presented speech sample is warmhearted or coldhearted by pressing one of two buttons as fast and correctly as possible. The listening tasks were carried out twice; once before fMRI measurement in a quiet sound proof room using all the verbal stimuli, and once with fMRI measurement using selected stimuli. The selection of verbal stimuli was done so as to minimize the sex-dependent differences in the correctness score. The response time and the percent correct were analyzed by ANOVA with three factors of Manner (Warmhearted vs. Coldhearted), Meaning (Positive vs. Negative), and Sex (Female vs. Male). For the control task used in brain imaging, tones with a rising pitch or a falling pitch were prepared. The length and the presentation level of the tones were adjusted to be the same as those of the speech samples. The control task was to judge whether the pitch of a tone raises or falls by pressing one of two buttons as fast and correctly as possible. 116 IMAIZUMI, HOMMA, OZAWA, MARUISHI, & MURANAKA

Fig. 1. F0 range with standard error bars. Empty bar: Negative linguistic meanings; Meshed bar: Positive linguistic meanings.

Functional Brain Imaging: All the subjects who took part in the listening task were scanned in the fMRI experiment. With the subject in the fMRI system (Siemens, Magnetic Symphony 1.5T), one of the listening tasks and the control task were performed alternately at 30-second intervals, 4 times each, and this block design, which took 4 minutes to complete, was performed under the 2 listening tasks. The speech and tone stimuli were presented at the most comfortable level for individual subjects once per 3 seconds. The subjects were required to express their judgment by pressing one of two buttons with their right index or middle fingers as quickly and correctly as possible. The scanning conditions were: The interscan interval (TR), 6 seconds; acquisition time (TA), 4.4 seconds; flip angle, 90°. The image data obtained were analyzed using a statistical parametric mapping (SPM2, the Wellcome Department of Imaging Neuroscience, http://www.fil.ion.ucl.ac.uk/spm/spm2.html). Cognitive subtraction analyses were carried out between each listening task and the control, between the two tasks directly, and between the two sexes. Voxels were identified as significant if they passed a threshold of PPWE-corr<0.05, corrected for multiple comparisons over the whole-brain volume.

RESULTS

Acoustic Characteristics Acoustic characteristics varied depending on Manner and Meaning. As illustrated in Fig. 1, ANOVA on F0 range revealed that the interaction effect of Manner and Meaning (F(1,160)=4.97, p=.027), the main effects of Manner (F(1,160)=63.21, p<.0001) and Meaning (F(1,160)=9.98, p<.005) were significant. The post-hoc test revealed that the warmhearted utterances had 104 Hz wider F0 range than the coldhearted utterances (p<.0001), and the linguistically negative phrases had 41 Hz wider F0 range than the linguistically positive phrases (p=.02). The warmhearted utterances of negative phrases had 70 Hz wider F0 range than the warmhearted utterances of positive phrases (p=.004), while coldhearted utterances had no significant differences between linguistically positive and negative phrases (p=.27). GENDER DIFFERENCES IN EMOTIONAL PROSODY PROCESSING 117

Fig. 2. Results of the emotion task. (a) Percent correct with standard error bars. (b) Response time with standard error bars. Filled bar: Phrases with negative linguistic meanings; Empty bar: Phrases with positive linguistic meanings. Cold: Coldhearted; Warm: Warmhearted.

Behavioral Results As shown in Fig. 2 (a), significant differences in the correctness scores in the emotion task were detected depending on the factors analyzed. Significant interaction effect was found only between Meaning and Manner (F(1,72)=7.8, p=.007). This significant interaction corresponded to the tendency that the coldhearted utterances of the negative phrases were more accurately judged than those of the positive phrases, while the warmhearted utterances of the positive phrases were more accurately judged than those of the negative phrases as shown in Fig. 2 (a). The post-hoc test revealed that the coldhearted utterances were 16% more accurately judged than the warmhearted utterances (p=.0004). The linguistically positive phrases were 8% more accurately judged than the negative phrases, although this was not significant (p=.06). The main effect of Sex on the correctness score was marginal but not significant (F(1,72)=3.9, p=.051). The main effect of Manner was significant (F(1,72)=12.5, p<.001), but that of Meaning was not (F(1,72)=2.3, p=.13). As shown in Fig. 2 (b), for response time (RT) in the emotion task, ANOVA revealed that the interaction effect of Meaning and Manner (F(1,72)=10.7, p<.005) were significant. The post-hoc test revealed that the male subjects needed 410 ms longer RT than the female subjects (p<.0001). The interaction effect of Meaning and Manner corresponded to the tendency that RT is shorter for warmhearted utterances of positive phrases than those of negative phrases, while it was shorter for coldhearted utterances of negative phrases than those of positive phrases. The main effect of Sex was also significant (F(1,72)=98.7, p<.0001). For the correctness score in the language task shown in Fig. 3 (a), only the interaction effect of Meaning and Manner (F(1,72)=8.3, p=.05) was significant. This significant interaction corresponded to the tendency that the coldhearted utterances of the negative phrases were more accurately judged than those of the positive phrases, while the warmhearted utterances of the positive phrases were more accurately judged than those of the negative phrases. The ANOVA and the post-hoc test revealed no other significant effects on the correctness score in the language task. As shown in Fig. 3 (b), for response time (RT) in the language task, ANOVA revealed that the main effects of Sex 118 IMAIZUMI, HOMMA, OZAWA, MARUISHI, & MURANAKA

Fig. 3. Results of the language task. (a) Percent correct with standard error bars. (b) Response time with standard error bars. Filled bar: Phrases with negative linguistic meanings; Empty bar: Phrases with positive linguistic meanings. Cold: Coldhearted, Warm: Warmhearted.

(F(1,72)=29.3, p<.0001) and Manner (F(1,72)=9.1, p=.004), and the interaction effect of Meaning and Manner (F=8.9, p=.005) were significant. The post-hoc test revealed that the male subjects needed 256 ms longer RT than the female subjects (p=.004), and the coldhearted utterances were judged 143 ms faster than the warmhearted utterances (p=.04). The significant interaction effect of Meaning and Manner corresponded to the tendency that RT was shorter for warmhearted utterances of positive phrases than those of negative phrases, while it was shorter for coldhearted utterances of negative phrases than those of positive phrases.

Brain Imaging Results: As shown in Fig. 4 and Table 1, the male subjects showed significant activation in more cortical areas than the female subjects for both tasks.

For the male subjects under the emotion task, significant activation (PPWE-corr<.05) was found in the bilateral middle temporal gyri including the superior temporal sulci (STS), the bilateral superior frontal gyri, the left posterior cerebellar lobe and the left inferior frontal gyrus. On the other hand, for the female subjects, the right posterior cerebellar lobe was the only area the activity of which was significant at PPWE-corr<.05 corrected for multiple comparison over the whole brain. As shown in Fig. 5, the activated cortical areas were similar in the male and female subjects for the language task when compared to the control task. There was no cortical area activated significantly when comparison was made between the male and female subjects for the language task. When compared to the female subjects, the male subjects showed significantly stronger activation only under the emotion task in only one cortical area in the right frontomedian cortex (PPWE-corr<.05), as shown in Fig. 6. Other SPM analyses including direct comparison between the emotion and the language tasks, and covariance analyses with RT or correctness score, revealed no significant activations at PPWE-corr<.05 corrected for multiple comparison over the whole brain. GENDER DIFFERENCES IN EMOTIONAL PROSODY PROCESSING 119

Fig. 4. Brain activity pattern revealed in the emotion task – the control task (p<.001 uncorrected for multiple comparison). (a) Female brain; (b) Male brain.

Table 1. Significantly Activated Cortical Areas in Cognitive Subtraction Between Emotion Task Versus Rest, Language Task Versus Rest, and Between Male Versus Female Subjects for Emotion Task Versus Rest. BA: Brodmann Area

Voxel-level Talairach coordinates Task Sex Cortical region BA Hemisphere t P corrected xyz

Emotion- Female posterior cerebellar lobe — Right 10.74 .037 20 –81 –33 Rest Male frontomedian cortex 9 Right 12.16 .010 6 56 32 superior temporal sulcus 21 Right 13.16 .005 57 –31 0 (STS) 21 Right 10.79 .029 63 –25 –2 21 Left 11.49 .018 –57 –24 –4 21 Left 11.06 .024 –61 –31 –4 posterior cerebellar lobe — Left 11.63 .016 –12 –85 –29 inferior frontal gyrus 47 Left 10.46 .036 –40 30 –15

Language- Female superior temporal sulcus 21 Right 12.30 .009 59 –14 –9 Rest (STS) inferior frontal gyrus 47 Left 10.27 .042 –52 41 –5

Male posterior cerebellar lobe — Right 16.23 .001 20 –79 –31 superior temporal sulcus 21 Left 12.68 .007 –63 –35 –2 (STS) 21 Left 11.84 .014 –57 –24 –4 21 Left 10.44 .049 –54 –18 –4

Emotion- Male- frontomedian cortex 9 Right 6.42 .035 6 58 38 Rest Female 120 IMAIZUMI, HOMMA, OZAWA, MARUISHI, & MURANAKA

Fig. 5. Brain activity pattern revealed in the language task – the control task (p<.001 uncorrected for multiple comparison). (a) Female brain; (b) Male brain.

Fig. 6. Significantly activated area of the male subjects when compared to the female subjects under Emotion task (p<.001 uncorrected for multiple comparison).

DISCUSSION

The male subjects in the present study show significant activation in the cortical areas including bilateral superior temporal sulci (STS), the left inferior frontal region, the GENDER DIFFERENCES IN EMOTIONAL PROSODY PROCESSING 121 medial frontal region, and the posterior cerebellum, while the female subjects show only non-significant weak activations in these areas except the posterior cerebellum. These results suggest that the mind reading system for speaking acts consists of neural resources distributed over these areas, and that there are sex-dependent differences in this mind reading system. Some of these cortical areas may be in connection with theory-of-mind processes (Baron-Cohen, Leslie, & Frith, 1985; Ferstl & von Cramon, 2002; Premack & Woodruff, 1978a, b; Siegal, Carrington, & Radel, 1996) or mentalizing processes (Frith & Frith, 1999), i.e. during the attribution of other people’s actions to their motivations, beliefs, or emotions. Frith and Frith (1999), for instance, proposed a distributed neural mentalizing system based on the homologs between studies of single-cell activity in monkeys and areas implicated in mentalizing from brain imaging studies. It consists of STS, for detection of the behavior of agents and analysis of the goals and outcomes of this behavior; (ii) inferior frontal regions, for representations of actions and goals; and (iii) anterior cingulate cortex / frontomedian regions, for representations of mental states of the self. The present study suggests that mind-reading for verbal acts relies also on the posterior cerebellum cooperating with the above mentioned meantalizing system. The activity in the posterior cerebellum was found for both sexes under both tasks, although the location of the maximum activity varied depending on sex and task. The maximum activity for the males located in the left posterior cerebellum under the emotion task and in the right under the language task, while it located in the right posterior cerebellum for the females under the emotion task. This activation suggests a role in the processing of language and emotion for the posterior cerebellum. Evidence has accumulated suggesting that the cerebellum contributes to cognitive functions not only linguistic semantic processing (Roskies, Fiez, Balota, Raichle, & Petersen, 2001), but also affect processing and mentalizing (Gundel, O’Connor, Littrell, Fort, & Lane, 2003), empathy with mirroring for emotive actions (Leslie, Johnson-Frey, & Grafton, 2004). Lesion studies (Schmahmann & Sherman, 1998) suggest that patients with lesions involving the posterior cerebellum are characterized by impairment of verbal executive functions such as verbal fluency, personality change with blunting of affect; and language deficits including agrammatism and dysprosodia. Although the primary role of the cerebellum has been considered to have to do with motor executive processes, recent studies suggest that the cerebellum plays an important role in constructing neural representations or inner models of cognitive actions (Imamizu, Kuroda, Miyauchi, Yoshioka, & Kawato, 2003; Imamizu et al., 2000). Such a modeling function of the cerebellum may also be involved in mind reading for verbal acts through connecting verbal acts with their goals. When compared to the female subjects, the male subjects showed significantly stronger activation in only one cortical area in the right frontomedian cortex only under the emotion task. The frontomedian cortex has been shown to be important for theory-of- mind processes (Frith & Frith, 1999). It has also been suggested that right-hemisphere- damaged but not left-hemisphere-damaged patients have difficulties for theory-of-mind processes (Siegal et al., 1996). Therefore, the present results may suggest that the right 122 IMAIZUMI, HOMMA, OZAWA, MARUISHI, & MURANAKA frontomedian area of the male subjects contributes to analyzing speaking behavior of the female speaker in conjunction with the representation of their own mental states, allowing them to make inferences about the intentions of the female speaker. A question arises as to why the female subjects do not need activations in this area. As shown in the behavioral data, the male subjects showed lower correctness scores with longer RT than the female subjects. Although there was no sex-dependent significant difference in the correctness score, the difference in RT remained significant when measured in the fMRI experiment using selected verbal stimuli. One possible interpretation of the present results is that the male subjects need conscious inferences about the verbal acts and intentions of the female speaker, which required stronger activation of relevant neural substrates with longer RT than the female subjects. Ferstl and von Cramon (2002) suggest that the essential role of the frontomedian area is inference, which is a core component for the theory of mind processes. The female subjects may need less conscious inferences to process emotional prosody. These results may coincide, in part, with previous studies suggesting that emotional prosody modulates perceptual word processing and that the time-course of this modulation differs for males and females, that is, women make an earlier use of emotional prosody during word processing as compared to men (Schirmer et al, 2002), and that women, but not men, show an interaction between prosody and word valence during a semantic processing stage (Schirmer & Kotz, 2003). If so, then, men, not women, would need to integrate linguistic semantics and emotional intent at a higher stage than the semantic processing stage, and therefore men may need extra activations of relevant neural resources and eventually need longer RT than women. As for the organization of the brain for processing language and prosody, the present results suggest that these two aspects significantly interact and are not mutually independent. As shown in Figs. 1, 2, and 3, significant interaction effects of Meaning and Manner were observed on the acoustic characteristics of utterances, such as F0 range, and also on the perceptual behavior evaluated by response time and judgment correctness. These results obtained in our experiment suggest that emotion modulates linguistic processes not only in speech production but also in speech perception, and such modulations may differ between the genders particularly in perceptual and cognitive processes. The correctness score was lower and RT was longer for the incongruent than the congruent stimuli for both sexes under both tasks. This suggests significant mutual interaction occurs between the neural processes of linguistic meanings and of emotional intents. Even though the subjects tried to attend to one aspect ignoring the other, the significant interaction suggests that this was not completely possible. This significant interaction in perceptual and cognitive processes may reflect bilateral hemispheric activations. This may explain why neither right-hemisphere predominance in the emotion task nor left-hemisphere predominance in the language task was observed in this study. In other words, mind reading for verbal acts relies, particularly for males, on bilaterally distributed neural resources. To conclude, the present results suggest that neural resources responsible for speakers’ mind reading are distributed over STS, inferior frontal regions, medial frontal GENDER DIFFERENCES IN EMOTIONAL PROSODY PROCESSING 123 regions and the posterior cerebellum, and sex differences exist in this mind reading system. Neural processes for language and prosody significantly interact to mentalize verbal acts.

REFERENCES

Baron-Cohen, S., Leslie, A. M., & Frith, U. 1985. Does the autistic child have a “theory of mind”? Cognition, 21, 37–46. Baxter, L. C., Saykin, A. J., Flashman, L. A., Johnson, S. C., Guerin, S. J., Babcock, D. R., et al. 2003. Sex differences in semantic language processing: A functional MRI study. Brain and Language, 84, 264– 272 Blonder, L. X., Bowers, D., & Heilman, K. M. 1991. The role of the right hemisphere in emotional communication. Brain, 114, 1115–1127. Borod, J. C. 1993. Cerebral mechanisms underlying facial, prosodic, and lexical emotional expression: A review of neuropsychological studies and methodological issues. Neuropsychology, 7, 445–463. Bradvik, B., Dravins, C., Holtas, S., Rosen, I., Ryding, E., & Ingvar, D. 1991. Disturbances of speech prosody following right hemisphere infarcts. Acta Neurologica Scandinavia, 84, 114–126. Bryan, K. 1989. Language prosody in the right hemisphere. Aphasiology, 3, 285–299. Dykstra, K., Gandour, J., & Stark, R. E. 1995. Disruption of prosody after frontal lobe seizures in the non- dominant hemisphere. Aphasiology, 9, 453–476. Emmorey, K. 1987. The neurological substrates for prosodic aspects of speech. Brain and Language, 30, 305–320. Ferstl, E. C., & von Cramon, D. Y. 2002. What does the frontomedian cortex contribute to language processing: Coherence or theory of mind? Neuroimage, 17, 1599–1612. Frith, C. D., & Frith, U. 1999. Interacting Minds — A Biological Basis. Science, 286, 1692–1695. Frost, J. A., Binder, J. R., Springer, J. A., Hammeke, T. A., Bellgowan, P. S., Rao, S. M., et al. 1999. Language processing is strongly left lateralized in both sexes. Evidence from functional MRI. Brain, 122, 199–208. Gündel, H., O’Connor, M. F., Littrell, L., Fort, C., & Lane, R. D. 2003. Functional neuroanatomy of grief: An fMRI study. American Journal of Psychiatry, 160, 1946–1953. Hampson, E., & Kimura, D. 1992. Sex differences and hormonal influences on cognitive function in humans. In J. D. Becker, M. Breedlove, & D. Crews (Eds.), Behavioral endocrinology. Cambridge, MIT Press. Imaizumi, S., Mori, K., Kiritani, S., Hosoi, H., & Tonoike, M. 1998. Task-dependent laterality for cue decoding during spoken language processing. Neuroreport, 9, 899–903. Imamizu, H., Kuroda, T., Miyauchi, S., Yoshioka, T., & Kawato, M. 2003. Modular organization of internal models of tools in the human cerebellum. Proceedings of the National Academy of Sciences of the United States of America, 100, 5461–5466. Imamizu, H., Miyauchi, S., Tamada, T., Sasaki, Y., Takino, R., Putz, B., et al. 2000. Human cerebellar activity reflecting an acquired internal model of a new tool. Nature, 403, 192–195. Koelsch, S., Maess, B., Grossmann, T., & Friederici, A. D. 2003. Electric brain responses reveal gender differences in music processing. Neuroreport, 14, 709–713. Kotz, S. A., Meyer, M., Alter, K., Besson, M., von Cramon, D. Y., & Friederici, A. D. 2003. On the lateralization of emotional prosody: An event-related functional MR investigation. Brain and Language, 86, 366–376. Ladd, D. R. 1996. Intonational phonology. Cambridge, Cambridge University Press. Leslie, K. R., Johnson-Frey, S. H., & Grafton, S. T. 2004. Functional imaging of face and hand imitation: Towards a motor theory of empathy. Neuroimage, 21, 601–607. McGlone, J. 1977. Sex differences in the cerebral organization of verbal function in patients with unilateral brain lesions. Brain, 100, 775–793. Mitchell, R., Elliot, R., Barry, M., Cruttenden, A., & Woodruff, P. 2003. The neural response to emotional prosody, as revealed by functional magnetic resonance imaging. Neuropsychologia, 41, 1410–1421. 124 IMAIZUMI, HOMMA, OZAWA, MARUISHI, & MURANAKA

Premack, D., & Woodruff, G. 1978a. Chimpanzee problem-solving: A test for comprehension. Science, 202, 532–535. Premack, D., & Woodruff, G. 1978b. Does the chimpanzee have a theory of minds? The Behavioral and Brain Science, 1, 515–526. Roskies, A. L., Fiez, J. A., Balota, D. A., Raichle, M. E., & Petersen, S. E. 2001. Task-dependent modulation of regions in the left inferior frontal cortex during semantic processing. Journal of Cognitive Neuroscience, 13, 829–843. Schirmer, A., Kotz, S. A., & Friederici, A. D. 2002. Sex differentiates the role of emotional prosody during word processing. Cognitive Brain Research, 14, 228–233. Schirmer, A., & Kotz, S. A. 2003. ERP evidence for a sex-specific stroop effect in emotional speech. Journal of Cognitive Neuroscience, 15, 1135–1148. Schmahmann, J. D., & Sherman, J. C. 1998. The cerebellar cognitive affective syndrome. Brain, 121, 561– 579. Shaywitz, B. A., Shaywitz, S. E., Pugh, K. R., Constable, R. T., Skudlarski, P., Fulbright, R. K., et al. 1995. Sex differences in the functional organization of the brain for language. Nature, 373, 607–609. Siegal, M., Carrington, J., & Radel, M. 1996. Theory of mind and pragmatic understanding following right hemisphere damage. Brain and Language, 53, 40–50. Starkstein, S. E., Federoff, J. P., Price, R. C., Leiguarda, R. C., & Robinson, R. G. 1994. Neuropsychological and neuroradiological correlates of emotional prosody comprehension. Neurology, 44, 515–522. Van Lancker, D. 1980. Cerebral lateralization of pitch cues in the linguistic signal. International Journal of Human Communication, 13, 101–109. Vikingstad, E. M., George, K. P., Johnson, A. F., & Cao, Y. 2000. Cortical language lateralization in right handed normal subjects using functional magnetic resonance imaging. Journal of the Neurological Sciences, 175, 17–27. Weintraub, S., Mesulam, M.-M., & Kramer, L. 1981. Disturbances in prosody: A right-hemisphere contribution to language. Archives of Neurology, 38, 742–744. Wellcome Department of Imaging Neuroscience (2003, May 12). Statistical Parametric Mapping: SPM2. (On-line). Available from http://www.fil.ion.ucl.ac.uk/spm/spm2.html Zatorre, R. J., Belin, P., & Penhune, V. B. 2002. Structure and lateralization of auditory cortex: Music and speech. Trends in Cognitive Science, 1, 37–46.

(Manuscript received March 29, 2004; Revision accepted June 8, 2004)