Psychology and Aging 2010, Vol. 25, No. 3, 702–707

© 2010 American Psychological Association 0882-7974/10/$12.00 DOI: 10.1037/a0019266

The Effects of Age on Using Prosody to Convey Meaning and on Judging Communicative Effectiveness Sarah K. Tauber, Lori E. James, and Paula M. Noble University of Colorado, Colorado Springs We tested the effects of aging on the use of prosody to convey meaning and the ability to monitor communicative effectiveness. Participants read aloud ambiguous sentences with the goal of clearly communicating one designated meaning. Young and older adults produced intonational boundaries consistent with the designated meaning equally often, but listener judgments indicated that older adults disambiguated the sentences more often than chance and young adults did so only marginally more often than chance. Young adults believed they communicated their message clearly, and older adults evaluated their own communication even more favorably. Participants were more confident for structurally ambiguous sentences than for lexically ambiguous sentences (which cannot be differentiated through prosody), and older adults demonstrated more overconfidence than young adults for both types of ambiguous sentences. Keywords: aging, communication, prosody, metacognition, confidence ratings

of pauses and vowel lengthening (which necessarily causes word lengthening) that assists the listener in interpreting speech. For example, the structurally ambiguous sentence She will order pork or chicken and fries could mean that she will order pork only or chicken plus fries or that she will order fries whether she orders pork or chicken. If the speaker pauses after pork and elongates the word or, the first meaning above is clearly intended, whereas if the speaker quickly produces pork or chicken and pauses, then elongates the word and, the second meaning is likely intended.1 Speakers often fail to make prosodic modifications that could clarify the meaning of structurally ambiguous sentences: Keysar and Henly (2002) found that speakers often failed to provide prosodic cues to disambiguate sentences produced in isolation, even when speakers knew that listeners had no situational, contextual, or physical information to assist in their interpretation of the ambiguity. Albritton, McKoon, and Ratcliff (1996) also found that participants generally failed to use prosody to disambiguate meaning, with the exception of trained professional speakers, who used prosody only when specifically instructed to do so. Much of the research on speech prosody in older adults has focused on patients with Parkinson’s disease (PD), who often fail to produce appropriate prosody when speaking (Canter, 1963). Unlike healthy age-matched controls, patients with PD (mean age ⫽ 65 years) did not make prosodic alterations to disambiguate noun compounds (e.g., a blackboard) from noun phrases (e.g., a black board; Darkins, Fromkin, & Benson, 1988). The prosody of patients with PD (mean age ⫽ 62 years) was rated as less normal than that of age-matched controls on a storytelling task (Caekebeke, Jennekens-Schinkel, van der Linden, Buruma, & Roos, 1991), and naive listeners were better able to identify the intended

The ability to use language to communicate our thoughts clearly would seem to be a highly practiced skill at which people generally succeed. Additionally, it seems that speakers would monitor their language production carefully to ensure listener comprehension. However, a growing body of literature indicates that potential miscommunications abound. The present research tested the ability of young and older adults to convey meaning clearly using prosody and their ability to monitor their communicative effectiveness.

Prosody in Language Production Prosody is the pattern of pausing, stress, and intonation produced for a spoken utterance, and it is an important part of linguistic communication. Prosody can arise for many reasons (Ferreira, 2007), but one clear goal of prosody is to help convey the meaning of an utterance. Ambiguity is ubiquitous in language, but some potentially ambiguous utterances can be clarified if the speaker alters his or her prosody. Correctly produced linguistic prosody has been shown to assist listeners in understanding spoken messages (e.g., Sanderman & Collier, 1997), specifically in interpreting otherwise-ambiguous utterances (e.g., Schafer, Carlson, Clifton, & Frazier, 2000). Speakers can create “intonational boundaries” (Cooper & Paccia-Cooper, 1980) using a combination

Sarah K. Tauber, Lori E. James, and Paula M. Noble, Department of Psychology, University of Colorado, Colorado Springs. Portions of this research were presented at the Cognitive Aging Conference, Atlanta, Georgia, April 2006, and the Rocky Mountain Psychological Association Conference, Denver, Colorado, April 2007. We thank Elizabeth Angel, Connie Isele, and especially Megan Miller for research assistance and Alan Castel for helpful discussion regarding the manuscript. Correspondence concerning this article should be addressed to Lori E. James, Department of Psychology, University of Colorado, Colorado Springs, 1420 Austin Bluffs Parkway, Colorado Springs, CO 80933-7150. E-mail: [email protected]

1 Lexical ambiguities (e.g., pipe in John was sad when the pipe broke, which could refer to either a water pipe or a smoking pipe) are not clarified by modifying the prosody of the sentence.

702

BRIEF REPORTS

meaning for control speakers (mean age ⫽ 63 years) than for speakers with PD (mean age ⫽ 62 years; Pell, Cheang, & Leonard, 2006).2 Overall, healthy older adults seem to use prosody more effectively than patients with PD do. The present study focused on comparisons of prosody use in healthy older versus young adults, because little is currently known about how aging impacts speech prosody beyond what can be inferred from comparing healthy older adults with patients with PD. Aging has been shown to have a positive effect or no effect on the use of prosody to express an emotion (Orbelo, Testa, & Ross, 2003). Kemper, Ferrell, Harden, Finter-Urczyk, and Billington (1998) compared the prosody of young and older adults describing a map route to a person whom they believed to be either a demented or a nondemented older listener. Young adults varied their pitch more, but the overall prosody of young and older adults was very similar (regardless of presumed cognitive status of the listener). Zanini, Bryan, De Luca, and Bava (2005) reported that adults in their 20s and 30s received higher ratings for prosody use in conversation than adults in their 70s. These contradictory findings are likely due to task differences between those studies (map description vs. conversation). The current experiment tested for age differences in prosody use when attempting to disambiguate the meaning of syntactically ambiguous sentences, which has not been explored in previous research.

Metacognitive Monitoring of Communicative Effectiveness Keysar and Henly (2002) tested the possibility that although speakers may produce ambiguous utterances, they carefully monitor listeners’ comprehension to be certain that communication succeeds. They found that participants systematically overestimated their communicative effectiveness: Speakers were highly confident that their meanings would be understood by naive listeners, even when listeners had to guess the intended meaning because the speaker provided no disambiguating prosodic clue. Goranson (cited in Kelley & Jacoby, 1996) also demonstrated that people overestimate their communicative success during a game by relying on their own knowledge to estimate whether others would be able to solve puzzles. Kruger, Epley, Parker, and Ng (2005) showed a propensity for miscommunication via e-mail messages intended to convey serious versus sarcastic content (e.g., I loved that movie, which can be interpreted literally or as the opposite of the literal meaning). Participants were not told how to convey sarcasm (an impossible task, in that an isolated sentence sent via e-mail does not allow visual or verbal cues), and they routinely failed to realize that the message recipient would be unable to determine the intended meaning. Older adults may have more difficulty than young adults in taking the perspective of others. For example, older adults make more egocentric error responses on Piaget-like visual tasks, show longitudinal declines in integrating multiple perspectives in their moral judgments, make more dispositional attributions for behaviors, and are less likely to adopt the perspectives of others when making judgments of behavior likelihood (e.g., Blanchard-Fields, 1994; Inagaki et al., 2002; Ligneau-Herve & Mullet, 2005; Pratt, Diessner, Pratt, Hunsberger, & Pancer, 1996). Older adults make judgment errors based on something akin to illusory transparency (e.g., Keysar, 1994) and the curse of knowledge (Kelley & Jacoby,

703

1996), under which a failure to adequately take others’ perspectives and ignore one’s own knowledge leads to inaccurate reasoning about others’ interpretations. This failure to monitor the difference between one’s own knowledge and that of a person attempting to comprehend a message is a metacognitive error. There has been a recent increase in attention to the metacognitive skills of older adults, especially their reduced ability to process information about their memory functioning. For example, a longitudinal study of metamemory in young and older adults found that older adults rely more on their theories of memory ability rather than actual monitoring of their memory performance (McDonald-Miszczak, Hertzog, & Hultsch, 1995). Older adults have sometimes, but not always, been shown to be more overconfident in their memory abilities than young adults are (e.g., Dodson, Bawa, & Krueger, 2007; see also Woo, SchmitterEdgecombe, & Fancher, 2007). Metacognitive assessments of older adults in domains other than memory have rarely been studied. The decreased likelihood of older adults taking others’ perspectives, combined with their potential for increased confidence, suggests that older adults will be overconfident to an even greater extent than young adults when assessing their communication effectiveness.

The Present Study We tested young and older adults’ ability to clearly communicate one designated meaning of an ambiguous sentence. For syntactically ambiguous sentences, prosodic alterations were required to differentiate the possible meanings. We predicted that our results would replicate those of Keysar and Henly (2002), namely, that neither young nor older adults would consistently make adequate alterations to their prosody to disambiguate their utterances. Speakers also rated their own communicative effectiveness, judging whether they believed a naive listener would understand the intended meaning of the isolated sentence they produced in the absence of any contextual information. We predicted larger mean confidence ratings for older than young adults, regardless of whether correct prosody was employed. Additionally, we expected confidence to vary by sentence type such that both age-groups would be more confident for structurally ambiguous sentences than for lexically ambiguous sentences.

Method Participants Forty-two young adults and 45 older adults were tested. Eight older participants’ data were excluded due to impaired vision, a lack of usable data, or low scores (26 and below) on the MiniMental Status Exam (MMSE; Folstein, Folstein, & McHugh, 1975). Therefore, we had usable data from 42 young adults (mean age ⫽ 20.63 years, SD ⫽ 2.05; 67% female) and 37 healthy older adults (mean age ⫽ 74.14 years; SD ⫽ 7.04, mean MMSE ⫽ 2 However, listeners in Pell et al. (2006) made many errors comprehending the speech of healthy controls, a finding that might reflect inadequate prosodic alterations by the speakers, incompetence of naive listeners in comprehending well-produced prosody (as in Keysar & Henly, 2002), or a combination of these factors.

704

BRIEF REPORTS

29.46, SD ⫽ 2.42; 68% female) who were native speakers of American English with self-reported normal hearing. Young adults had marginally less education (M ⫽ 14.42 years, SD ⫽ 1.61) than older adults did (M ⫽ 15.49 years, SD ⫽ 3.02), t(71) ⫽ 1.88, p ⬍ .06. Young participants received course credit, and older participants were paid $10.

Materials and Procedure We selected 15 ambiguous sentences containing phrases judged to have similar frequency for each of two interpretations, based on pilot testing in which young and older participants judged the percentage of time a given word or phrase (without any context) reflected each of two possible meanings. Among the stimulus sentences, four were “structurally ambiguous” and contained ambiguities that could be easily disambiguated with appropriate use of prosody to create intonational boundaries. For example, The lake froze over a month ago can mean “the lake froze more than one month previously” if the word froze is elongated and a pause is inserted after it to create an intonational boundary, or it can mean “the lake completely froze one month ago” if the intonational boundary is created by elongating the word over and inserting a pause after it. Five sentences contained lexical ambiguities that could never be disambiguated with prosody (e.g., Chris sat on the board). There were six additional filler sentences, which were not analyzed. Two structurally ambiguous practice sentences preceded the experimental sentences. Two paragraphs were developed for each sentence. Each paragraph was intended to evoke one particular interpretation of the ambiguous sentence. Stimuli were divided into two sets with each set containing one scenario (i.e., one meaning) for each sentence, and half of the participants in each age-group were randomly assigned to each set. Participants were given a booklet that had three pages associated with each sentence. On the first page, the participant saw the sentence in isolation, read it, and attempted to generate two possible meanings for the sentence. If the participant did not generate two viable and distinct meanings, the experimenter provided the second meaning and discussed it with the participant until it was clear that the meaning was understood. This ensured that all participants knew precisely what needed to be disambiguated in their subsequent production of the sentence. On the next page, participants read the paragraph containing the biasing context with the target sentence printed below it. They read the scenario silently and then produced the target sentence aloud into a tape recorder. Participants were told that someone would listen to the recording at a later time to determine which meaning they intended to convey and that the listener would hear only the sentence (without reading the biasing scenario), so they should be sure to convey their meaning clearly. The final page included summary paraphrases of each possible meaning, and participants circled the meaning they had intended to communicate. This ensured that speakers correctly understood the scenario and that their intended meaning was the designated one. Beneath this was a 6-point rating scale on which speakers indicated how confident they were that they had conveyed the intended meaning (0 indicated “I was unable to clearly communicate the meaning I intended”; 5 indicated “I was able to clearly communicate the meaning I intended”). For lexical ambiguities the correct rating was always 0, but for syntactic ambigu-

ities ratings should vary with prosody use (but participants were not told this). All participants were tested individually and in quiet laboratory rooms to ensure quality tape recording.

Results We first removed trials on which the participant indicated that he or she intended to convey the incorrect meaning of the sentence, a fairly rare occurrence that occurred less for young adults (M ⫽ 5%, SD ⫽ 7%) than for older adults (M ⫽ 12%, SD ⫽ 15%), t(77) ⫽ 2.75, p ⬍ .01. These problems are assumed to primarily reflect a failure to correctly comprehend the passage leading to attempted production of the incorrect meaning of the sentence, but they may also reflect memory failures (i.e., participants failing to recall which meaning they had attempted to communicate).

Use of Prosody to Convey Meaning We examined prosody for the four structurally ambiguous sentences as determined by listening to the recorded responses. There were three judges, all blind as to which meaning the participant was attempting to convey in producing the sentence. Two judges were trained to identify the prosody corresponding with each possible meaning of these sentences, and they independently listened to every sentence produced by each speaker and identified whether the prosody was consistent with either of the two possible meanings. For 75% of sentences, the two judges agreed about the prosody. The third judge listened to all sentences for which the two initial judges disagreed, and she agreed with one of their determinations for another 19% of sentences. The remaining sentences were excluded from further analyses. The percentage of trials with unusable data was virtually identical for young adults (M ⫽ 7%) and older adults (M ⫽ 6%). The percentage of syntactically ambiguous sentences successfully disambiguated via prosody was numerically but nonsignificantly greater for older adults (M ⫽ 66%, SD ⫽ 41%) than for young adults (M ⫽ 59%, SD ⫽ 31%; t ⬍ 1). T tests comparing results from each age-group against 50% (chance) indicated that older adults correctly disambiguated sentences through prosody more often than chance, t(36) ⫽ 2.37, p ⬍ .05, a difference that was marginally significant for young adults, t(41) ⫽ 1.89, p ⬎ .06. We then compared how many young versus older adult participants correctly used prosody to disambiguate all syntactically ambiguous sentences. Significantly more older adults (49%) than young adults (24%) consistently disambiguated all sentences, t(77) ⫽ 2.35, p ⬍ .05. To improve upon the necessarily subjective scoring of prosody used in that analysis, we digitized the sentence recordings into Sound Edit 16 (Macromedia, discontinued) and measured intonational boundary times from the visual waveform. We defined intonational boundaries as pause length plus pronunciation duration for critical words (Cooper & Paccia-Cooper, 1980) and measured the intonational boundary times at the “correct” and “incorrect” locations within the sentence. For example, for the sentence The lake froze over a month ago, a correct intonational boundary occurs when a participant elongates the word “over” and pauses after it when he or she intends to communicate the possible meaning that the lake was completely frozen (i.e., “frozen over”) one month ago. We measured both the correct intonational bound-

BRIEF REPORTS

ary duration (at the location in the sentence that would communicate the desired meaning) and the incorrect intonational boundary duration (at the location in the sentence that would communicate the other, nondesired meaning). Intonational boundary durations were analyzed in a 2 (correct vs. incorrect) ⫻ 2 (young vs. older adults) analysis of variance (ANOVA; see Table 1). Correct boundary durations were significantly longer than incorrect boundary durations, F(1, 77) ⫽ 27.77, p ⬍ .001, ␩2p ⫽ .27, and older adults had longer boundary durations than young adults did, F(1, 77) ⫽ 9.76, p ⬍ .01, ␩2p ⫽ .11. The pattern of correct and incorrect intonational boundary durations did not differ by age (F ⬍ 1).

Confidence Ratings A 2 (young vs. older adults) ⫻ 2 (structurally ambiguous vs. lexically ambiguous sentences) mixed ANOVA was conducted on mean confidence ratings to compare ratings for the structurally ambiguous sentences and the sentences with lexical ambiguities (which cannot be disambiguated via prosody). As shown in Figure 1 (left panel), there was a main effect of age, F(1, 77) ⫽ 17.33, p ⬍ .01, ␩2p ⫽ .18, and a main effect of sentence type, F(1, 77) ⫽ 24.70, p ⬍ .01, ␩2p ⫽ .24, and a marginally significant interaction of age and sentence type, F(1, 77) ⫽ 3.78, p ⬍ .06, ␩2p ⫽.05. Young adults provided higher confidence ratings for structurally ambiguous sentences than for sentences with lexical ambiguities, t(41) ⫽ 4.63, p ⬍ .01, and older adults showed the same pattern but with a smaller numerical difference, t(36) ⫽ 2.34, p ⬍ .05. Older adults were significantly more confident than young adults that they had disambiguated the structurally ambiguous sentences, t(77) ⫽ 3.29, p ⬍ .01; however, they were also significantly more confident that they had disambiguated the sentences with lexical ambiguities, t(77) ⫽ 3.73, p ⬍ .01. Because lexically ambiguous sentences cannot be disambiguated via prosody, this finding indicates that older adults overestimate their communicative effectiveness more than young adults do. Next we compared the confidence ratings for correctly disambiguated sentences and those that were not disambiguated (as determined by the judges) in a 2 (young vs. older adults) ⫻ 2 (correct vs. incorrect prosody) mixed ANOVA. Many participants (16 young and 25 older adults) were excluded from this analysis for not having produced at least one sentence with correct prosody, and this exclusion is reflected in the degrees of freedom. There was a main effect of age, F(1, 36) ⫽ 4.57, p ⬍ .05, ␩2p ⫽ .11, because older adults rated their effectiveness more highly than young adults did, and a main effect of prosody use, F(1, 36) ⫽ 4.70, p ⬍ .05, ␩2p ⫽ .12, because ratings were higher for sentences with

Table 1 Mean Correct and Incorrect Intonational Boundary Durations for Young and Older Participants Intonational boundary duration (ms)

Young adults

Older adults

Correct Incorrect Overall

550 (224) 413 (80) 482 (126)

667 (282) 489 (100) 578 (147)

Overall 605 (258) 449 (100)

Note. Standard deviations are shown in parentheses. Intonational boundary duration ⫽ pause time plus pronunciation duration.

705

correct prosody than for those with incorrect prosody. However, there was no interaction between age and prosody use (F ⬍ 1; see Figure 1, right panel).

Discussion Aging and Prosody Use Both young and older participants produced intonational boundaries consistent with the correct prosody to communicate a particular sentence meaning. Listener ratings generally comported with the intonational boundary analysis: We did not find a significant age difference in listener judgments of use of correct prosody to convey meaning, but young adults used correct prosody marginally more often than chance, and older adults used correct prosody significantly more often than chance. We also found that more older than young adults were judged to have consistently altered their prosody to disambiguate all of the syntactically ambiguous stimulus sentences. These findings suggest that even when speakers make correct adjustments to intonational boundaries, their modifications are often too subtle to be detected by listeners. This result comports with those of Keysar and Henly (2002), who found that listener identification of the intended meaning of a spoken structurally ambiguous sentence was statistically above chance but that communication via prosody use was also far from perfect. On the whole, older adults produced longer intonational boundaries than young adults did, a difference that may have made older adult prosody easier for judges to detect when listening to the recordings. In other words, perhaps older adult prosody was benefited by overall slower speech rates, resulting in more successful disambiguation. An interesting question for future research is whether young and older speakers can be taught to produce intonational boundaries that are more salient to listeners, thereby reducing the potential for miscommunication.3 The obtained age differences in prosody production favor older adults, indicating that in spite of age-related declines in many cognitive functions that can negatively impact language production (e.g., Burke & MacKay, 1997; Thornton & Light, 2006) older adults are not impaired in the ability to use prosody to convey sentence meaning. Our measure of intonational boundary production did not indicate age differences in prosody use. Both agegroups made correct alterations to their prosody to convey meaning (e.g., longer pausing and word lengthening in the correct than incorrect location), a finding consistent with the lack of age differences in prosody production found by Kemper et al. (1998). In sum, older adults use prosody to communicate at least as effectively as young adults do, with several factors likely contributing to this pattern. For example, prosody may be one of many aspects of language processing that do not decline with age (e.g., Burke & Shafto, 2008). Additionally, greater experience with the pragmatic and communicative aspects of speech may have sensitized older adults to the importance of clarity in communication (e.g., James, Burke, Austin, & Hulme, 1998; Orbelo et al., 2003). 3

Kraljic and Brennan (2005) indicated that the presence of a live addressee might influence the likelihood of disambiguating speech. The role of a live addressee on prosody use in young versus older adults should be investigated in the future.

BRIEF REPORTS

706 5.00

Young Adults 4.50

Older adults

4.00

Mean Rating

3.50

3.00

2.50

2.00

1.50

1.00

0.50

0.00

Lexical

Structural

Sentence Type

Correct

Incorrect

Prosody for Structurally Ambiguous Sentences

Figure 1. Mean confidence ratings for young and older adults by sentence type (left panel) and by correct vs. incorrect use of prosody for structurally ambiguous sentences (right panel). Error bars represent 1 SE.

Aging and Judgments of Communicative Effectiveness Older adults were more confident than young adults in their ability to disambiguate the sentences with structural ambiguities, and given that they used correct prosody (nonsignificantly) more often than young adults did, these assessments may be realistic. However, older adults were also more confident than young adults that they disambiguated both syntactic ambiguities for which they did not produce correct prosody and lexical ambiguities, where prosodic disambiguation is impossible. These findings indicate age-related increases in the overestimation of communicative effectiveness.4 There was some evidence that speakers from both age-groups can be sensitive to their communicative accuracy in that ratings were higher for correctly disambiguated than for nondisambiguated sentences. Perhaps the most interesting finding is that both young and older adults judged their communication of lexical ambiguities to be quite effective, in spite of the fact that lexical ambiguities can never be clarified through prosody but rather require linguistic or physical context for disambiguation. Our data dovetail with the metacognitive literature that demonstrates inaccuracies in metacognitive awareness related to memory (e.g., Koriat, 1997; Koriat, Bjork, Sheffer, & Bar, 2004) and an increase in these errors during healthy aging (e.g., McDonaldMiszczak et al., 1995). Correcting metacognitive errors in prosody production by reducing overconfidence (i.e., monitoring errors) should lead to more effective communication. Future research might explore methods of improving self-reflective sensitivity of communication abilities in order to see improvements in the use of

prosody. Also of interest, our results indicated a marginal interaction in confidence assessments such that young adult confidence varied more by sentence type (i.e., structural vs. lexical ambiguities) than did older adult confidence. This may indicate that young adults are somewhat better at detecting differences in successful prosody production based on type of ambiguity, a pattern that future research should further explore. In sum, participants often failed to adequately adjust their prosody to disambiguate their language production, and in spite of this failure to appropriately modify speech to avoid miscomprehension by listeners, speakers were remarkably overconfident that their language would be understood. There are a variety of domains in which people are unaware of their own incompetence (e.g., Dunning, Heath, & Suls, 2004), but the finding that young and older adults have this problem in language, a skill with which people have extensive experience, is particularly intriguing. The importance of accurate communication suggests that future research should explore strategies or training to improve our communication effectiveness or at least to ensure that we more accurately monitor our communication abilities.

4 Confidence ratings may differ across age due to cohort effects. That is, perhaps different generations differ in how conservatively they judge their performance on all tasks. The present findings may therefore allow us to conclude only that older adults are more overconfident than young adults, not that aging causes increased overconfidence.

BRIEF REPORTS

References Albritton, D. W., McKoon, G., & Ratcliff, R. (1996). Reliability of prosodic cues for resolving syntactic ambiguity. Journal of Experimental Psychology: Learning, Memory, and Cognition, 22, 714 –735. doi: 10.1037/0278-7393.22.3.714 Blanchard-Fields, F. (1994). Age differences in causal attributions from an adult developmental perspective. Journal of Gerontology, 49, 43–51. Burke, D. M., & MacKay, D. G. (1997). Memory, language, and ageing. Philosophical Transactions of the Royal Society B: Biological Sciences, 352, 1845–1856. doi:10.1098/rstb.1997.0170 Burke, D. M., & Shafto, M. A. (2008). Language and aging. In F. I .M. Craik & T. A. Salthouse (Eds.), The handbook of aging and cognition (3rd ed., pp. 373– 443). Mahwah, NJ: Erlbaum. Caekebeke, J. F. V., Jennekens-Schinkel, A., van der Linden, M. E., Buruma, O. J. S., & Roos, R. A. C. (1991). The interpretation of dysprosody in patients with Parkinson’s disease. Journal of Neurology, Neurosurgery & Psychiatry, 54, 145–148. doi:10.1136/jnnp.54.2.145 Canter, G. J. (1963). Speech characteristics of patients with Parkinson’s disease: I. Intensity, pitch and duration. Journal of Speech and Hearing Disorders, 28, 221–229. Cooper, W. E., & Paccia-Cooper, J. (1980). Syntax and speech. Cambridge, MA: Harvard University Press. Darkins, A. W., Fromkin, V. A., & Benson, D. F. (1988). A characterization of the prosodic loss in Parkinson’s disease. Brain and Language, 34, 315–327. doi:10.1016/0093-934X(88)90142-3 Dodson, C. S., Bawa, S., & Krueger, L. E. (2007). Aging, metamemory, and high-confidence errors: A misrecollection account. Psychology and Aging, 22, 122–133. doi:10.1037/0882-7974.22.1.122 Dunning, D., Heath, C., & Suls, J. M. (2004). Flawed self-assessment: Implications for health, education, and the workplace. Psychological Science in the Public Interest, 5, 69 –100. doi:10.1111/j.15291006.2004.00018.x Ferreira, F. (2007). Prosody and performance in language production. Language and Cognitive Processes, 22, 1151–1177. doi:10.1080/ 01690960701461293 Folstein, M. F., Folstein, S. E., & McHugh, P. R. (1975). Mini-mental state: A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research, 12, 189 –198. doi: 10.1016/0022-3956(75)90026-6 Inagaki, H., Meguro, K., Shimada, M., Ishizaki, J., Okuzumi, H., & Yamadori, A. (2002). Discrepancy between mental rotation and perspective-taking abilities in normal aging assessed by Piaget’s ThreeMountain task. Journal of Clinical and Experimental Neuropsychology, 24, 18 –25. doi:10.1076/jcen.24.1.18.969 James, L. E., Burke, D. M., Austin, A., & Hulme, E. (1998). Production and perception of “verbosity” in younger and older adults. Psychology and Aging, 13, 355–367. doi:10.1037/0882-7974.13.3.355 Kelley, C. M., & Jacoby, L. L. (1996). Adult egocentrism: Subjective experience versus analytic bases for judgment. Journal of Memory and Language, 35, 157–175. doi:10.1006/jmla.1996.0009 Kemper, S., Ferrell, P., Harden, T., Finter-Urczyk, A., & Billington, C. (1998). Use of elderspeak by young and older adults to impaired and umimpaired listeners. Aging, Neuropsychology, and Cognition, 5, 43– 55. doi:10.1076/anec.5.1.43.22 Keysar, B. (1994). The illusory transparency of intention: Linguistic perspective taking in text. Cognitive Psychology, 26, 165–208. doi:10.1006/ cogp.1994.1006

707

Keysar, B., & Henly, A. S. (2002). Speakers’ overestimation of their effectiveness. Psychological Science, 13, 207–212. doi:10.1111/14679280.00439 Koriat, A. (1997). Monitoring one’s own knowledge during study. A cue-utilization approach to judgments of learning. Journal of Experimental Psychology: General, 126, 349 –370. doi:10.1037/00963445.126.4.349 Koriat, A., Bjork, R. A., Sheffer, L., & Bar, S. K. (2004). Predicting one’s own forgetting: The role of experience-based and theory-based processes. Journal of Experimental Psychology: General, 133, 643– 656. doi:10.1037/0096-3445.133.4.643 Kraljic, T., & Brennan, S. E. (2005). Prosodic disambiguation of syntactic structure: For the speaker or for the addressee? Cognitive Psychology, 50, 194 –231. doi:10.1016/j.cogpsych.2004.08.002 Kruger, J., Epley, N., Parker, J., & Ng, Z.-W. (2005). Egocentrism over e-mail: Can we communicate as well as we think? Journal of Personality and Social Psychology, 89, 925–936. doi:10.1037/0022-3514.89.6.925 Ligneau-Herve, C., & Mullet, E. (2005). Perspective-taking judgments among young adults, middle-aged, and elderly people. Journal of Experimental Psychology: Applied, 11, 53– 60. doi:10.1037/1076898X.11.1.53 McDonald-Miszczak, L., Hertzog, C., & Hultsch, D. F. (1995). Stability and accuracy of metamemory in adulthood and aging: A longitudinal analysis. Psychology and Aging, 10, 553–564. doi:10.1037/08827974.10.4.553 Orbelo, D. M., Testa, J. A., & Ross, E. D. (2003). Age-related impairments in comprehending affective prosody: Comparison to brain damaged subjects. Journal of Geriatric Neurology and Psychiatry, 16, 44 –52. doi:10.1177/0891988702250565 Pell, M. D., Cheang, H. S., & Leonard, C. L. (2006). The impact of Parkinson’s disease on vocal-prosodic communication from the perspective of listeners. Brain and Language, 97, 123–134. doi:10.1016/ j.bandl.2005.08.010 Pratt, M. W., Diessner, R., Pratt, A., Hunsberger, B., & Pancer, S. M. (1996). Moral and social reasoning and perspective taking in later life: A longitudinal study. Psychology and Aging, 11, 66 –73. doi:10.1037/ 0882-7974.11.1.66 Sanderman, A. A., & Collier, R. (1997). Prosodic phrasing and comprehension. Language and Speech, 40, 391– 409. Schafer, A., Carlson, K., Clifton, C., Jr., & Frazier, L. (2000). Focus and the interpretation of pitch accent: Disambiguating embedded questions. Language and Speech, 43, 75–105. doi:10.1177/00238309000430010301 Thornton, R., & Light, L. L. (2006). Language comprehension and production in normal aging. In J. Birren & W. Schaie (Eds.), Handbook of the psychology of aging (6th ed.). San Diego, CA: Academic Press. doi:10.1016/B978-012101264-9/50015-X Woo, E., Schmitter-Edgecombe, M., & Fancher, J. B. (2007). Memory prediction accuracy in younger and older adults: A cross-sectional and longitudinal analysis. Aging, Neuropsychology, and Cognition, 15, 68 – 94. doi:10.1080/13825580701626936 Zanini, S., Bryan, K., De Luca, G., & Bava, A. (2005). The effects of age and education on pragmatic features of verbal communication: Evidence from the Italian version of the Right Hemisphere Language Battery (I-RHLB). Aphasiology, 19, 1107–1133. doi:10.1080/02687030500268977

Received June 17, 2009 Revision received January 14, 2010 Accepted January 29, 2010 䡲