Intergenerational communication beliefs across thelifespan: comparative data from India.

This article examines U.S. American and East Indian young adults' perceptions of intergenerational communication. Irrespective of culture, as age of target increased, so did attributions of benevolence, norms of politeness and deference, and communicative respect and avoidance; conversely, attributions

of personal vitality and communication satisfaction decreased linearly. Young adults' avoidant communication with older people negatively predicted their communication satisfaction. In India, age stereotypes and politeness norms predicted communication satisfaction. The less young Indians felt a need to be polite to older people and the more they perceived them as benevolent and active, the more communication satisfaction they reported with elders.

Keywords: Age Norms; Age Stereotypes; Communication Beliefs; Communication Satisfaction; Deference; Elderly; India; Respect

Introduction

Research in Western societies has shown that young people construe communication with nonfamily elderly adults as largely problematic and unsatisfactory Williams & Nussbaum, 2001). For example, young adults report that they are often patronized by older adults in the sense that the latter seem overly nurturing, do not listen, and express negative stereotypes to them about their young age group (Giles & Williams, 1994). Older adults can also be perceived as "nonaccommodative" via their communication styles in that they are perceived to be authoritarian, dismissive, and inattentive. In addition, young people report feeling obliged to be polite and to defer to the older adult in such exchanges (Williams & Giles, 1996). Given these communicative gulfs, Coupland, Coupland, and Giles (1991) concluded that "younger and older people may find themselves conversing across a cultural divide, predisposed by their predictably varying social experiences, social attitudes and priorities for interaction" (p. 152).

This same pattern of intergenerational communication dissatisfaction, at least from the perspective of young adults, has also been documented in many East Asian contexts (Ota, Giles, & Somera, 2007). For instance, while intragenerational communication among young adults is reportedly very similar between Asian and Western sites, studies have shown that the intergenerational communication climate in Australia and the USA is perceived to be more favorable than in the People's Republic of China, Hong Kong, Japan, and Taiwan (for review, see Giles, McCann, Ota, & Noels, 2002). In fact, despite reporting higher respect and obligation for older adults, East Asian young adults report more nonaccommodation from, and avoidance of, older adults than do their Western counterparts.

Current research then underscores the critical role of culture in understanding intergenerational beliefs about communication. Obviously, data from other settings are important to garner as we move toward formulating more robust theoretical models. Accordingly, we extend this program of research into India as one means of geographically bridging our previous foci on Asia, it being an area of demographic and communication technological growth. Additionally, our choice here was predicated on the consistent finding that norms of filial piety in Asian settings did not translate into more positive reports of communication to older people from young people (Giles et al., 2002). The present study provided us with the unique possibility of examining yet another culture that purportedly values older people in its religious philosophy and practices.

Theoretical Underpinnings and Theory Building

This research has astrong theoretical basis and is built on well-established principles inherent in communication accommodation theory (CAT; cf Gallois, Ogay, & Giles, 2005). According to CAT, our communicative behaviors are, at least in part, fueled by social stereotypes, many of which are negative, but some of which are positive (Hummert, 1994). Recent studies have shown that as age of target increases from young adulthood, to middle age, and to elderliness, so do trait attributions of benevolence (e.g., kindness), while attributions of personal vitality (e.g., attractiveness) decrease linearly (Giles, Makoni, & Dailey, 2005; McCann, Dailey, Giles, & Ota, 2005).

Perhaps not surprisingly, negative age stereotypes (such as declining personal vitality) have potent behavioral implications and, in a number of more specific theoretical models of communication and aging, these have been regularly posited as triggers for overaccommodating talk to older people (e.g., Ryan, Giles, Bartolucci, & Henwood, 1986). Such speech would manifest in a slowing down of talking rate, using overly polite nonverbals, and crafting simpler grammatical forms, which have been perceived as demeaning and patronizing by cognitively and socially active older people (e.g., Ryan, Hummert, & Boich, 1995).

In addition, given findings overviewed above, these models suggest how cultural norms for talking with older people, such as obligations to respect and defer to them, contribute to the equation (Barker, Giles, & Harwood, 2004). Further, beyond the reports of communicative behaviors towards older people, researchers have focused on how these behaviors may impact younger adults' satisfaction with and enjoyment of intergenerational conversations (e.g., Giles et al., 2005). In pursuing the theoretical goal of mapping out young people's age stereotypes (i.e., benevolence and personal vitality), age norms (i.e., politeness and deference), and their communicative behaviors (i.e., respect and avoidance) across different cultures, we previously conducted two studies surveying students in the USA, Ghana, and South Africa (Giles et al., 2005; McCann et al., 2005). Cross-cultural differences notwithstanding, reported communication behaviors (viz., respect and avoidance) were the strongest predictors of satisfying and enjoyable conversations with older people.

Communication and Aging in India

Intergenerational communication scholars have afforded little empirical attention to aging in this cultural context (Liebig, 2003). Interestingly, India has many parallels with comparable work in other parts of Asia. As elsewhere, India is presently experiencing an exponential growth in its aging population that will have serious repercussions in terms of the availability of health care services as well as pension and social security payments. Ranjan, Sarma, and Mishra (2003) projected an increase of older persons aged 80 years and over from 71 million in 2001 to 179 million by 2031. Many older Indians live in multi-generational households and a substantial proportion of them suffer physical ailments and endure psychological distress (Joshi, Kumar, & Avasthi, 2003). Indeed, many of these individuals experience forms of dementia that are "better tolerated in Indian homes where it is considered by most people to be a natural concomitant of senescence" (Neki, 1987, pp. 87-88). The tradition in India is that older people must be respected and revered: "One repays one's duty to the parents and also paves one's way to salvation" (Fonseca, 1999, p. 162). Indeed, the Hindu religion maintains that failure to take care of older parents signals harsh consequences in the afterlife.

However, other studies have underscored an erosion of societal values as the younger generations search for new identities encompassing economic independence (Bhat & Druvarajan, 2001). There is a general consensus among young people that their elders have not adapted to changing values and, thus, not only do young people have negative images of older people, but also express little desire to interact with them and do not enjoy their company (Patel & Prince, 2001). Furthermore, older people are often and even actively avoided, with intergenerational conflicts being reportedly frequent (Nayar, 2000). Given the above contrastive perspective and this being the first study of its kind in India, our preference was to conservatively proffer research questions below.

The Present Study: Extending Previous Research

Herein, we also continue our attention on midlife, for although this is a potent developmental phase (e.g., Helson & Soto, 2005), little attention has been afforded it in intergenerational communication research (however, cf Harwood & Giles, 1993). Therefore, recent studies have invited young adults to rate the communicative features of conversations with older people and their peer-aged but have now included also middle-aged people into the evaluative frame alongside them (e.g., Giles et al., 2005). These investigations have also examined young adults' norms of respect for older people in terms of politeness and deference (e.g., Gallois et al., 1999) as well as communication satisfaction with young adult, middle-aged, and older adult targets (Williams & Giles, 1996).

Multivariate analyses of the data from these two studies (Giles et al., 2005; McCann et al., 2005) showed that as age of target increased from young adulthood through to elderliness, so did trait attributions of benevolence, norms of politeness and deference, and communicative respect and avoidance; however, attributions of personal vitality as well as communicative satisfaction and enjoyment decreased linearly. These "staircase" profiles were evident among Black African (in Ghana and South Africa) as well as U.S. American students. Hence:

RQ1: Will Indian young adults manifest the same linear patterns regarding age stereotypes, age norms, communicative behaviors to and satisfaction/ enjoyment with younger, middle-aged, and older people as their U.S. American counterparts?

Another aim of our recent work was to determine what factors are predictive of young adults' reported communications with older adults as well as how these, in turn, might contribute to communicative satisfaction with, and enjoyment of, conversations with them. Although a restriction on generalizability given prior studies have used students, we maintained that focus for comparability purposes; the same reasoning underlay our reliance on the same dependent measures as previously adopted. Thus:

RQ2: What are the predictors of communication behaviors for both U.S. Americans and Indians?

RQ3: Will the overall pattern of predictors of communication satisfaction and enjoyment with older people in India be different from those in the USA?

Method

Participants

In total, 232 undergraduate students from two nations participated in this study: 140 (60.3%) from the United States (southern California) and 92 (39.7%) from India (Calcutta). The language of testing was also English in the latter context given that it is the medium of tertiary education. Indian respondents were Hindu whose first language was Bengali. They were drawn from a large urban university which recruits its students from different social, economic, and caste backgrounds as well as both urban and rural areas. The heterogeneity of the students reflects some of the diversity of the Hindu community. Overall, the mean age of participants was 20.09 (SD = 1.57), and relatively equal numbers of males and females participated (47.4% male). Descriptive statistics for each sample and overall are provided in Table 1.

Procedures and Instruments

Using a within-subjects design, questionnaires asked participants to indicate their interaction experiences with nonfamily members or nonclose friends regarding three target ages: young, middle-aged, and older adults. Participants were allowed to self-define these age ranges, and means of these ranges are presented in Table 1. Order of target age was counterbalanced through two versions distributed randomly. In half of the questionnaires, young adults were placed first, followed by middle-age adults, and elderly adults last; this order was reversed in the other half.

For each target age, the questionnaires included measures of communication behaviors (i.e., how they actually behave during interactions with the target age), communication satisfaction with the age group, perceived stereotypes of the age group, and norms of respect (i.e., how they believe they should generally interact with the age group). All items for these scales (as well as their original sources) are provided in McCann et al. (2005). Exploratory factor analyses (i.e., principal components analyses using an oblimin rotation focusing on eigenvalues greater than one) were performed to determine the subdimensions present in the communication behaviors, stereotypes, and norms or respect scales. Separate analyses were conducted for the Indian and U.S. American samples. Kaiser-Meyer-Olkin sampling adequacy measures for all factor analyses were greater than .70 indicating the data were suitable for such analyses. The analyses showed the same factors between nations and across target ages for age stereotypes and norms of respect, which are identical to the factors revealed in other cultures (e.g., Giles et al., 2005). Only minor differences were revealed for communication behaviors. To create the same factors across nations and targets, the items included were those that maximized the reliabilities across all groups. Separate reliability coefficients (Cronbach's [alpha]) for the resulting factors by each target were calculated for each sample (see below).

Communication behaviors

Communication behaviors during interactions with the three target ages were assessed through 10 items. Similar to previous use of this scale (Giles et al., 2005; McCann et al., 2005), the exploratory factor analyses (EFAs) revealed two dimensions: respect and avoidance;, across the nations and target ages, the total variance explained in the EFAs ranged from 46.81% to 59.55%. The respect factor included six items (e.g., "I accommodated to them" and "I showed respect because of their age"), and the avoidance factor included two items (e.g., "I did not know what to say" and "I looked for ways to end the conversation"). Two items were excluded from the factors. Similar to previous uses of this scale (Giles et al., 2005), one item ("I held back my opinions") did not cohere with either factor. The second item ("I did not act like myself") significantly decreased the reliability of the Indian factors for each target age and was thus excluded. Across the age groups, reliability coefficients for both respect ([alpha] = .74 to .86) and avoidance ([alpha] = .67 to .72) were sufficiently high with the exception of U.S. American young adults which had a lower reliability of .48.

Age stereotypes

Participants' stereotypes regarding each age group were assessed through a nine-item semantic differential scale. Similar to previous use of this scale, two factors were revealed by the EFAs (total variance explained ranged from 49.73% to 58.19%): personal vitality ([alpha] = .73 to .82) and benevolence ([alpha] = .63 to .76). The personal vitality factor included six items (e.g., attractive vs. unattractive, strong vs. weak, active vs. inactive), and the benevolence factor included three items (i.e., generous-ungenerous, kind-unkind, and wise-unwise).

Norms of respect

Participants' beliefs about how they should act with each age group were assessed through seven items (Gallois et al., 1999). Similar to previous use of this scale, two dimensions were revealed by the EFAs (total variance explained ranged from 59.64% to 74.32%). The politeness factor refers to norms people feel they should enact during interactions with the age group and included three items (e.g., "I should speak politely to them" and "I should listen patiently to them"). The deference factor refers to a restraint of behaviors during interactions and included four items (e.g., "I should defer to them" and "I should hold back my opinions from them"). Reliability coefficients were sufficient for politeness ([alpha] = .77 to .97) and deference for U.S. Americans ([alpha] = .68 to .78); however, reliability coefficients were lower for Indian deference ([alpha] = .45 to .55).

Communication satisfaction

This was assessed with each age using two items: "I enjoyed my conversation with them" and "I was not satisfied with my conversation with them." The second item was reverse-coded so that greater scores indicated greater satisfaction and enjoyment. Although some of the reliability coefficients were relatively low ([alpha] = .53 to .77), all correlations were significant ([p.sub.s] < .001) and were combined to create one overall score of communication satisfaction for each target age.

Results

Cross-cultural perceptions of age categories were first examined to explore any differences in how cultures viewed when young, middle, and elderly ages begin and end. Separate ANOVAs were conducted on each age category question; all revealed differences (p < .01) across the two nations with the exception of the beginning of middle age (see Table 1 for the means of each nation). Interestingly, U.S. American participants reported lower ages for the beginning and ending of young adulthood, yet they reported the highest ages for the end of middle age and beginning of older adulthood. Thus, the Indian participants viewed young adulthood as lasting longer, middle adulthood as lasting shorter, and elderly adulthood beginning earlier.

Research Question 1

Separate 3 (age target: young, middle-age, elderly adults) by 2 (nation: U.S. American and Indian) by 2 (sex) repeated measures MANOVAs were conducted to identify differences in each of the dependent variables: communication behaviors, communication satisfaction, age stereotypes, and norms of respect. With the exception of satisfaction which had only one factor, both factors of each dependent variables were included in the same MANOVA. Both factors of each dependent variable were included in the same MANOVA. Target age was a within-subjects variable; nation and sex were between-subjects variables. Across the MANOVAs, target age and the interaction between target age and nation were the most substantial predictors of differences in communication perceptions and behavior. As such, the report of the MANOVA analyses focus on these results. [1]

Communication behaviors

The MANOVA for communication behaviors revealed a main effect for target age, [lambda] = .425, F (4, 222) = 74.461, p < .001, [[eta].sup.2] = .575, and an interaction between target age and nation, [lambda] = .786, F (4, 220) = 15.016, iv < .001, [[eta].sup.2] = .214. The univariate analyses for the target age main effect revealed a staircase pattern in which ratings increase as target age increased from young adults to elderly adults for respect (F = 190.415, p < .001, [[eta].sup.2] = .461; Ms = 4.02, 4.78, 5.56, respectively) as well as for avoidance (F = 52.606, p < .001, [[eta].sup.2] = .199; Ms = 2.69, 3.28, 3.61, respectively). The univariate analyses also showed that the interaction between target and nation was significant for respect only (F = 33.615, p < .001, [[eta].sup.2] = .131). Although both U.S. Americans and Indian participants reported greatest respect for elderly adults and lowest respect for young adults, U.S. Americans reported more respect for elderly adults and less respect for young adults as compared to the Indian participants. Thus, the range regarding the staircase pattern for U.S. American participants was wider than the pattern for Indian participants.

Communication satisfaction

The MANOVA for communication satisfaction yielded a significant main effect for target age, [lambda] = .886, F (2, 220) = 14.095, iv < .001, [[eta].sup.2] = .114. No other main effects or interactions were significant. The means for each target age revealed that satisfaction decreased with increasing target age (Ms = 5.36, 5.02, 4.87, respectively). Thus, again, a staircase pattern emerged in the combined samples.

Age stereotypes

The MANOVA examining age stereotypes revealed main effects for target age, [lambda] = .227, F (4, 223) = 189.722, p < .001, [[eta].sup.2] = .773, and a significant interaction effect for target age and nation, [lambda] = .911, F (4, 223) = 5.445, p < .001, [[eta].sup.2] = .089. The univariate effects showed that ratings of both vitality (F = 428.713, p < .001, [[eta].sup.2] = .668) and benevolence (F = 210.198, p < .001, [[eta].sup.2] = .482) varied by target age. Again, staircase patterns emerged; ratings of vitality decreased as target age increased from young adults to elderly adults (Ms = 5.55, 4.76, 3.38, respectively), whereas ratings of benevolence increased as target age increased (Ms = 4.71, 5.00, 5.66, respectively).

Univariate analyses showed significant interactions between nation and target age for both vitality (F = 5.611, p = .009, [[eta].sup.2] = .024) and benevolence (F = 10.943, p < .001, [[eta].sup.2] = .046). For vitality, Indians and U.S. Americans rated young adults and middle-aged adults fairly equally, but the Indians rated elderly adults higher than did U.S. Americans. For benevolence, Indians and U.S. Americans rated elderly adults fairly equally, Indians rated middle-aged adults slightly higher and young adults substantially higher than U.S. Americans.

Norms of respect

The MANOVA regarding norms of respect revealed main effects for target age, [lambda] = .430, F (4, 222) = 73.552, p < .001, [[eta].sup.2] = .570 as well as an interaction effect for target age and nation, [lambda] = .953, F (4, 222) = 2.753, p = .029, [[eta].sup.2] = .047. For the target age main effect, the staircase pattern again emerged for both politeness (F = 201.209, p < .001, [[eta].sup.2] = .472) and deference (F= 150.653, p < .001, [[eta].sup.2] = .401). Participants reported the least politeness with young adults and most with elderly adults (Ms = 4.56, 5.47, 6.18, respectively). Similarly, participants reported the least deference with young adults and the most with elderly adults (Ms = 2.90, 3.79, 4.51, respectively).

Univariate analyses revealed that the interaction between target age and nation was significant for both politeness (F= 3.714, p = .037, [[eta].sup.2] = .016) and deference (F = 8.628, p < .001, [[eta].sup.2] = .037). Both Indians and U.S. Americans reported fairly equal (and low) levels of politeness with young adults. In contrast, although both nations reported greatest levels of politeness with elderly adults, U.S. Americans reported greater politeness with middle-aged and elderly adults than did Indians. For deference, both U.S. Americans and Indians reported similar levels of deference with elderly adults (both reporting highest levels of deference with elderly adults), but Indian participants reported slightly higher deference with middle-aged adults and greater deference with young adults than did U.S. Americans.

Overall, in answering RQ1, Indian young adults reported a similar linear pattern in their stereotypes, norms of respect, communication behaviors, and communication satisfactions with the three target ages. Although small differences emerged between nations, generally these results are similar to reports by other cultures (Giles et al., 2005; McCann et al., 2005).

Research Questions 2 and 3

Because much of the literature focuses on the communication between young and older adults, and to replicate the analyses of McCann et al. (2005) and Giles et al. (2005) with additional cross-cultural data, separate multiple regressions were conducted to determine the predictors of communication behaviors with elderly adults for each nation (RQ2). Although direction of causality cannot be tested with the current data, norms of respect as well as stereotypes should influence communication behaviors during interactions. Additional regressions were also conducted to determine the predictors of communication satisfaction (RQ3). Here, norms or respect, stereotypes, and communication behaviors may play a role in determining participants' satisfaction ratings regarding interactions. See Tables 2 and 3 for a summary of the regression results.

In predicting communication behaviors (RQ2), similarities as well as differences emerged across the two nations. Both politeness and deference significantly predicted respect toward elderly adults in the U.S. American data. Self-reported displays of respect increased as participants' beliefs regarding politeness and deference when interacting with elderly adults increased. In contrast, only beliefs regarding politeness (again positively) predicted respect in the Indian data. In addition, the U.S. American and Indian results were similar in that stereotypes of elderly adults (i.e., personal vitality and benevolence) did not predict respect shown with elderly adults during interactions.

When predicting avoidance, only the stereotype of personal vitality was significant for the U.S. American data; as participants' view of elderly adults increased in vitality, avoidance during interactions with elderly adults decreased. In contrast, several factors predicted avoidance in the Indian data. Both the stereotype regarding personal vitality and norms of politeness were negatively related to avoidance, whereas the respect norm of deference was positively related to avoidance.

When all communication behaviors, stereotypes, and norms of respect were used to predict communication satisfaction with elderly adults (RQ3), only avoidance was significantly related to satisfaction in the U.S. American data--as avoidance increased, satisfaction decreased. In contrast, for the Indian participants, avoidance, personal vitality, benevolence, and politeness all significantly predicted communication satisfaction. Similar to the U.S. American data, greater avoidance during interactions with elderly adults by Indian participants was related to less satisfaction with the interactions. In addition, both of the stereotypes, personal vitality and benevolence, were positively related to satisfaction. In other words, when participants perceive that elderly adults are more healthy and active, they are more satisfied with their interaction with elderly adults. Additionally, politeness norms were negatively related to satisfaction; as participants' beliefs about the need for politeness with older people increased, their satisfaction in communicating with them decreased.

Discussion

The Indian findings very much mirror the U.S. American data in terms of the staircase patterns for stereotypes, norms, communicative behaviors, and communicative satisfaction. In other words, in India as well as elsewhere, as people increasingly age, young adults confer more respect and deference, provide them with such respect, and accord them increasing benevolence. Yet at the same time, the more people age the more they are attributed lowered personal vitality, the less satisfying conversations are with them, and the more they are avoided. Interestingly, and again complementing earlier data in regions of Asia (see Giles et al., 2002), peer-young targets were favored on many of the measures more by Indian than U.S. American respondents. However, and interestingly in contrast to prior stereotype data (Williams, Pandey, Best, Morton, & Pande, 1992), Indians rated their elderly targets as possessing more personal vitality and benevolence than did U.S. Americans. That said, this was paralleled by the latter claiming they communicated more respect for elders than their Indian counterparts.

These findings, as before (e.g., Giles et al., 2005), emerged largely irrespective of respondents' sex (McCann et al., 2003); a finding that is perhaps somewhat surprising given the large-scale social discrepancies existing in India historically between males and females across the lifespan (e.g., Jejeebhoy, 2002). For instance, mother-in-law typically hands over the responsibilities of managing the household to the daughter-in-law and "graduates" into old age. While the father passes on responsibilities to his son when he marries, this is more of a nominal exercise than a real transfer of power, underscoring the complicated ways in which aging is gendered. In future work, sex of target rated could be profitably manipulated as well as perhaps other physical and social attributes of the target, such as family relationship.

It is also imperative to have middle-aged and elderly respondents join the empirical fray. If we are, ultimately, to move towards implementing robust intervention programs to tackle communication management problems pertaining to intergenerational relations, then obtaining data from the latter groups is vital to informing the design of an approach that involves all generations in these issues. Accordingly and obviously, we are exceedingly reticent about generalizing to the complexity of inherent heterogeneous mosaic that represents multilingual India (Khilnani, 1998), and the various ethnic, religious, and caste structures within this country, many of whose speakers are illiterate in English--clearly a rich arena for further empirical exploration. Indeed, it is possible that growing older is less salient for the upper caste which might not have to rely as heavily on their own physical labor as the lower castes. In this vein, survey language can be manipulated as different varieties can be meaningfully relevant to triggering some distinct value structures for the same bilingual person. Subsequently, we need to further explore the relevance, as well as cultural equivalence, of our social and communicative dependent measures for the Indian subcontinent. Indeed here, the process of aging has unique culture meanings that do not correspond with chronological age, which is appealing for empirical exploration (Cohen, 1992). For instance, the Hindu lifespan is divided into four, functionally different and idealistic stages (Bhat & Druvarajan, 2001; Tilak, 1989): Brahmachrya (student life with sexual abstinence); Grihastha (married with righteous living); Vanapratsha (retired life with religious reflections); and Sanyasa (living a spiritual life based on these prior foundations). Furthermore, we know little about the exact impact of castes on how aging is framed in India (Qigley, 1993), and the ways in which it may differ between castes.

The answer to RQ2 was affirmative in that the largest predictor of communication satisfaction with elderly people was, once again, communicative avoidance. However, and in answer to RQ3, the profile of predictors was different to the extent that three other factors in addition to communicative avoidance emerged as significant. Among these was the age stereotype of personal vitality similar to the African sample; this time (like the Ghanaians) the link was positive but so, too, was the other stereotype factor, benevolence. In other words, age stereotypes in the Indian cultural context directly predicted communication satisfaction with older people (whereas personal vitality's role was only in predicting communicative avoidance in the U.S. American sample). Paradoxically then, this does (more than U.S. American data) support Western-based models of intergenerational communication which afford age stereotypes a crucial mediating role in determining communicative outcomes (e.g., Hummert, Garstka, Ryan, & Bonnesen, 2004; Ryan et al., 1986). Also interestingly, was that the norm of politeness was negatively predictive of communication satisfaction: the less young people felt pressured into being polite with their elders, the more satisfaction they reported in communicating with them. Again, this is a potent rider to the traditional role of age norms in shaping communicative outcomes (Barker et al., 2004). Why it was that more factors predict communication in India than the USA is not at all clear and worthy of empirical pursuance.

These findings should be interpreted in light of some limitations. First, a few of the factors yielded low reliabilities, specifically U.S. American young adults' avoidance, and Indian deference for all three target ages. The results regarding these factors should be interpreted with caution until additional research can substantiate these findings and provide more reliable and valid factors. Additionally, the communication satisfaction factor was assessed using only two items. Although consistent with previous research (Yeh, Williams, & Maruyama, 1998), validated scales of communication satisfaction should be utilized in future research to enable stronger conclusions regarding predictors of communication satisfaction. Further, although perceptions of actual communication behaviors were assessed, methodologies employing observational analyses should be incorporated to verify participants' accounts of the behaviors enacted during interactions with each of the target ages. In addition, we need to move beyond student samples. Finally, given the differences in self-defined age ranges of young, middle, and older adulthood for the two nations, future research would also benefit by examining how these differences may affect ratings of stereotypes, norms of respect, and other factors assessing regarding intergenerational communication.

Despite these constraints, we have provided unique empirical data with an understudied population in intergenerational communication research. It is important to continue the kind of work described herein as most older people in India are cared for by young adults and, in this sense, intergenerational communication is an integral component of care.

References

Barker, V., Giles, H., & Harwood, J. (2004). Inter- and intragroup perspectives on intergenerational communication. In I. F. Nussbaum & J. Coupland (Eds.), Handbook of communication and aging research (2nd ed., pp. 139-166). Mahwah, NJ: Erlbaum.

Bhat, A. K., & Dhruvarajan, R. (2001). Ageing in India: Drifting intergenerational relations, challenges, and options. Ageing and Society, 21, 621-640.

Cohen, L. (1992). No aging in India: The issue of gerontology. Culture, Medicine, and Psychiatry, 16, 123-161.

Coupland, N., Coupland, J., & Giles, H. (1991). Language, society, and the elderly: Discourse, identity and aging. Oxford, UK: Blackwell.

Fonseca, M. (1999). Role of family and aging in India. In S. D. Gokhale, P. V. Ramamurthi, N. Pandit, & B. S. Pendse (Eds.), Aging in India (pp. 161-167). Mumbai, India: Somiya Publications.

Gallois, C., Giles, H., Ota, H., Pierson, H. D., Ng, S. H., Lim, T.-S., et al. (1999). Intergenerational communication across the Pacific Rim: The impact of filial piety. In J.-C. Lasry, J. Adair, & K. Dion (Eds.), Latest contributions to cross-cultural psychology (pp. 192-211). Lisse, The Netherlands: Swets & Zeitlinger.

Gallois, C., Ogay, T., & Giles, H. (2005). Communication accommodation theory: A look back and a look ahead. In W. Gudykunst (Ed.), Theorizing about intercultural communication (pp. 121-148). Thousand Oaks, CA: Sage.

Giles, H., Makoni, S., & Dailey, R. M. (2005). Intergenerational communication beliefs across the lifespan: Data from Ghana and South Africa. Journal of Cross-Cultural Gerontology, 20, 191-211.

Giles, H., McCann, R. M., Ota, H., & Noels, K. A. (2002). Challenging intergenerational stereotypes: Across eastern and western cultures. In M. S. Kaplan, N. Z. Henkin, & A. T. Kusano (Eds.), Linking lifetimes: A global view of intergenerational exchange (pp. 13-28). Honolulu, HI: University Press of America, Inc.

Giles, H., & Williams, A. (1994). Patronizing the young: Forms and evaluations. International Journal of Aging and Human Development, 39, 33-53.

Harwood, J., & Giles, H. (1993). Creating intergenerational distance: Language, communication and middle-age. Language Sciences, 15, 1-24.

Helson, R., & Soto, C. J. (2005). Up and down in middle age: Monotonic and nonmonotonic changes in roles, status, and personality. Journal of Personality and Social Psychology, 89, 194-204.

Hummert, M. L. (1994). Stereotypes of the elderly and patronizing speech. In M. L. Hummert, J. M. Wiemann, & J. F. Nussbaum (Eds.), Interpersonal communication in older adulthood (pp. 162-184). Thousand Oaks, CA: Sage.

Hummert, M. L., Garstka, T. A., Ryan, E. B., & Bonnesen, J. L. (2004). The role of age stereotypes in interpersonal communication. In J. F. Nussbaum & J. Coupland (Eds.), Handbook of communication and aging research (2nd ed., pp. 91-114). Mahwah, NJ: Erlbaum.

Jejeebhoy, S. (2002). Convergence and divergence in spouses' perspectives on women's autonomy in rural India. Studies in Family and Planning, 33, 299-308.

Joshi, K., Kumar, R., & Avasthi, A. (2003). Morbidity profile and its relationship with disability and psychological distress among elderly people in Northern India. International Journal of Epidemiology, 32, 978-987.

Khilnani, S. (1998). The idea of India. New Delhi, India: Penguin.

Liebig, P. S. (Ed.). (2003). An aging India: Perspectives, prospects, and policies. New York: Haworth.

McCann, R. M., Dailey, R. M., Giles, H., & Ota, H. (2005). Beliefs about intergenerational communication across the lifespan: Middle age and the roles of age stereotyping and filial piety. Communication Studies, 56, 293-311.

McCann, R. M., Ota, H., Giles, H., & Caraker, R. (2003). Perceptions of intra- and intergenerational communication among adults in Thailand, Japan, and the U.S.A. Communication Reports, 16, 1-23.

Nayar, V. K. (2000). Low intensity conflicts in India. New Delhi, India: United Service Institution of India.

Neki, J. (1987). Psychological stressors in ageing and old age in various subcultures in India. In L. Levi (Ed.), Society, stress, and disease, volume 5: Old age (pp. 85-93). New York: Oxford University Press.

Ota, H., Giles, H., & Somera, L. B. (2007). Culture and communication across age groups: People's perceptions of intra- and intergenerational communication in Japan, the Philippines, and the United States. Communication Studies, 58, 173-188.

Patel, V., & Prince, M. (2001). Ageing and mental health in a developing country: Who cares? Qualitative studies from Goa, India. Psychology of Medicine, 31, 29-38.

Qigley, D. (1993). The interpretation of caste. New York: Clarendon Press.

Ranjan, S. T., Sarma, P. S., & Mishra, U. S. (2003). Demography of Indian aging 2001-2051. Journal of Aging and Social Policy, 15, 11-30.

Ryan, E. B., Giles, H., Bartolucci, G., & Henwood, K. (1986). Psycholinguistic and social psychological components of communication by and with the elderly. Language and Communication, 6, 1-24.

Ryan, E. B., Hummert, M. L., & Boich, L., (1995). Communication predicaments of aging: Patronizing behavior toward older adults. Journal of Language and Social Psychology, 13, 144-166.

Tilak, S. (1989). Religion and aging in the Indian tradition. Delhi, India: Sri Satguru Publications.

Williams, A., & Giles, H. (1996). Retrospecting intergenerational conversations: The perspective of young adults. Human Communication Research, 23, 220-250.

Williams, A., & Nussbaum, J. F. (2001). Intergenerational communication across the lifespan. Mahwah, NJ: Erlbaum.

Williams, J. E., Pandey, J., Best, D. L., Morton, K. R., & Pande, N. (1992). Young adults' view of old adults in India and the United States. In S. Iwawaki, Y. Kashima, & K. Leung (Eds.), Innovations in cross-cultural psychology (pp. 227-234). Amsterdam: Swets & Zeitlinger.

Yeh, K.-H., Williams, A., & Maruyama, M. (1998). Approving and disapproving grandmothers and strangers: Young Taiwanese and American comparisons. Journal of Asian Pacific Communication, 8, 125-149.

Notes

[1] Though predicting substantially less variance than target age, the MANOVAs revealed main effects for nation for stereotypes, [lambda] = .914, F (2, 225) = 10.598, p < .001, [[eta].sup.2] = .086, and norms of respect, [lambda] = .968, F (2, 224) = 3.659, p = .027, [[eta].sup.2] = .032. The univariate analyses for stereotypes revealed significant differences for both vitality, F= 8.904, p = .003, [[eta].sup.2] = .038, and benevolence, F = 20.500, p < .001, [[eta].sup.2] = .083. Indian participants reported greater vitality (M = 4.70) than U.S. Americans (M = 4.47) as well as greater benevolence (M = 5.18) than U.S. Americans (M = 4.81). The univariate analyses for norms of respect showed that ratings of deference only varied significantly by nation, F = 4.170, p = .042, [[eta].sup.2] = .018. U.S. American participants (M = 3.64) reported lower deference than did Indian participants (M = 3.90). Sex differences were found only for age stereotypes, [lambda] = .969, F (2, 225) = 3.558, p < .03, [[eta].sup.2] = .031. The univariate analyses showed that only benevolence significantly varied by sex, F = 7.119, p = .008, [[eta].sup.2] = .031; females reported greater benevolence (M = 5.11) than did males (M= 4.89).

Howard Giles (PhD, University of Bristol, 1971) is a Professor in the Department of Communication at the University of California, Santa Barbara. Rene M. Dailey (PhD, University of California, Santa Barbara, 2005) is an Assistant Professor in the Communication Studies Department at the University of Texas, Austin. Jayashree M. Sarkar (PhD, Central Institute of English and Foreign Languages, 1994) is an Associate Professor in the State English Language Teaching Institute, Calcutta. Sinfree Makoni (PhD, University of Edinburgh) is an Associate Professor in the Department of Applied Linguistics and African and African American Studies at the Pennsylvania State University, College of Liberal Arts. We are grateful to Alicia Mandac and Kimberly Khoang for their assistance, three anonymous reviewers for their valuable comments, and the Editor for his comprehensive feedback. Correspondence to: Howard Giles, Department of Communication, University of California, Santa Barbara, Ellison Hall, Santa Barbara, CA 93106-4020, USA. E-mail: HowieGiles@cox.net

Table 1 Descriptive Statistics of Sample

                   Overall        U.S. American      Indian

N                  232            140                92
% Male             47.40%         50.00%             43.50%
Age range          16-28          16-28              18-25
Age mean (SD)      20.09 (1.57)   19.98 (1.56)       20.27 (1.57)
Young adulthood    16.95 (2.14)   16.60 (2.26) (a)   17.55 (1.78) (b)
    begins:
    ([dagger])
  Range            10-21          10-21              13-21
Young adulthood    29.93 (6.53)   28.24 (5.57) (a)   32.75 (7.06) (b)
  ends:
  Range            18-50          18-44              24-50
Middle age         32.92 (5.87)   32.67 (5.82)       33.33 (5.96)
    begins:
  Range            21-50          21-50              24-50
Middle age ends:   54.11 (7.88)   55.85 (7.85) (a)   51.22 (7.09) (b)
  Range            34-80          35-80              34-65
Elderly age        57.80 (8.14)   60.61 (7.73) (a)   53.13 (6.53) (b)
    begins:
  Range            35-81          36-81              35-66

Note. ([dagger]) Nine Indian participants did not provide
estimates of age ranges; additionally, two Indian and four
U.S. American participants' reports were excluded from the
ranges and means as their reports deviated more than three
standard deviations from the overall mean. Superscripts (a)
and (b) indicate significant differences between nations.

Table 2 Regression Results for U.S. American Data

                   [beta]     F           df         [R.sup.2]

Communication
    Behaviors

Respect                       27.251 **   (4, 135)   .43
  Vitality         -.029
  Benevolence      -.023
  Politeness        .591 **
  Deference         .157 *

Avoid                         5.203 **    (4, 135)   .11
  Vitality         -.283 **
  Benevolence      -.120
  Politeness        .052
  Deference         .167

Communication
    Satisfaction              8.932 **    (6, 133)   .26

  Respect          -.020
  Avoid            -.466 **
  Vitality          .096
  Benevolence       .079
  Politeness        .166
  Deference        -.033

* p < .05. ** p < .001.

Table 3 Regression Results for Indian Data

                             [beta]        F         df     [R.sup.2]

Communication Behaviors

Respect                                 49.470 **   (4,81)      .70
  Vitality                    .005
  Benevolence                -.024
  Politeness                  .787 **
  Deference                   .123
Avoid                                   7.803 **    (4,81)      .24
  Vitality                  - .220 *
  Benevolence                 .121
  Politeness                - .480 **
  Deference                   .358 **

Communication Satisfaction              14.760 **   (6,78)      .50

  Respect                     .371
  Avoid                     - .656 **
  Vitality                    .218 *
  Benevolence                 .186 *
  Politeness                - .360 *
  Deference                  -.054

* p < .05. ** p < .001.