Twelve older adult informants participated in the study. This is a very small sample in respect of the older adult population in Mumbai. However, given the importance of cultural context for understanding conceptualizations of PSOC [20] [39] , a in-depth and context-sensitive analysis was prioritized over sample size.

To recruit older adult informants, scientific staff at the Tata Institute of Social Science (TISS) invited older adults by phone. The following admission criteria were used; above 60 years of age, speak English and living in Mumbai. A combination of purposive and convenience sampling was used in the study. Ensuring a sample that could inform us about PSOC in old age from different perspectives, we included members of the Lions Club in Mumbai (three men, two women) and older adults who were not active in social or political groups (three women). Finally, older adults living in institutions (three men; two of them active as volunteers, one woman) were included. Two day-care centres were chosen as an institutional setting suitable for recruiting older adults sufficiently healthy to participate in the study. All the informants were acquaintances of one of the involved staff members from TISS. Furthermore, the following demographic variables characterized the sample (Table 1). All informants reported to be Hindu.

Qualitative interviews have been a useful method for getting information about

¹Informants are represented with codes indicating gender (F, female and M, male) and the informant’s age.

PSOC in non-Western cultural contexts [13] . Aiming for data that provided a detailed understanding of the samples meaning of PSOC we chose semi-struc- tured interviews, making sure that the informants had a chance to express their understanding and at the same time have some overall structure to the conversation [40] . The interview guide consisted of two sections; part one contained questions about demographical variables (age, gender, marital status, highest level of education, residential area of Mumbai, years of residence, number of children, and number of family members residing in their home). Part two included open-ended questions about the meaning of community and PSOC (e.g. “What comes to your mind when you think of the word ‘community’?”, “What meaning do you relate to the concept ‘sense of community’?” and “In which way would you say a sense of community is important to you as an elderly person?”).

The interviews were performed in English. Two research assistants were present during the interviews to make sure language did not become a challenge. The interviews were recorded using a digital voice recorder and ranged from 13 minutes to 1 hour and 36 minutes. The study was approved by national ethical guidelines (Norwegian Centre for Research Data) and TISS.

Thematic Analysis. In psychology thematic approaches are a common way of analyzing data from interviews. Braun and Clarke (2006) provide an accessible, systematic and rigorous approach to identify and analyze frequent and central themes as part of peoples meanings [41] [42] [43] . This was essential for the aim of the study, and the approach was used as the initial analysis of the data set, consisting of verbatim transcriptions of the second section of the interviews. As will be further discussed, we decided to combine the thematic analysis with discourse analysis in order to explore the samples meanings of PSOC in depth and in a context-sensitive way.

The first four steps of the thematic analysis were performed to capture the most frequently mentioned and common themes. First, the researcher (corresponding author) read through the transcripts to become familiar with the material and generate initial codes. To make sure codes were not missed out in the first read through, the researcher went back to the data set to search for additional codes. To develop themes the most central characteristics were noted for each candidate theme and sub-theme. Then, the researcher reviewed themes by going through the data-set to see if and how each theme interacted with other themes and demographical variables. Finally, the review was extended by highlighting themes in the data set with different colors, checking if the parts that were not colored provided further insight to the samples meanings of PSOC. In this process ten themes were identified. They were: “Helping and caring for others”, “Adaptation and acceptance”, “Reciprocity”, “Safety and welfare structures”, “Shared elements”, “Sense of belonging and attachment”, “Community satisfaction”, “Respect”, “Residential investment and action” and “Heterogeneity”.

Discourse Analysis. An a priori assumption in discourse analysis is that meanings of concepts change with time. This process happens through discourse, the interaction between spoken language (the way individuals and groups speak about something), and culture as part of meaning systems prescribing and describing the common sense of society [44] [45] [46] [47] . Discourse analysis provides several approaches and tools [45] [48] . Gee [49] offers an approach to analyze meaning systems in a data set and is considered appropriate for our study. He argues that what is taken for granted or common sense is most essential for understanding meaning. As one the tools within his approach ‘The making strange tool’ focuses on the taken for granted elements in the data set. The researcher enquires into what the informants implicitly take as natural and true. Applying the tool it became obvious that some words used by the participants reoccurred in the data set. These words were used to search by in the data set, as a way of concretizing the use of the tool. Eight words were typically used when talking about taken for granted notions: “Should/Shouldn’t”, “Must”, “Have to”, “Absolutely”, “Normal/Normally”, “Isn’t/That is not”, “Right” and finally “Of course”. Searches were made for these words and sections from the data set including the words were extracted and put into a separate document to analyze these with the discourse analysis tool. Four recurring taken for granted aspects of PSOC were identified: “Community development”, “Owing the community”, “Community membership”, and “Social influence”. Using the tool to review the themes gave an in-depth understanding of what was taken for granted, and provided the possibility of revealing additional themes. The findings from this analysis were evaluated and then merged with the findings from the thematic analysis, making sure that the most central and common themes as well as the taken for granted dimensions of PSOC in the data set were included in the final themes.

Context is a key concept in discourse analysis [45] [46] [48] . To understand the role of context in the samples meaning of PSOC we chose to use Gee’s “Framing problem tool”. This tool can be used to analyze meaning in context by continually asking and analyzing how contextual aspects affect the meaning of peoples’ utterances. Any aspect of context can affect the meaning of an utterance and the researcher uses his/her knowledge of the context to see if aspects of the context are relevant [49] . This tool was applied in the analysis of the core themes in the informants’ meaning of PSOC and community (presented below) by asking how any of the related utterances from the sample reflected the Indian context. This way of thinking gave a more context-sensitive understanding by connecting the samples meaning with the larger meaning-systems in their cultural context.

It is almost expected that people speak about help and care when asked about their meaning of PSOC. In our material it was the most evident theme and it operated as a “taken for granted” part of the informants meaning. Words and expressions like “benefit”, “work” and “doing something” were typically used. For several informants it was part of the very meaning of PSOC:

M65: “Sense of community is the caring of others.” M81: “…sense of community there first of all needs to be an objective. And the objective should be common for all. And most of the objectives should be positive, that is in a helping type”.

However, while Western conceptualizations of PSOC typically describe help and care between somewhat equal individuals within a community (e.g. at work) the informants described it as something they provided for less fortunate individuals:

F82: “… our senior citizens used to go there and all the patients used to come, those patients where illiterate, they don’t know where to go, so we used to help them…volunteering.” F61: “… we go to the slums… and help with education of the uneducated people…”

Further on it was explained that helping and caring was a way to “give back” to the community:

F82: “… from the community we came up to this level, so we must give it back to the community. That is what I feel is sense of community means.”

In this type of community service, some demographical variables seemed to matter. First of all educational background; older adults who identified themselves as “highly educated” (four of the informants) expected themselves and other educated people to make an effort in the community:

M65: “All the educated should have come together and they should try to impart whatever they know… The behavior should be good. Social behavior like helping each other, so because helping others is not an obligation, it is a duty.”

Secondly, (widowed) older adults who lived alone seemed to find an especially important purpose and well-being in giving help to the larger community:

M81: “I need them [the family] ... but I don’t want… this is better. I am not God himself, but he has given me some power to do something, impact something. So, let me impact.” F61: “If you are in a community you feel ‘I am more safe than being outside the community’, then you will feel alone…. A sense of safety, a sense of belonging, and sense of wellbeing, mentally and physically. That I am in a good company, and I can also contribute…”

There are several elements within the context, which can help us to understand this theme in a context-sensitive way. First, concept “seva”; “service to others” [50] , often understood as a selfless act of worship for a Guru or a social cause [51] [52] . Just as the community efforts the informants described, seva is frequently practiced by volunteering in NGOs [53] . Secondly, according to Hinduism and the caste system there are duties which Hindu individuals, as our informants were, are expected to perform are prescribed by caste [31] . The older adult informant who identified as ‘highly educated’ most likely belong to the highest caste of “Brahmin”. In urban areas of India, such as Mumbai, religious practice (including social duties) has been demonstrated as of greater importance for “Brahmins” compared to lower casts [54] . Finally, Hindu philosophy prescribes older adult individuals to withdraw from the material world and the social bonds, such as the family and community life [31] [32] [33] . Then, the finding of community investment as an important and salient part of the “educated” older adults meaning (particularly for those who living alone), seem to be consistent with prior findings that the main motivation for doing seva in old age is a sense of belonging [51] , but actually contradicts central age related traditional expectations within their context.

Established definitions of PSOC are founded upon a membership-similarity understanding of community. In order for PSOC to grow community members need to be alike. The sample from Mumbai, however, shared an understanding of community as consisting of members who were dissimilar. Typically, words like “different”, “cosmopolitan” and “mix” were used. Differences in regional religious affiliation, age, interest, professional background and group-identities were described as important in a community:

M67: “I believe that those who stay together, maybe of different regions, maybe of different religions…. But… they should continue to stay together peacefully… that is what I feel a community means.” F61: “A community… is people… not only of the same age, different age groups. Different groups of people having different professions, different interests, but they are staying in the same area.” M65: “Community means that it is a group of people belonging to different race, different caste…but all are living happy. There is no distinction as such.”

To adapt to and accept community members’ differences was something the informants identified with as “Mumbaikars”:

F67: “… we [from Mumbai] are none of us in a place of origin. We do have a certain culture, a basic culture at home, but we are very adaptable to the other cultures of my neighbours… Here people have… that sense of acceptance, and that is something very special about this Mumbai…”

To understand the informant meaning of community in a context-sensitive way we have to look Mumbai as a context. Mumbai can be described as a “Mini- India” where multiple diverse Indian cultures co-exist. The city is also the most exposed context to Western meaning-systems within the nation [38] . As such the citizens, especially those who have resided there for a long time (as many older adults are likely to) are used to a great heterogeneity of meaning systems. In addition it is likely that this form of heterogeneity attracts residents preferring it, and that those who do not favour it have to adapt to it in order be accepted, strengthening diversity as part of the culture further.

As presented, personally investing in the well-being of the larger city community was a very important part of the Indian older adults meaning system of PSOC. This form of social responsibility has been acknowledged as a central part of PSOC, particularly, in the element “personal investment” included in the “Membership” dimension [55] [56] [57] [58] [59] . In Nowell and Boyd’s recent corollary conceptualization of PSOC [58] the concept is understood as “sense of community responsibility” “… the feeling of personal responsibility for the individual and collective well-being of a community of people, not directly rooted in expectations of personal gain” [59] . The way our sample described PSOC―as helping and caring for others in a way benefiting the larger community-largely resemble PSOC-R by definition. Moreover it mirrors values of collectivism, selflessness and community investment embedded in the Indian cultural context [50] [51] [60] . Then, the informant’s meaning exemplify that PSOC-R can be a context-sensitive concept, reflecting key aspects of culture. Furthermore, as our findings suggest, with changes in the family structure (constructing old age outside the family as the norm) and the traditional culture (increasing individualism), this form of responsibility may be especially important for urban older adults PSOC today, despite traditional prescriptions of old age as a withdrawn from social bonds.

One of the primary values of community psychology has been to build positive communities and at the same time recognize the centrality of diversity. In the field today there is a ongoing discussion about this community-diversity dialectic; One side using agent-based simulation models to explore the dialectic argue that the contextual conditions that foster respect for diversity run in opposition to the ones that foster PSOC [61] [62] [63] . On the other side Brodsky [64] and Hill [65] state that there is no inherent dialectical relationship or conflict inherent to the concepts. Our informants’ discourse of community was founded upon diversity as a positive and required aspect, enabling inclusion, acceptance and co-existence of the many meaning systems within the urban context. As such the older adults concept provides an example from the “real world” illustrating community and diversity as interrelated and highly contextual concepts [64] , which should be understood from within peoples multiple identities, communities and meaning systems.