Situation and Social Network in Cities

Situation and Social Network in Cities*

N O E L J. C H R I S M A N / PomonaCoZZege

Un problkme signillcatif dans l'klaboration et l'utilisation de r6seaux sociaux en recherche urbaine se rapporte B la sp6cification du contenu des relations de r6seau. Dans une 6tude sur le terrain chez les Danois am6ricains de la Cali- fornie, certains facteurs furent de grande utilit6 dans la d6finition des liens de r6seau. Ce sont : le cadre institutionnel B l'inthrieur duquel s'6laborent les liens de r6seau; les situations qui leur permettent d'ctre op6rants et la nature atTec- tive de ces liens de rkseau. En plus de leur valeur optiratoire, ces facteurs permettent au chercheur d'ivaluer la qualit6 et la diversit6 de la vie urbaine. A significant problem in the construction and use of social networks in urban research is that of specifying the content of network relationships. In fieldwork among Danish-Americans in California, the institutional framework in which network ties were recruited, the situations in which such ties were activated, and the af€ective nature of network links were useful in defining network bonds. In addition to their operational utility, these factors allow the investigator to assess the quality and diversity of urban life.

In order for network analysis to be more than an exciting potentiality in social research, it is necessary that there be agreement on what social networks are (cf. Mayer, 1966) and that there be means to describe them. Previous studies have provided us with some aspects of networks which must be included in a d e ~ t i o n . For example, Barnes discusses the fact that individuals who are related to one another by ties of friendship and acquaintance may be grouped together as a network. He stresses that this is a residue left behind when other principles of grouping in the society (territory and occupation) are removed ( 1954). Elizabeth Bott contributes the notion of connectedness; that individuals in networks may or may not know each other independently of a central ego (1957). The location of boundaries in connection with network definition continually plagues us (cf. Barnes, 1968; Mayer, 1966). Regularity of interaction and closeness of ties are characteristics considered by Epstein in his discussion of effective and extended network (1961). Each field worker had contributed the characteristics of networks which aided him in his own research. Symposia such as the present one in which contributors may pool their own recent thoughts are helpful in refining the notion of network and contributing new dimensions to be considered in network analyses. In this discussion, I wish to present three factors I found useful in my own fieldwork.

* An earlier version of this paper was given at the meetings of the Central States

Rev. canad. Soc. & AnthJCanad. Rev. Soc. & Anth. 7(4) 1970

Anthropological Association in Milwaukee, Wisconsin, May 1969. 245

A difiiculty in network analysis as a methodological approach at the present time has to do with the relation between the structure and the content of the component relations. For example, “close-knit” seems to refer to a structural property of interconnectedness; “effective” appears to describe aspects of both structure and content since the individuals included are seen frequently (a structural property) and opinions carry weight with the central ego (an aspect of content). The characteristics of structure and content must be analytically separated to the greatest possible extent. (This is, of course, difiicult since they are observably the same thing - interaction patterns). Barnes (1968) has made a major contribu- tion to a consideration of the structural attributes of networks. Mitchell’s (1969) further specification of these characteristics is useful as well. How- ever, approaches to understanding and specifying the content of component links continue to be problematic.

In research1 done among Danish-Americans in “Pacific City,” California (Chrisman, 1966), I used a series of factors at an operational level to attempt an Understanding of the nature of the interpersonal links. I deter- mined the way in which an interpersonal relationship had been recruited, the situations in which such relations were activated, and made a subjective judgment as to the afEective nature of the bond. Data on recruiting included the original context of a bond formation as well as subsequent institutional bases of interaction. For example, most of the Danes I inter- viewed were members of at least one Danish voluntary association and had formed personal links in that context. A second characteristic I used was situation. That is, I attempted to ascertain the number and types of situations in which network members interacted. For example, lodge members interacted at lodge meetings, special events, sports activities, and visits to each others’ homes. The two factors of recruitment and situation give an indication of the content of the link in that they are related to the interests and other personal characteristics of the people involved.

The judgment about the affective nature of the network link is, to some degree, based upon the above factors of recruitment and situation, That is, those individuals who regularly activated the roles from which their personal relationships were recruited and who frequently interacted in a series of different situations were often found to have a higher degree of interpersonal commitment than persons who saw each other less regularly. It is clear that frequency and regularity of interaction are not the only indi- cators of the commitment and durability of a link since many Americans maintain close ties (through the mail, telephone, etc.) with individuals with whom they seldom interact personally. For this reason, I also made subjective judgments from interview data. A structural characteristic which seems to be related to the personal nature of the social bonds is the goal

1 This research was supported in part by a pre-doctoral fellowship from the Be- havioral Sciences Project of the School of Public Health, University of California, Berkeley. (NIMH Grant 5T1 MH-8104.) 246

orientation of the role link; i.e., its instrumental or expressive nature (Par- sons, 1951).

A goal in the development of network analysis as a new methodological tool has been to generate something that rivals kinship diagrams in shpli- city and ease of description. We long for our own “lineage” and the se- curity which accompanies a description of a kinship system. One knows that the notions of patrilineage, bilateral kindred, and the like will be understood by others and that they are reasonably descriptive of some reality. Since the results of the operationalization of these terms may be represented on paper in the form of charts, this is an objective for network analysts as well. A problem in charting a social network, however, is that the meaning of the connecting lines is not so clear as in kin charts. We can merely note the degree of connectedness of such lines. As well, individual roles in the mesh of network relations do not have the explicit, comparable meaning of father’s mother’s brother, for example. One reason for this is that kin roles are dependent, using Nadel’s terminology. That is, “a given role is so conceived that, by its character, it requires to be enacted vis4- vis another counterpart or correlative role” (Nadel, 1957:79- 80). This is not necessarily the case in the relationships included in a social network. Since social ties in a personal set may be drawn from a wide variety of structural arenas, the actual content of these roles is variable; we know only that a relationship of some kind exists.

Before we generate little but a serious inferiority complex, it might be useful to gather some h i t s from kinship studies about directions in which we might go. What is the nature of the societies in which kinship is most commonly used for social analysis? Generally, they are the non-Western groups which anthropologists are so fond of studying. Characteristically, roles in these societies overlap to a great degree. Thus, father is also culti- vator, lawyer, priest, and hunter. Kinship is useful for analysis because it openly identifies one of the roles. Then, because of the nature of the societies we examine, not because of the tool, we may investigate other aspects of life. If descent groups did not also act as major political, economic, and ritual units, we would not have such a versatile tool in kinship. Network analyses have been attempted where traditional kinship approaches would not have been as productive. It is perhaps in vain that we expect network charts to tell us as much at a glance as do kin charts. It is not, however, unreasonable to expect that the roles and linkages of networks can be described fully and accurately.

RECRUITMENT AND INSTITUTIONAL ALTERNATIVES

Rather than beginning with the chart, or over-all aspect of network, it might be worthwhile to examine the component social relationships. Speci- fying their nature is useful since the structure of the network may vary de- pending upon the content of the roles within it. For example, Bott links 247

close-knit networks to an overlapping of social roles - one individual filling a number of roles at the same time (Bott, 1957:66). There are a number of ways the component role relationships of a network may be described - kin, neighbour, friend, and the like. Operationally, we can consider these to be examples of how the relationships were recruited. For example, Bott and others state that a network may consist in part of kin ties. The kinship system, as an institutional framework, insofar as it has provided for certain individuals to have a role relationship, has supplied the opportunity for those individuals to know each other and to interact.

The many institutional frameworks of a society are the structural features in the social system which can influence the formation and persistence of groups and which can also provide the arena within which network relations may form. People who share a common institutional membership have at least the possibility of minimum and structured interaction. For example, there were about 150 members of the Pacific City branch of the Danish Brotherhood. Each had been initiated (and learned the secret lore of the association) and had continued to pay dues. Any one of these men could, therefore, go to a Friday night meeting, enjoy the Danish atmos- phere and fellowship, participate in discussions, and vote. They were all eligible for office in the organization. That only about twenty-five men usually attended the meetings does not decrease the importance of the voluntary association as a potential arena for interaction. Should a man wish to activate his associational role for some purpose, he would be able to do so.

Urban dwellers belong to a series of institutions which may be activated with greater or lesser frequency. The Danish-Americans were also members of institutions in the areas of occupation, residence, kinship, religion, and the like. Association membership was only one aspect of a life composed of participation in a series of diverse institutions which might be spatially and socially unrelated. Within each institution which is activated, the individual has the choice of the depth and manner with which he interacts with others around him. He may invest much in the relationship so that rights and obligations with respect to a role-partner transcend the requirements of the formal relationship - he may, in other words, form a dyadic contract (Foster, 1961) with another person. The individual might interact with others in an institution only to the extent of the roles provided by institutional structure; or as Foster points out, he may not honour role obligations at all. In addition, individuals who share membership in the same institution and know the role obligations may still not interact because of lack of physical opportunity. For example, machinists on alter- nate shifts, in separate areas of a factory, or in Merent towns can share close institutional ties, but not be able to activate them. Danish-Americans who belong to difEerent chapters of the same lodge share knowledge of common role behaviours, but unless one physically visits another chapter, no interaction takes place and the potential role relationship is left un-

248 activated.

Individuals who share membership within an institutional framework fulfill one of the preconditions for being recruited into a personal network, but do not conform to the interactive component of network relations. Katz (1966) refers to these possibilities for networks as ‘‘potential networks.” He notes, however, that “in an ‘open’ society all members are potential contacts of ego, at least in principle” (1966:204). This most inclusive notion of network - that which includes individuals in relationships which may never connect in an interaction situation - is too broad to be a unit of research and analysis in urban areas. In addition to the breadth of the problem, there is also the diflicdty of differentiating between those potential links which are the result of shared institutional membership and those which may be drawn from personal relationships. That is, one must distinguish between role sets and personal sets (Wolfe, 1969). In role sets, individuals may form a relationship while engaging in a common institutional task; whereas a personal set may be drawn from many institutional contexts or from interpersonal interaction outside of large-scale institutions with set tasks.

Recruitment as one of the mechanisms in network formation leads us first to descriptions of potential networks. For any ego in a society, we may examine those institutions and groups of which he is a part and determine the individuals with which he may potentially come into contact. In my research on the Danish-Americans, such institutional frameworks could be listed. In addition to frameworks shared by many Americans such as kinship, occupation, residence, neighbourhood, and the like, these Danes shared membership in one or more Danish-American voluntary associations. Such a listing of institutional frameworks and the potential network resulting from role relations can be useful in specifying the extent and nature of an individual’s possible participation in the wider society and exposure to varying populations. In an investigation of ethnic communities within a large urban centre, this information is valuable in a determination of the completeness (Breton, 1964) of a person’s commitment to his own community. Most Danish-Americans participated in non-Danish occupa- tions; they were not localized in an ethnic residential area; and they utilized generally available non-ethnic city services such as supermarkets, theatres, and department stores. Thus, only one formal aspect of life - voluntary association membership - had an ethnic component. And even this sphere of life was not totally restricted on an ethnic basis. Some belonged to non- ethnic fraternal organizations such as the Masons and Moose lodge.

SITUATION

A statement of potential networks does not, however, get at the quality of urban social life; we do not discover the actual extent and meaning of participation in particular aspects of society. As well, such a network is so unwieldy and ill-defined that it gives us little information about personal groupings. To secure data on everyday social life, the interactive compo- 249

nent must be investigated. That is, we must discover which of an individual’s many potential role relationships are activated, in which situations this occurs, and the nature of that interaction. Network analysis provides an important method for gathering these data.

Networks can be seen in terms of the kind of institution within which they were recruited and activated. There may be areas within an individual’s potential network which are actualized only in specific situations in which the interaction is limited and its focus is very specific. For example, a factory worker might interact with his fellow workers only while on the job. Such a network could be very closeknit, but its generalizability into other spheres of life might be limited. We know from the Western Electric studies and other small group research (6. Homans, 195 1 ) that the content of such interactive links in the occupational sphere relates to more than just the specific job. However, these relationships might be segmented from other areas of the urban dweller’s life; that is, there may be little inter- actional overlap between members of a network activated in a single situation and people whom the city resident sees in other aspects of his life.

Using Parsons’ terminalogy (1951), we can speak of the instrumental nature of the content of such network relations. Thus, of an individual’s potential network of relations - as defined by the institutional frameworks of which he is a part - some will be more oriented toward pragmatic ends. The content of those relations when activated is more likely to be specific to the role relationship and these links are more likely to be activated in a limited number of situations; situations largely confined to immediate tasks.

Members of the Danish-American community maintained varying num- bers and types of such instrumental networks. Occupation, for example, provided the mechanism for recruiting people into social networks which were activated only during the work day. One informant told me that he disliked seeing fellow dock workers during off hours because he preferred to relax and not be forced verbally, “to reload every ship in my own living room.” Neighbourhood ties too were seen in a highly instrumental fashion. One lady told me that it was important to her to remain on good terms with her neighbours so they could be called upon for assistance in emer- gencies. She, however, carried the relationships no further because they had little - other than neighbourhood affairs - to talk about.

To examine relationships in networks whose members interact with each other on bases beyond the activation of single role relations, situation becomes an important variable. Networks based upon instrumental role sets are likely to be activated only in situations calling for the completion of a particular task in society; the links are not generalized into other interactive contexts. However, many individuals maintain networks whose component social bonds are multi-stranded. Different individuals in the network can have different types of relationships with ego and the content of any particular dyadic link in the network may be diverse. Networks with multi-stranded links contain bonds which could have been initially re- 250

cruited within an institution which implies limited interaction situations, but the ties are activated in other, institutionally unrelated, contexts. For example, one of my informants expanded a minimal neighbouring role with the man who lived next door. Not only did they engage in neighbouring patterns (Keller, 1968) but also they belonged to the same (non-Danish) voluntary association and engaged in an extensive visiting pattern. The addition of these two other situations presented them with different tasks to be confronted, the opportunity for additional periods of interaction, and the chance to explore new facets of each other’s interests, beliefs, and the like (“to get to know each other better”). The Dane’s personal set included four couples from the Danish lodge, two couples related by kinship, and his neighbours. Although these relationships may be structurally designated on the basis of their boundaries and degrees of connectedness, it is also important to specify the content since the various segments of this personal set were recruited in difTerent ways and interact in separate situations. The two aspects of content, recruitment and situation, influence the structure of the personal network.

RECRUITMENT AND SITUATION

In extraordinary situations, such as the twenty-fifth wedding anniversary party held by one Danish-American couple, segments of a personal set which normally do not interact may be included in the same social context. At the Madsen’s anniversary party, people were present who had relations with the couple based on kinship, occupation, and association membership. I was greeted at the door by “another Mrs Madsen.” The obvious identification of the kin tie and my response with an association link was a necessary social tool for sorting guests into categories. On the patio, where most of the guests congregated, there was some inter-penetra- tion of individuals from separate networks. This was particularly true for kin and association members since they could discuss Danish topics relatively easily and because some could have been together in other, less extraordinary, interaction situations. However, the few individuals drawn from Mr Madsen’s occupation-based network, one of whom was black, interacted almost exclusively with each other or with members of the Madsen nuclear family. There were few shared topics of conversation because of the discrepancy in recruiting frameworks.

The relation between recruiting institution and situation is a close one; however, they must be kept analytically separate. The institutional framework provides for role relationships among people and the possibility of a further social relationship among individuals (potential network). As well, the institution provides at least one social situation in which tasks must be accomplished. When network relations are abstracted from the role set, one aspect of the content of those links is the shared roles of the participants. In addition, the experiences of individuals in 251

task-oriented situations becomes part of the store of knowledge a person has about those around him. In institutions with largely instrumental goals, there is little institutional provision for a wide range of shared personal knowledge about others in the same situation. That is, it is not likely that there will be varied situations for interaction to occur and participants may consistently see only one aspect of their co-participants. The content of the relationship is thus restricted.

If network ties are described specifying the institutional frameworks from which the links were recruited and in which they continue to be activated, the analyst has an operational means of designating the content of the network relationships. One can expect that those links which contain only one institutional affiliation are more likely to be segmented from other relationships in society and might be best considered as a role relationship with less personal content included. However, if interaction occurs in a series of varied social situations, the interpersonal content may be more complex. Thus, both recruitment context and situation must be specified.

A network link which includes recruitment from more than one institutional framework can be expected to have more varied content - by definition on the institutional level. (I continue to refer to the role aspect of a link as recruitment even though it is obvious two individuals who have a relationship from one context are not “newly recruited” when they meet in another context. They are, however, newly recruited for that aspect - the new role relationship - of their multiplex social bond.) For example, the lodge member who recruits an occupation mate into the lodge shares at least two role relationships with him and can see him in at least one more social context - a context in which difEerent tasks are primary to the interaction and other aspects of personality may be discovered. The overlap of interaction situations and the possibility of activating other kinds of social roles - e.g., those from a voluntary association - in addition to the original recruitment arena is reminiscent of the kind of role overlap and participation in a series of institutions which occurs in tribal societies. An important difference here is that the overlap is on an individual level and is due to personal choice to a greater extent, rather than membership in an ascribed grouping which acts within interlocked institutional frameworks. This, of course, makes research more dif€icult since the investigator must discover the congruences of role for each person rather than finding general principles.

I have been discussing the ways in which the notions of situation and recruitment are useful for specifying the content of network links. Those factors have been related to the potential for relationships to form and to be expanded. I have suggested that a greater number of shared roles in institutions and of situations for interaction allow individuals the opportunity to discover more information about co-participants and to form clearer ideas about shared beliefs and attitudes. Newcomb’s research on the acquaintance process (1957) suggests that people who share similar value orientations are more likely to continue their relationship than are 252

those whose orientations differ. Shared institutional membership and the opportunity for several interaction situations provide for the presence of and discovery of shared beliefs in two ways. Institutional recruiting criteria can restrict membership to individuals who are more likely to be similar to each other. For example, most Danes who joined’the lodges were more traditional (more “Danish” in some sense), had a lower-middle class back- ground, and were “lodge men.” (That is, they appreciated the type of sociality in a fraternal organization and were more likely to be or to have been members of other fraternal orders.) Danes who did not join or who did not attend meetings, often did not share these characteristics. Secondly, interaction situations within the institutional context and those stemming from it (e.g., going out for a beer after work) can provide for a series of general small-talk subjects to be discussed which indicate the opinions of participants. The individual who disagrees or is disagreeable can be ex- cluded from the more informal aspects of interaction even though formal role requirements may continue to be met. Those who continue to interact have the occasion to expand their relationship on a more personal level, resulting in a difference in content from that implied by their existing role relationship.

DYADIC CONTRACTS AND AFFECTIVE SOCIAL TIES

It is in the examination of relationships which transcend those formally set up by the institutional framework that Foster’s dyadic contract is most relevant. His discussion provides an important element in the description of network ties which have been generalized beyond the recruiting institutional framework. Foster hypothesizes that:

... every adult organizes his societal contacts outside the nuclear family by means of a special form of contractual relationship. These contracts are informal, or implicit, since they lack ritual or legal basis. They are not based on any idea of law, and they ace unenforceable through authority; they exist only at the pleasure of the contractants. The contracts are dyadic in that they occur only between two individuals; three or more people are not brought together. The contracts are noncorporate, since social units such as villages, barrios, or extended families are never bound (1961 : 1174).

In an industrialized urban area, dyadic contracts may be formed within any of the institutional frameworks within which one interacts. Those frameworks which provide for sustained interaction over time - even though the types of situations in which interaction occurs are relatively similar - are more likely to be the arenas within which dyadic contracts are formed. Shared type and place of occupation, thus, is more likely to result in dyadic contracts than merchantcustomer or doctor-patient relations. Voluntary associations, especially those which are expressive (Rose, 1953) can be an important arena for the creation of dyadic contracts and subsequent informal social networks.

Social networks which are based upon dyadic contracts among the com- 253

ponent individuals differ from the more role-based networks in several respects. Most importantly, since the dyadic contract is a relationship generated interpersonally and is not totally dependent upon the outside institution, there is a much greater range of possible behaviour. The content of interaction may cover many areas of urban life. Under conditions where sustained interaction may and does take place, Homans (1961; cf. Newcomb, 1957) suggests that one consequence is an increase in the degree to which individuals like each other. If and when this occurs, the relationships may be described as having more affective content than would be expected in a link based solely upon roles.

Another type of network, then, is based upon affective ties. Relation- ships in such a network - while they may be recruited from any institu- tutional framework - might be more likely to be based upon recruitment from some institutional frameworks than others. (Which ones, of course, becomes an interesting empirical problem - both for the type of framework leading to affective ties, and for the type of individual life-style involved.) In addition, the number and type of situations in which network ties are activated differ . In a consideration of networks with a high degree of affective commit-

ment as part of the content of the links, it is important to distinguish between the affective role content of a link and the personal aspect of a link. Kinship systems provide a good example of a role structure with affective goals. Relations with distant or disliked kin may be characterized by the outward forms of a positive emotional relationship without the interpersonal commitment of a dyadic contract. In the Danish-American lodges, a high value is placed upon fellowship - members are expected to be cordial, friendly, and brotherly. A man can expect a positive reception at the lodge even if he attends rarely. One Dane ceased attending meetings because he did not receive the warm reception he expected. On one occasion, I asked a lodge member who his friends were. He waved in the direction of the lodge room and exclaimed: “Why, they’re all my friends!”

NETWORKS I N T H E DANISH-AMERICAN COMMUNITY

The affective nature of interpersonal ties is related to the recruiting framework (insofar as some institutions might be more likely to generate affective links) ; the number and diversity of interaction situations (because of Homans’ notion of interaction and liking) ; and, perhaps as a consequence of the first two factors, the presence or absence of dyadic contracts. These factors may be illustrated by referring again to the Danish-Americans. The population with which I worked included all formal members of Danish- American voluntary associations. These people were, therefore, all included in each other’s potential networks by virtue of common membership in an institutional framework. There were, as well, patterned behaviours of friendship and fellowship which were almost mandatory between fellow 254

association members. Thus, there was an affective component in the role relationships because of group membership. Members (as members) gathered together in two kinds of situations: semi-monthly formal meetings at which few were present, and less frequent large celebrations such as parties, picnics, and formal installations of officers. The celebrations were much better attended and, for many, were the opportunity of briefly renew- ing seldom-activated social ties. From the point of view of the seldom-seen member, these social links are part of his extended network. He devotes his time to other aspects of life and presumably has more continuous social relationships there. The errant member is seen in much the same way by regular attenders - part of their extended network.

From the observer’s point of view, we may speak of the wide scale, loose-knit network based upon Danish association membership. The ties were, therefore, recruited through the associations. Such links persist since members see one another at meetings and may be casual friends. In addition, some amount of personal information about other members is gathered over time. A lodge member, for example, may know details about other men’s lives - men whom he never sees except at lodge meetings. In this way, a man can judge shared orientations and the potential for a close relationship at a distance. Knowledge of and acquaintanceship with others is extended through cities in the metropolitan area and elsewhere by means of various social events which take place. For example, all the lodges co- operate on a picnic every June. At this event, people from different lodges who previously did not know each other might be introduced by mutual friends. New members are included at the tables of family and friendship groups and are introduced to the people who stroll around to greet seldom- and often-seen friends. At large social occasions sponsored by a single lodge, such as an installation of officers, intra-lodge ties may be formed. For example, Ed Hansen and Eric Svendsen and their wives sat near three other couples at a dinner when they first joined the lodge. These couples have since expanded their relationships into close friendships. It is significant that the possibility of interaction such as that illustrated above is provided by common membership in the association. Expansion of a relationship is then based upon factors of common interests and shared values. In the case of the two couples at the dinner, similarity of age was an important factor.

Within the two Danish-American fraternal lodges, it was possible to identlfy a few close-knit networks which were composed of close friends. These informal social networks could be observed interacting on meeting nights. Before and after meetings, there was a much greater likelihood that members of an informal social network (an altercentric set) would interact among themselves than with other members. This was particularly the case for those who enjoyed playing cards. One network composed of about seven men regularly attended meetings of both lodges and spent most of their time engaged in card playing. (Four would play and the others either 255

watched or talked with members outside the cluster.) As well, members of these informal clusters saw one another in other situations more frequently than they saw other Danish association members. The card players gathered at each other's houses on a regular basis for evenings of cards. As well, there was informal visiting among the couples. This intensification of institutional bonds occurred partly because the men and their wives had discovered they liked each other and enjoyed similar entertainments and formed friendships beyond the roles implied by common lodge membership. Participation in a series of difEerent situations - e.g., organization meetings, informal visiting, and recreation - allows participants the opportunity to discover more of their common attributes.

These close-knit informal social networks, generated from the common institutional framework, can be seen as partly isolated clusters - or social sets, to use Mayer's terminology - which can activate a series of aspects of urban life. Membership remains relatively constant and institutions and situation may fluctuate. The constancy of participants in such a set meant, in the Danish community, that there were similarities of behaviour in those varying situations. One set, recruited from the more conservative of the two lodges, frequently engaged in ethnic behaviour. For example, at an association meeting, the Danish picnic, or in informal visiting, Danish was spoken, a Danish card game played, and Danish food eaten.

Informal social network membership also had an effect upon attendance at formal meetings. These men attended more regularly than others, presumably because they knew they would have another opportunity to engage in interaction with their friends. In contrast, a member of the ethnic association who rarely attended had formed his primary relationships from his occupation mates. His primary social network did not overlap at all with his potential role partners in the Danish lodge.

I suspect that the high affective content of the interpersonal ties in these sets has an influence on the fact that the sets may be found in different situations. A network based largely upon role relations - no matter how tightly knit - is probably restricted to a smaller number of interaction situations. In addition, these situations might be more closely tied to the institutional framework in which recruitment to the network took place.

Data on the recruiting institution or institutions, the situations in which network interaction takes place, and the nature of the interpersonal ties involved are useful operational means to dehing the content of bonds in informal social networks. The focus upon the arenas of recruitment allows the investigator to relate interpersonal relations to the structural features of the general society. By considering the situations in which interaction occurs and the nature of the interpersonal tie, one may discover the relation between personal network formation and the activation of institutional ties, arid the impact of societal structural features and interpersonal relationships upon the behaviour of urban dwellers may be gauged. 256

REFERENCES

Barnes, J. A. 1954 “Class and committees in a Norwegian island parish.”

1968 “Networks and political process.” Pp. 107-130 in Marc J. Swartz (ed.),

Bott, Elizabeth 1957 Family and Social Network: Roles, Norms, and External Relationships

Breton, Raymond 1964 “Institutional completeness of ethnic communities and the personal

Chrisman, Noel 1966 Ethnic Muence on Urban Groups : The Danish-Americans.

Epstein, A. L. 1961 ‘The network and urban social organization.” Rhodes-Livingstone

Foster, George 1961 “The Dyadic contract: a model for the social structure of a Mexican

peasant village.” American Anthropologist 63: 1173-1 192. Homans, George 1951 The Human Group. New York: Harcourt, Brace and Company. 1961 Social Behavior: Its Elementary Forms. New York: Harcourt, Brace

Katz, Fred E. 1966 “Social participation and social structure.” Social Forces 45 : 199-210. Keller, Suzanne 1968 The Urban Neighborhood: A Sociological Perspective. New York:

Mayer, Adrian 1966 “The sigdcance of quasi-groups in the study of complex societies.”

Human Relations 7:39-58.

Local Level Politics. Chicago: Aldme.

in Ordinary Urban Families. London: Tavistock.

relations of immigrants.” American Journal of Sociology 70: 193-205.

Unpublished Ph.D. dissertation. Berkeley: University of California.

Journal 29 : 29-62.

and World, Inc.

Random House.

Pp. 97-123 in Michael Banton (ed.), The Social Anthropology of Complex Societies. London: Tavistock.

(ed.) , Social Networks in Urban Situations. Manchester: Manchester University Press.

Mitchell, J. Clyde 1969 “The concept and use of social network.” Pp. 1-50 in J. C. Mitchell

Nadel, S. F. 1957 The Theory of Social Structure. New York: Free Press. Newcomb, Theodore 1957 The Acquaintance Process. New York: Holt, Rinehart and Winston, Inc. Parsons, Talcott 1951 The Social System. Glencoe: The Free Press. Rose, Arnold 1953 Theory and Method in the Social Sciences. Minneapolis:

Universitv of Minnesota Press. wolfe,-Alvin ’ 1969 “Proposal for Analysis of Social Networks in Cities.” Manuscript.

257

Documents

Situation and Social Network in Cities