2.3.1 Network structuring

Network structure regulates patterns of contact between individuals. The likelihood of two unacquainted people meeting increases as the number of connections they share increases and the number of links separating them decreases (Parks 2007, 2011). Access to alternative partners will also depend in part on how one is embedded in a network, so that, for example, romantic partners whose networks are densely interconnected will be less likely to encounter alternative partners than those whose networks are loosely knit.

The role of interpersonal contacts in the diffusion of innovations is well recognized (Rogers 1962; Valente 1995). Early studies demonstrated that purchasing decisions were often more heavily influenced by consumers’ conversations with each other than by advertising or marketing (Ryan and Gross 1943). The power of this “word-of-mouth-marketing” has been amplified with the addition of online social networks (e.g., Kozinets et al. 2010).

Networks carry sentiment and influence through the closely related processes of transitivity and contagion. Transitivity occurs as sentiments, including feelings of liking and friendship, spread across the network. If X befriends Y and Y befriends Z, then X and Z are likely to become friends as well (Davis 1970). Many other aspects of behavior in networks appear to be transitive as well. Stress in one relationship, for example, may become transitive by affecting the level of stress experienced by the individual’s partner in another relationship (Haines, Marchand, and Harvey 2006).

The contagious spread of fears, rumors, and hysterias in social networks has been recognized for hundreds of years (Boss 1997). More recently, the social contagion model has been extended in an effort to explain patterns of obesity, happiness, loneliness, and smoking (Christakis and Fowler 2009). Research on the mirror neuron system also suggests that there may be a physiological basis for at least some aspects of social transitivity and contagion, although empathy also seems to require higher-order cognitive perspective-taking (e.g., Iacoboni 2009; Baird, Scheffer, and Wilson 2011).

It is also clear that networks can be deliberately activated by individuals or groups seeking a particular goal. Researchers focused on public health campaigns, for instance, have explored techniques for ensuring or accelerating the spread of information in community networks (Dearing, Maibach, and Buller 2006). Particular attention has been given to identifying and targeting key linkers or influencers within networks on the assumption that persuading them will spread information and influence more efficiently than broader efforts (Boster et al. 2011; Valente and Pumpaung 2007). There is also mounting evidence documenting the beneficial effects of recruiting friends and family to help people attempting to overcome substance abuse or cope with chronic health conditions (e.g., Glazer et al. 2003).

2.3.2 Relational sense-making

It is axiomatic that humans are constantly motivated to manage uncertainty and to make sense of their social environment (see Chapter 13, Knobloch and McAninch). Our ability to manage uncertainty about a relational partner or relationship depends not only on interaction with the partner, but also on our interaction with members of the partner’s network (Parks and Adelman 1983).

Interactions with network members contribute to relational sense-making in at least two ways. First, network members provide information about partners that the partners will not or cannot provide themselves. Partners may, for example, reveal “family secrets” (Afifi and Olson 2005; Vangelisti and Timmerman 2001). Network members are rich sources of such hidden information as well as the partner’s relational history. Network members also function as observers and may be able to tell people things about their relationships that they are not yet able or willing to see, as in the case of network members’ sensitivity to signs of dominance or abuse in relationships (Arriaga and Oskamp 1999).

Network members also serve as points of reference and comparison. As important network members become more obese, for example, individuals’ sense of “normal” weight may be skewed upwards, ultimately causing obesity to spread within the network (Christakis and Fowler 2009). Network members may be singled out for active social comparison, as when married couples compare their marriages to the marriages of their friends and family (Titus 1980). In other cases, individuals will actively seek the opinions or guidance of network members when they are uncertain about an opinion or a course of action (Goldsmith 2004).

3.1.1 Networks as a resource

Social support may be defined as “verbal and nonverbal communication between recipients and providers that reduces uncertainty about the situation, the self, the other, or the relationship, and functions to enhance a perception of personal control in one’s life experience” (Albrecht and Adelman 1987: 19). Researchers customarily distinguish several different types of social support (Cutrona and Russell 1990; Xu and Burleson 2001). Emotional support involves expressions of love or concern for another individual (Cutrona 1996) and has received the greatest attention in research because of its strong connections with positive psychological and physical health outcomes (Cunningham and Barbee 2000; Ikeda and Kawachi 2010).

Several other types of support have been articulated in the literature (Cutrona and Russell 1990; Xu and Burleson 2001). These include esteem support (communication of emotions that validate the other person), network support (behaviors and expressions that make an individual feel connected and a part of the larger social group), tangible support (provision of material goods or services to assist another person) and informational support (sharing information about a situation of interest or concern with network members).

Support seekers behave strategically because they recognize that network members vary widely in their willingness and ability to provide specific types of support. Emotional support, for example, is usually sought only from those with whom one has a strong relationship because of the degree of familiarity and trust necessary for the effort to be perceived as successful or even appropriate (Burleson 2003; Goldsmith and Parks 1990). Both parties are likely to be uncomfortable when these conditions are not met (High and Dillard 2012).

On the other hand, weaker, less developed relationships may be more effective in providing access to new information and diverse resources. In a classic study, Granovetter (1973) showed that job seekers were most successful at finding new work when they targeted their weak network ties – that is, the friends of friends within their network – rather than close friends for informational and network support. Other strategic considerations may prompt people to seek support, even emotional support from weak ties. They may, for example, wish to prevent strong ties from knowing about a problem (Parks 2007). In other cases, weak ties found in comparatively anonymous online support groups may help people disclose embarrassing problems, find information not known by members of one’s face-to-face networks, or even find support at times when regular network members are not available (Walther and Boyd 2002).

3.1.2 Research on adequacy of support networks and health

A network with a diversity of ties that provides access to different types of support benefits physical health and psychological well-being (see Chapter 21, Duggan and Thompson). In a seminal study examining social support and mortality, Berkman and Syme (1979) found that a lack of community ties was associated with an increased risk of death at a nine-year follow-up, even controlling for socioeconomic status and risk factors such as smoking, alcohol consumption, and obesity. Subsequent studies corroborated these findings in different populations and countries (Kaplan et al. 1988; Orth-Gomér and Johnson 1987). Having a network containing strong, supportive relationships has been linked to specific health outcomes including enhanced immune response (Cohen 2004; Pressman et al. 2005), decreased levels of anxiety, depression, and psychological stress (Cohen and Willis 1985), and lower risk of cardiovascular disease (Kaplan et al. 1988). For a summary of these and related findings, see Parks (2007) or Ikeda and Kawachi (2010).

Two explanations have been advanced to account for the role of support networks in health and well-being. The first, called the stress-buffering hypothesis, proposes that social support serves a protective function by mediating the effects of stress on physiological and psychological functioning. That is, support from network members buffers or reduces the negative effects of stress by helping people cope, which, in turn, limits their negative emotional and physiological responses to the stressor (Lovell, Moss, and Wetherell 2012; Stephens, Alpass, Towers, and Stevenson 2011). For example, a study of elderly Norwegians found a strong association between perceived social support and psychological distress, with those reporting lower levels of support experiencing significantly higher levels of distress and poorer physical health (Bøen, Dalgard, and Bjertness 2012).

The second, or direct or main effects hypothesis, argues that strong social relationships are directly beneficial independent of the level of stress experienced by the individual. Social connectedness ensures regular positive interactions, a sense of predictability and stability in one’s life, and a sense of belonging that is protective and promotes positive health outcomes (Cohen and Willis 1985). Network characteristics, such as overall network size, density, and scope (having connections in multiple different domains) have been shown to directly impact resistance to disease, depressive symptoms, cardiovascular disease, cognitive decline, and all-cause mortality (Berkman and Syme 1979; Crooks et al. 2008; Fratiglioni et al. 2000; Haines, Beggs, and Hurlbert 2008; Vogt et al. 1992). Cohen and colleagues (1997), for example, found that individuals who were exposed to the common cold virus were less likely to contract a cold if they had more diverse social networks. This was true regardless of their level of social support and may suggest a direct network role in immune functioning.

Social networks may thus provide people with access to material resources and a sense of meaning and purpose for their lives that protects and enhances their overall well-being (Berkman et al. 2000; Rook 1990). In a 12-year study tracking the relationship between social engagement and cognitive health in an elderly cohort, for example, those who had regular contact with a greater number of social ties experienced lower levels of cognitive decline even after adjusting for factors such as initial cognitive performance, sensory impairment, physical activity, education, and risk factors such as depression, alcohol use and smoking (Bassuk, Glass, and Berkman 1999). The direct effects and stress-buffering hypotheses are not mutually exclusive. Just being connected to others appears to confer health benefits, but there are obviously times when people actively seek emotional, material, and other kinds of support in an effort to cope with life stressors.

3.2.1 Network activation

Seeking support can be a complicated business. People must weigh multiple concerns when seeking support, including what their request might imply about their current situation, how it may cause them to appear, and how it might affect their relationship with the support provider (Albrecht and Adelman 1987; Goldsmith 2004). They may also be concerned with if and how their disclosures will be shared with other network members (Parks 2007).

Politeness theory (Brown and Levinson 1987; see also Goldsmith 1992) provides one framework for understanding these multiple concerns by grouping them into two broad categories: positive face (the desire to be seen positively by others) and negative face (the desire to maintain autonomy). An individual desiring support might worry, for example, that they will be perceived as needy or incapable (a threat to positive face), and the person being asked for help may feel imposed upon or burdened (a threat to negative face). Similarly, unskilled or unsuccessful efforts to gain support threaten one’s positive face, while unwanted offers of support jeopardize negative face.

Sensitive interaction systems theory (SIST) offers another framework for examining how people navigate these complex interactions (Barbee and Cunningham 1995; Barbee et al. 1993). According to SIST, support seekers select strategies based on factors such as the severity of the problem and their own and others’ personal characteristics. Their strategies can be classified along two distinct axes: direct vs. indirect and verbal vs. nonverbal. A direct verbal behavior might involve explaining the problem and simply asking for help, whereas a direct nonverbal behavior might involve an outward display of distress, such as crying. Examples of an indirect verbal behavior might be complaining or hinting about the problem, while sighing or skulking reflect indirect nonverbal efforts to elicit support.

The support provider’s response can also be classified along two axes (Barbee and Cunningham 1995): problem/emotion (whether the focus is on the problem or the emotions associated with it) and approach/avoid (whether the response engages the problem or emotion or instead tries to minimize its importance or distract the support seeker from it). The combination yields four distinct interactive coping behaviors. Solve behaviors, which are problem-focused/approach behaviors, involve efforts to clarify the problem and respond to it. Similarly, solace behaviors (emotion-focused/approach) attempt to comfort or soothe the distressed individual. On the opposite side of the spectrum are dismiss (problem-focused/avoidance) and escape (emotion-focused/avoidance) behaviors. SIST has been applied to a range of support encounters within networks, including how HIV positive individuals interact about their illness with social network members as well as how romantic couples negotiate support following adverse medical diagnoses (Derlega et al. 2003; Yankeelov et al. 1995). SIST principles (such as the direct/indirect and approach/avoid axes) have informed more recent typologies of support activation, such as those used when eliciting support for a romantic relationship from network members (Crowley 2012).

3.2.2 Strategic management of support in personal networks

Efforts to obtain support are sometimes thwarted when others respond with unhelpful behavior (Ford and Ellis 2004). Unhelpful or unwanted support is often a concern, especially when one’s problem is apparent to others, as in the case of people with visible disabilities. Strategies commonly used in such situations include ignoring offers of unwanted assistance, directly refusing assistance, and using humor to deflect the offer of assistance (Braithwaite and Eckstein 2003). Additional insight comes from a study of how obese and overweight young adults who were trying to lose weight managed unsupportive responses from network members (Faw in press). Unhelpful or undermining responses were common, and participants reporting using strategies that ranged greatly in their directness. Some showed a high level of creativity, as in the case of participants who would plan ahead by offering to cook at home or selecting a healthy restaurant so as to avoid temptations introduced by network members. Others deliberately ignored unsupportive network members or confronted them directly.

Social support poses other drawbacks as well. Providing support can be an exhausting and burdensome task (Lu and Argyle 1992). Professional support providers as well as lay caregivers providing long-term care for others frequently experience stress, exhaustion, and poorer health (Lee and Ashforth 1996; Lovell and Wetherall 2011; Rohleder et al. 2004).

This problem may also occur episodically. For example, an individual who has drawn on network members for support in the past may be overwhelmed when network members converge with reciprocal requests for support (Lu 1997). In other cases, well-intentioned support efforts may end up in a debilitating cycle of corumination in which participants repeatedly dwell on problems and negative affect. Paradoxically, co-rumination results in stronger bonds between participants even as it simultaneously enhances stress, anxiety, and in some cases depression (Byrd-Craven, Granger and Auer 2011; Rose, Carlson and Waller 2007).

We still have much to learn about the complicated processes of social support in a network setting, particularly about how people manage the dilemmas of seeking support or selecting strategies for activating support within a network. We know much more about single encounters than we do about how people manage support across a broader network. We do know, however, that the burdens experienced by lay and professional caregivers can be at least partially offset if they can regularly obtain social support themselves (Lovell and Wetherall 2011). Clearly social support is a network affair. So, too, as we see in the next section, is the way we go about forming interpersonal relationships.

4.1.1 Social proximity and relationships

Even before they meet for the first time, two people may be indirectly related within a larger network structure. Rick and Albert, for example, were indirectly linked through John and Nick’s close friendship (Figure 2). They were separated by only three links (degrees). According to Social Contextual Theory, the likelihood of any two unacquainted people meeting increases as the number of links separating them decreases and the number of links between the members of their networks increases (Parks 2007).

Research in the U.S. and Europe demonstrates that between 50% and 70% of married, cohabitating, and dating couples as well as same-sex friends remember having had one or more common contacts prior to meeting for the first time (Kalmijn and Flap 2001; Laumann et al. 1994; Parks 2007). In some cases, meetings may have occurred simply as a result of social proximity, but in many other cases network members appear to take an active role. Research is limited but suggests that over 50% of romantic or sexually involved couples were introduced to one another by network members (Laumann et al. 1994; Parks 2007).

Social proximity continues to play a role even after relationships come apart. Divorced couples who no longer interact with each other directly, for example, may nonetheless continue to keep track of one another and influence each other indirectly through children, friends, and other contacts they still hold in common. There is currently no research that would establish the prevalence of this phenomenon, but it is clear that the growth of online social networking sites such as Facebook facilitates such monitoring and that users are often concerned about unwanted tracking and observation by other network members (Binder, Howes, and Smart 2012). Although such monitoring can be benign or even helpful, such “cyberstalking” can also result in harassment or worse (Lyndon, Bonds-Raacke, and Cratty 2011; Spitzberg and Hoobler 2002).

4.1.2 Contact and communication with network members

The development of personal relationships both drives and is driven by contact and communication between participants and the members of their partner’s networks. As relational participants meet more members of their partner’s network, their relationship becomes closer. We view this association as mutually causal, so that increasing linkage with the partner’s network also reinforces the relationship and enhances its stability. Evidence of the links between contact with the partner’s network and the emergence of a shared network comes from a variety of studies on adolescent and young adult same-sex friendships and adolescent and young adult romantic relationships (Agnew, Loving, and Drigotas 2001; Kim and Stiff 1991; Milardo, Johnson, and Huston 1983; Parks 2007). The effect also appears in marital relationships where satisfaction and commitment tend to be greater in couples who have more contact and overlap between their networks (Julien and Markman 1991; Stein et al. 1992).

The amount of communication one has with members of the partner’s network, independent of the number of people one has met, also appears to be linked with measures of development. Studies of same-sex friendship and romantic relationships in both adolescent and young adult samples consistently report that how often one communicates with members of the partner’s network is strongly associated with measures of closeness, commitment, and communication between the partners (see Parks 2007 for summary).

The overall pattern of research results clearly links the development of personal relationships with increases in the amount of contact and communication with the partner’s network and with the emergence of a joint network. But several qualifications should be noted. First, although the evidence supports a mutual causal relationship between development and network engagement, there is as yet insufficient longitudinal work. Furthermore, it may be that perceptions of relationship development or quality may be more closely related to network factors for one partner than for the other (Kearns and Leonard 2004) and that the positive effects of network engagement may be greater early in relationships and after significant relational transitions (e.g., Sprecher and Felmlee 2000).

In addition, the members of the respective partners’ networks should have more contact with each other as the partners’ relationship becomes closer and less as it deteriorates. We have already noted that networks tend to become more shared, overlapping as the partners’ relationship develops. Relationships in which participants are embedded in more densely linked, overlapping networks are also less likely to terminate (Agnew, Loving, and Drigotoas 2001; Feeley 2000; Parks and Adelman 1983). Researchers have paid less attention to the level of contact between the members of relational partner’s separate networks, although there is some evidence suggesting that this cross-network contact is also positively associated with indices of relational development (Parks 2007).

Differences in the level of network contact that characterize relationships of differing levels of development can be clearly seen in a comparison of Figures 1 and 2. Acquaintances Rick and Albert have no close friends in each other’s networks and few members of their respective networks are closely connected to each other (Figure 2). Close friends Marc and Daniel, however, each have close friends in the other’s network and members of their respective networks have more close connections with each other (Figure 1).

4.2.1 Appraisal and support

Network members may provide emotional and esteem support as well as information or tangible goods and services that are intended to enhance the participants’ relationship. In this way, network appraisal and support may be seen as a special class of social support in which the support is directed toward a relationship rather than an individual.

Researchers interested in relationship development have usually focused on generalized perceptions of support or opposition from network members rather than on assessments of particular kinds of support. Lewis (1973), for instance, asked dating partners to indicate how often “friends and family” invited them to social events together and made positive comments about them as a couple. Other researchers have asked relational partners for global assessments of “approval/ disapproval” among key groups (e.g., “partner’s family”) or from individuals listed as important to the relational participants (e.g., Felmlee 2001; Parks, Stan, and Eggert 1983).

Findings have consistently demonstrated that personal relationships become more intimate and more stable when network members are perceived to be supportive and more likely to dissolve when they are not. This appears to be the case for both heterosexual romantic relationships and same-sex friendships (Felmlee et al. 1990; Felmlee 2001; Lewis 1973; Johnson and Milardo 1984; Parks and Adelman 1983; Parks 2007). It is possible that opposition from network members may induce psychological reactance (Brehm 1966) that, in turn, causes couples to become more committed, but the evidence supporting this “Romeo and Juliet effect” is limited and subject to a number of qualifications (Driscoll, Davis, and Lipetz 1972; for summary see Parks 2007). Perceived support from the partner’s network and from one’s own network are associated with relational closeness and commitment in almost identical ways, although there is some evidence to suggest that in interethnic romantic relationships, participants’ feelings for their partners are more closely linked to perceived support from their own network than from the partner’s network (Parks 2007).

A second body of research related to network appraisal and support deals with the provision of financial and material aid by network members. In stressful times, such as those following a natural disaster, people typically seek assistance from family and friends before turning to social service agencies (e.g., Aten et al. 2012). Financial assistance, child care, and assistance with the tasks of daily living are among the more important forms of aid provided by network members (e.g., Swartz 2009). More attention has been focused on emotional and informational support, but the negotiation of financial and material assistance poses dilemmas for information and impressions that are now being explored by communication researchers (e.g., Edwards, Allen, and Hayhoe 2007; Plander 2013).

4.2.2 Managing network appraisals and support

Relational participants are not simply passive registers of support or opposition from network members. Instead they actively “work the network” so as to manage the image of their relationship, the demands placed on it, and the resources available to it. This was particularly apparent in one of our studies of how network members helped individuals initiate romantic relationships (Parks 2007). We found that network members played a variety of roles, including arranging for prospective partners to meet (direct initiations), saying positive things about one to the other (attraction manipulations), and providing coaching and other services to assist the partners in the early stages of their relationship (direct assists). In more than 85% of the cases, however, one or both recipients had either explicitly asked for the network member’s assistance or had strongly hinted that it was desired.

Relational partners also actively attempt to manage support and opposition from network members. Like managing others’ views of the self, managing impressions of one’s relationship raises concerns of positive and negative face (Brown and Levinson 1987). And indeed many of the same concerns with image and autonomy apply at both the individual and relational level. Relational partners, for example, provide, withhold, or selectively present information to network members in order to promote a positive image of their relationship (e.g., Leslie, Huston, and Johnson 1986; Parks 2007). They may attempt to shield their relationship from network members who are perceived to be disruptive (Felmlee 2001). When troubles occur inside the relationship, they strategically select sympathetic listeners in the network and try to prevent others in the network from knowing (Goldsmith and Parks 1990).

It should not be assumed that the appraisals being sought are always positive ones. In deteriorating relationships, for example, a disaffected partner may seek validation for leaving the relationship from network members. Finding network members who accept the individual as he or she transitions out of a relationship promotes post-relational adjustment (Jacobs and Sillars 2012; Krumrei et al. 2007). These individuals may be especially important because network members are typically lost following the termination of a close personal relationship such as marriage (Terhefl, Van Groenou, and Van Lilburg 2004).

Managing appraisals and support poses dilemmas for relational partners and network members alike. Dilemmas of image involve positive face concerns including the desire to have one’s self, partner or perhaps the overall relationship be seen in a good light, while at the same time not being seen as too needy or manipulative. Those providing appraisals seek to be responsive without being seen as too intrusive or judgmental. Dilemmas of autonomy arise as relational partners attempt to maintain relational boundaries and avoid being imposed upon, while at the same time obtaining the validation and resources they seek. Support seekers may worry about future expectations for reciprocation, while support providers may wish to help without feeling exhausted or otherwise burdened (Albrecht and Adelman 1987; Parks 2007).