Getting Down to Basics: A Situated Model of Conflict in Social Relations

Social conflict is a fact of life. It can be a basic driver of human development, constructive social change, innovation, and cultural evolution, or it can lead to roiling animosities between disputants, dysphoric rumination, cycles of revenge, and violent atrocities. For decades, research has provided important insights into the nature of social conflict and conflict resolution (see Deutsch 2002; Kriesberg 2003; Pruitt and Kim 2004; Deutsch, Coleman, and Marcus 2006; De Dreu 2010 for summaries). But these efforts have tended to be atomistic, focusing on separate features of conflict. Some researchers, for example, have investigated the effects of power asymmetries on conflict and negotiations (see Gurr 1970, 2000; Rubin and Brown 1975; Tjosvold 1981, 1991; Blalock 1989; Zartman and Rubin 2002; Kim, Pinkley, and Fragale 2005; Van Kleef et al. 2006; Magee, Galinsky, and Gruenfeld 2007), while others have investigated the effects of different motives operating in isolation (see Deutsch 1973, 2006a; Carnevale and Lawler 1986; De Dreu and Van Lange 1995; Beersma and De Dreu 2002; Johnson and Johnson 2005), different degrees of dependency (Ng 1980; Mannix, Thompson, and Bazerman 1989; Gerhart and Rynes 1991; Kim and Fragale 2005), and different conflict styles or strategies for intervention (Schelling 1960; Thomas 1976; Tjosvold 1991; Johnson and Johnson 2003; Kressel 2006). As a result, the findings from research on social conflict are often piecemeal and contradictory (Zartman and Rubin 2002; Coleman 2003, 2004a; Kim, Pinkley, and Fragale 2005; Tjosvold and Wisse 2009; De Dreu 2010) and thus provide an unsteady foundation for effective conflict resolution.

What is more, the gap between research and practice in the field of conflict resolution continues to be unacceptably large (Honeyman 1997; Honeyman, McAdoo, and Welsh 2001; Deutsch, Coleman, and Marcus 2006). More than a decade ago, researchers evaluating centers that study conflict resolution found that the work of most practitioners surveyed had been largely unaffected by the contributions of these centers (new theory, tactics, evaluation tools, etc.; Honeyman 1997). In addition, much of the research conducted at these centers was found to be removed from practical realities and constraints. A similar chasm was exposed in the medical field more than one hundred years ago when the Carnegie Foundation released the Flexner Report (Flexner 1910), which identified a wide rift between science and physician practice and led to a considerable uproar and significant reform in the education and licensing of physicians. This divide presents one of the more serious challenges to the field of peace and conflict today (Gerami 2009).

In this article, we present an integrative model for the study of conflict in dyadic social relations. Our goal is to return to research on the more basic aspects of social relations to gain a fresh perspective on a topic that is characterized by inconsistent conclusions despite decades of research. Rather than isolating specific variables and looking at their separate effects, here we build on research that has identified the fundamental features of social relations and then employ these features to create a basic framework for the study of conflict. Thus, our model is integrative and heuristic, weaving together previously disparate strands of research to establish a more comprehensive framework for research and practice.

We begin by summarizing relevant theories and conclusions that have been generated by conflict scholars, highlighting their diffuse and disparate natures. We then describe the fundamental dimensions that underlie social relations, as identified by Morton Deutsch (1982, 1985, 2007, 2012) in his work on social relations and psychological orientations (POs). Next, we present our situated model, which attempts to recast empirical generalizations about social conflict in terms of the interplay of these fundamental dimensions of social relations. We then employ our model to generate a set of propositions concerning various conflict scenarios. Some of these propositions find support in existing literature, while others are heuristic and suggest an agenda for further empirical research. In doing so, we shed light on how future research on conflict can help us to better inform practice and further the development of coherent, cohesive, and integrated theories.

Although the literature on social conflict is extensive (see Pruitt 1998; Deutsch 2002; Kriesberg 2003; Pruitt and Kim 2004; Deutsch, Coleman, and Marcus 2006; De Dreu 2008 for overviews), our approach will be focused. Our goal in summarizing some key aspects of the literature here is not to be comprehensive. Rather, we seek to capture what notable research has taught us to date about conflict dynamics in social relations and to underscore both important gaps in this research and opportunities for integration to inform both our model and future research in this area.

The ways in which social conflict has been theoretically conceptualized and operationally defined in psychological research have overlapped considerably, although important distinctions do exist. Mary Parker Follett, an early scholar of organization behavior and pioneer in the study of conflict, defined conflict simply as “difference” (1973: 30). Pioneering social psychologist Kurt Lewin defined conflict more broadly as “a situation in which oppositely directed forces of about equal strength play upon a person simultaneously” (1936: 114). Morton Deutsch (1973), a leading conflict scholar and student of Lewin characterized conflict as existing whenever incompatible activities (opposing goals, claims, beliefs, values, wishes, actions, feelings, etc.) occur. Psychologists and conflict scholars Dean Pruitt and Sung Hee Kim (2004) logically extended these prior definitions by describing conflict as arising from perceived divergence of interest (see also Coser 1956; Schelling 1960; Boulding 1962; Kriesberg 2003 for similar definitions).¹

Our definition of social conflict builds on the work of Follett, Lewin, Deutsch, and others but also acknowledges contemporary calls to view conflict not as a single event or situation occurring at a specific moment in time, but rather as a process unfolding in relationships over time (Pondy 1967; De Dreu and Gelfand 2008; De Dreu 2010). Accordingly, we define social conflict as a relational process influenced by the presence of incompatible activities. These processes typically occur in a relational context that has a history and a normative trajectory. In other words, conflicts, or incompatible activities, often only perturb the flow of ongoing psychosocial processes.

Five theoretical models of dyadic conflict and conflict resolution stand out in the literature as particularly influential in the field today and to some degree, amenable to integration, although their development has been largely independent or in parallel. They are:

• social interdependence theory;

• social motivation theory;

• dual‐concern theory;

• power dependence theory; and

• game theory.²

Each of these models has focused generally on understanding the conditions and processes that lead to constructive versus destructive conflict dynamics and outcomes (see Deutsch 1973, 2002; Deutsch, Coleman, and Marcus 2006; De Dreu 2008), although they differ in their assumptions and areas of emphasis and applicability. Here, they are summarized briefly.

Based on Lewin's insight that interdependence is the essence of group dynamics (Lewin 1936, 1948), this theory specified the basic conditions and processes involved in constructive versus destructive conflict (Deutsch 1973). Deutsch's (1949a, 1949b) earlier research showed how different types of goal interdependence between parties — positive or cooperatively linked goals versus negative or competitively linked goals — affect constructive versus destructive processes and outcomes in groups, respectively. Deutsch (1973) found that constructive conflict resolution processes are similar to cooperative problem‐solving processes in which the conflict is seen as a mutual problem by the parties and addressed jointly, and that destructive conflict processes are similar to competitive processes in which the conflict is framed and approached as a win–lose struggle. The research on cooperative and competitive goal interdependence (see Deutsch 1973, 2006a; Johnson and Johnson 1989, 2005) has regularly demonstrated the contrasting effects of cooperation versus competition in conflict. It has shown that the perception of cooperative goals between people and between groups, when compared with competitive goals, leads to more friendliness, helpfulness, respect, better communication, better coordination, a sense of similarity in values and beliefs, a willingness to enhance the other's power, and the framing of conflicting interests as mutual problems to be solved together (Deutsch 1949a, 2006a; Johnson and Johnson 1989, 2005). Empirical studies on negotiation have found similarly that when parties perceive themselves to share a cooperative orientation, they function more effectively than if they share a competitive orientation (Rubin and Brown 1975; Zartman and Rubin 2002).This basic idea cascaded into a variety of propositions (outlined in Deutsch 1973) that have provided a general intellectual framework for understanding conflict and for developing methods of constructive conflict intervention.

Since its publication, Deutsch's theory of conflict resolution has been seen as one of the most important advances for the study of conflict in the last century (Jones 1998). It has been validated by a large canon of empirical studies (see Deutsch 1973; Johnson and Johnson 1989, 2005) and has led to a wide array of practical methodologies and trainings for the constructive resolution of conflict (see Tjosvold and Johnson 1983; Tjosvold 1991; Coleman and Deutsch 2001; Coleman and Lim 2001; Lewicki et al. 2003; Johnson and Johnson 2005; Deutsch, Coleman, and Marcus 2006).

Deutsch's theory, however, has limitations (see Johnson and Johnson 2005; Deutsch 2011); two of the most consequential are its assumptions regarding social power and interdependence. The original formulation of the theory assumed both equal power and high degrees of interdependence between the parties in conflict (Johnson and Johnson 2005). Thus, the empirical studies supporting the original theory occurred under those conditions.

Subsequent research on conflict and goal interdependence in asymmetrical power relations produced mixed results. Several studies demonstrated the important role of cooperative interdependence in fostering more constructive power dynamics between parties whose power was unequal (Tjosvold 1981, 1985a, 1985b; Tjosvold, Johnson, and Johnson 1984; Tjosvold, Coleman, and Sun 2003; Coleman 2004b). And cooperative goals, when compared with competitive and independent goals, were found to induce “higher expectations of assistance, more assistance, greater support, more persuasion and less coercion and more trusting and friendly attitudes” (Tjosvold 1997: 297) between those with more and less power.

Other laboratory and case‐based research produced contradictory evidence, however. Researchers reported that under conditions of asymmetrical power, high‐power parties tended to behave in a more domineering manner even when they shared common goals, acting coercively and exploitatively to fulfill their wishes (Dwyer and Walker 1981; McAlister, Bazerman, and Fader 1986; De Dreu 1995; Lin and Germain 2003; see also Rubin and Brown 1975; Zartman and Rubin 2002 for summaries). These inconsistent findings have yet to be reconciled theoretically or empirically.

Another influential model of social conflict, informed by the work of Harold Kelly and John Thibaut (Thibaut and Kelley 1959; Kelley and Thibaut 1978) on interdependence and Charles McClintock (1976) on social motives, describes how both individual and situational differences in interdependence affect people's social orientations and thus their values and behaviors when negotiating disputes (see De Dreu et al. 2007; Van Lange et al. 2007 for summaries). Scholars have identified a variety of different social motives, including altruistic, competitive, and individualistic; however, most of the research on negotiation and conflict has focused primarily on the effects of pro‐self versus pro‐social motives. Pro‐self motivation combines both individualistic and competitive goals into one motivational orientation, and pro‐social motivation combines both cooperative and altruistic goals.

Research has shown that negotiators with a pro‐self motive seek to maximize their own outcomes, have little or no concern for the other party's outcomes, tend to see negotiation as a competitive game in which wielding power and winning are key, and selectively search for and process information that is consistent with this competitive view (De Dreu et al. 2007). In contrast, negotiators with a pro‐social motive seek fair outcomes that maximize both their own and the other party's goals; tend to view negotiations in more cooperative terms in which harmony, solidarity, and fairness are important; and seek information that validates this view. This research has found that while social motives are largely determined by personality differences (De Dreu and Van Lange 1995), the tendency to adopt a pro‐self or pro‐social orientation can also be triggered by rewardstructures (e.g., Weingart, Bennet, and Brett 1993) and social cues (e.g., Burnham, McCabe, and Smith 2000).

The research on social motives has indicated that motivational orientations affect short‐term responses to conflict and has also shed light on some of the social‐cognitive processes associated with pro‐self and pro‐social motives, but it has its limitations. First, by treating distinct motives such as competition and individualism in conflict similarly, it often conflates what may in fact be important conflicting tendencies in some social situations (maximizing one's own outcomes versus accepting less in order to defeat the other party). In addition, social motives researchers have investigated the comparative effects of having distinct motivational orientations (cooperation, egalitarianism, altruism, etc.) on conflict but have yet to fully address how these orientations can be integrated into an account of how social motives function together in conflict (see Van Lange et al. 2007).

The dual‐concern theory was originally developed as a model of individual differences in conflict resolution styles (Blake et al. 1964; Filley 1975; Thomas 1976; Rahim 1983, 1986) but has subsequently been developed into a predictive theory of choice and strategy under different motivational conditions in conflict (Pruitt and Rubin 1986; Pruitt and Kim 2004). It proposes that differences in disputants' two basic concerns — concerns for their own outcomes and concerns for the other parties' outcomes (ranging from weak to strong) — combine to affect the strategies people choose in conflict, including yielding, avoiding, contending, problem solving, and compromising. With this model, Kenneth Thomas (1976) extended Deutsch's one‐dimensional model (competitive or cooperative) to include a second dimension, as he saw self‐concern and other‐concern as orthogonal interests that can function independently. These two concerns can differ according to social conditions (reward structures, social and cultural norms, etc.) and individual differences in style preferences.

Dual‐concern theory has received some empirical support (e.g., Sternberg and Dobson 1987; Van de Vliert and Kabanoff 1990) and has also been used to address how motivational differences (concerning both oneself and others) can operate orthogonally and vary by degrees of importance. Subsequent research on this model has primarily investigated the conditions that foster self‐ and other‐concerns in conflict, but has also indentified how strategic choice is moderated by people's resistance to yielding (Druckman 1994; Pruitt 1998) and the perceived feasibility of employing different strategies in particular contexts (Kelley 1967; Pruitt 1981; Pruitt and Kim 2004). The model has yet to address, however, how differences in power distribution between the parties affect the predictions of the theory, severely limiting its implications.

Many studies of power differences in negotiations are based on the theory of power dependence (e.g., Ng 1980; Mannix, Thompson, and Bazerman 1989; Gerhart and Rynes 1991; Kim and Fragale 2005). Also situated within the broader framework of interdependence theory (Thibaut and Kelley 1959; Kelley and Thibaut 1978), power dependence theory states that “the power of A over B is equal to and based upon the dependence of B on A” (Emerson 1962: 32–33). Dependence is based on two dimensions of the negotiation situation: it is directly proportional to the value attributed by a party to the outcome at stake and it is inversely proportional to the availability of this outcome through alternative sources. Laboratory research on negotiations has generally supported this model, finding that negotiators who hold more attractive best alternatives to a negotiated agreement (BATNAs) or who may achieve desired outcomes through alternative means or who are able to increase the other party's dependence are less dependent on their negotiation partners and thus possess greater power relative to them and obtain better outcomes in negotiations (Mannix 1993; Pinkley, Neale, and Bennett 1994; Kim 1997; Kim and Fragale 2005).

Power dependence theory has been particularly predictive in the realm of distributive or competitive negotiations but also has its limitations (see Kim and Fragale 2005). For instance, by defining and operationalizing power solely as “asymmetries of dependence,” the model overlooks many other types of power and influence that can shape conflict dynamics, such as social status, charisma, moral authority, access to resources, and so on. The model also fails to incorporate the potential for change in the differential dependencies among parties to a conflict over time (Kim, Pinkley, and Fragale 2005).

Another of the more influential paradigms for conflict and negotiation research emerged from economics and the study of games (Von Neumann 1928; Von Neumann and Morgenstern 1944; Schelling 1960). In 1944, John Von Neumann and Oskar Morgenstern published their now famous Theory of Games and Economic Behavior, which formulated conflicts of interest in mathematical terms. Like Lewin's (1947, 1948) work in psychology, Von Neumann and Morgenstern's approach emphasized the interdependent nature of disputants' interests, behaviors, and fates. It focused, however, on rational decision making in conflict, positing that there is typically a “rational” choice that is the best counterchoice to that of one's opponents. This approach has been most effective in predicting behaviors and outcomes in competitive, “zero‐sum” situations, but it recognized that many, if not most, conflict situations were of a mixed‐motive (cooperative and competitive) nature (Schelling 1960). Scholars contend that it is most predictive in competitive situations because of the underlying assumptions inherent in the theory of what constitutes “rational” choice (e.g., efficiency, maximization of goals, transitivity of preferences, etc.; Deising 1962; Deutsch 1985). These assumptions limit the applicability of the model to more purely distributive, economic types of conflicts.

Conflict research from a game theoretical perspective has focused largely on the rules and strategies for winning conflict games,³ but it has also sought to identify the conditions for achieving a state of equilibrium or stability between parties. An important finding from this research is that over time the outcomes of players who pursue only their self‐interest in these games tend to be worse than those of players who consider the other players' concerns when making choices.

Methodologically, game theory had a huge impact on conflict research by introducing game matrices, which are precise and efficient abstract representations of conflicts that allow for the investigation of dynamics over time. In 1977, Dean Pruitt and Paul Kimmel reported that more than one thousand studies had been published employing experimental games. But, as Deutsch wrote, “Much of this research . . . was mindless — being done because a convenient experimental format was readily available” (2002: 313). In other words, although the research shed light on more rational, competitive conflict dynamics, no broader theory emerged from game theory to better inform our understanding of the vast majority of social conflicts that involve mixed motives (cooperative and competitive goals) with both rational and “irrational” elements that operate both consciously and automatically (Pruitt 1998).

The five models we have outlined have helped theorists understand how different types and levels of interdependence, importance, and power can shape motivational orientations and affect constructive versus destructive conflict in social relations. Nonetheless, our theoretical understanding of conflict dynamics remains fractured. It currently comprises a series of mid‐level or micro‐level models of conflict antecedents, processes, and outcomes that have yet to become convergent with a more general theory of social relations. Clearly, the time is ripe for a more comprehensive approach to social conflict.

Deutsch's (1982, 1985, 2007, in press) theory of social relations and POs provides the conceptual scaffolding for our situated model. This theory emerged from his earlier empirical research identifying the fundamental dimensions of interpersonal relations (Wish, Deutsch, and Kaplan 1976). Through multidimensional scaling analysis of survey data, this research identified five basic dimensions of social relationships:

• cooperation–competition;

• power distribution (equal–unequal);

• task‐orientation versus social‐emotional orientation;

• formality versus informality; and

• degree of importance.⁴

Deutsch (1982) later theorized that variations in these dimensions, when combined in situations, afford distinctive types of social relations and that a particular type of social relation would promote a particular type of PO in people (see Thibaut and Kelley 1959; Triandis 1972; Kelley and Thibaut 1978; Kelley 1979, 1984, 1991 for similar models).

POs are more or less a consistent blend of cognitive, motivational, moral, and action orientations that guide one's behaviors and responses (Deutsch 2007, 2011). Because of both internal and external pressures for consistency, specific types of situations will tend to encourage appropriate POs that “fit” the situation, and different types of POs will tend to propel people toward social relations that are consistent with their orientations — when they have a choice. For example, a parent's PO when he is in conflict with his child over breaking curfew will usually differ dramatically from his PO when arguing with a policeman or a judge over a traffic ticket. According to the theory, people can develop the capacity to employ different POs in different situations. Some people, however, develop strong, chronic, rigid orientations that can be inconsistent with a particular situation, which can cause them to respond inappropriately or ineffectively (imagine addressing a police officer as you would a petulant child).

Deutsch's theory sketches out the normative dynamic tendencies between certain types of social situations and distinctive types of POs. Deutsch argued that the social relationship cannot by itself determine the PO (nor the resulting behavior) and vice versa, but rather that a fit between POs and social relations tends to develop that will lead to changes in one or both until congruence is achieved. In other words, we will tend to seek out relationships and situations that fit with how we prefer to view the world (e.g., dominant–submissive relations). But when we are unable to escape a situation (such as someone in prison or a student in elementary school), we will tend to develop an orientation and strategies that fit in that situation. At times, POs will be easier to change, at other times social relations will be more easily altered.

Several aspects of this general model of social relations are relevant to our current purposes. First, Deutsch's empirical research (Wish, Deutsch, and Kaplan 1976), which is convergent with others (Thibaut and Kelley 1959; Triandis 1972; Kelley and Thibaut 1978; Kelley 1979, 1984, 1991), helped identify dimensions for the study of social relations that provide a sturdy foundation for our model. Consistent with the conflict models we summarized earlier, these include type of interdependence (cooperative–competitive), power (in)equality, and degrees of dependence (relational importance).

Second, Deutsch's (1982, 1985, 2007, 2011) elaboration of the theory of social relations and POs provides us with a framework that characterizes a person's social behavior as the dynamic result of her psychological variables interacting with the different social conditions (combinations of the basic dimensions) that she confronts, and this behavior is to some degree propelled by a basic need for congruence between her orientation and these social conditions. Finally, Deutsch's theory highlights the importance in social relations of adaptivity: the capacity to employ different types of POs as they are necessary in different types of situations.

Deutsch's theory of social relations and POs offers a promising sketch of some basic ideas relevant to social conflict. It currently, however, requires more rigorous theoretical development, has yet to be explicitly developed with regard to conflict dynamics, and lacks empirical support (Deutsch 2007). The present model addresses the first two issues and sets an agenda for the third.

In recent years, social psychology has moved toward the construction of theoretical models that situate individual cognition and behavior in the context of specific social and cultural forces (see Jost and Kruglanski 2002 for a summary). Our model builds on Deutsch's (1982, 1985, 2007, 2011) theory of social relations and POs as well as on other influential models of social relations and conflict (Boulding 1956, 1959; Thibaut and Kelley 1959; Triandis 1972; Kelley and Thibaut 1978; Rouhana and Fiske 1995; Van Lange et al. 1997; Alexander, Brewer, and Herrmann 1999) to develop such a situated model of conflict. These models all differ to some degree in their characterization of the most basic dimensions of social relations, but they emphasize the primacy of three:

1. the nature of the parties' goal interdependence;

2. the relative distribution of power among the parties; and

3. the degrees of total goal interdependence and relational importance.

In other words, the model suggests that when people are faced with a conflict, they have three primary considerations:

1. Is the other party with me or against me or some combination of both?

2. What is my power relative to the other party (high, equal, or low)?

3. To what extent are my goals linked to the other party's goals, and therefore, how important is this conflict and relationship to me?

Thus, our model incorporates these dimensions and integrates them into a basic three‐dimensional model of conflict in social relations (see Figure One).⁵ Each is specified below.

Following Lewin (1936, 1948), Deutsch (1973), Kelly (Thibaut and Kelley 1959; Kelley and Thibaut 1978), Thomas Schelling (1960), and many others (Boulding 1956, 1959; Triandis 1972; Pruitt and Rubin 1986; Rouhana and Fiske 1995; Van Lange et al. 1997; Alexander, Brewer, and Hermann 1999; Pruitt and Kim 2004; Johnson and Johnson 2005; Tjosvold, Johnson, and Johnson 2006; De Dreu et al. 2007), the first dimension of our model of social conflict is the nature of goal interdependence in social relationships. It constitutes the x‐axis of the model, with pure positive forms of goal interdependence (where all goals between parties in conflict are complementary at the extreme left of the x‐axis), pure negative interdependence (where all goals are contradictory at the extreme right of the x‐axis), and mixed‐motive types (combinations of both cooperative and competitive goals) along the middle of the x‐axis. Thus, conflicts of a purely cooperative nature (such as between members of a humanitarian organization who share the goal of trying to reduce suffering in a developing country but who may have different ideas about the best ways of achieving those goals) are located on the far left of the dimension. Those of a more competitive, distributive nature (conflicts over land, political office, or other scarce resources) in which one party's success necessitates the other's failure are located on the far right. Along this dimension we also have various forms of mixed‐motive interdependence, from those weighted more positively (on the left side of the continuum) to those weighted more negatively (on the right side of the continuum), with relatively balanced forms of positive and negative interdependence located near the middle.⁶

The mix of goal interdependence in social relationships can be influenced by a wide variety of factors at any point in time, including the parties' history of relations with one another, the quality of their relationship, their degree of similarity, the demand for scarce resources, the nature of task structures and reward systems, the parties' abilities to communicate effectively with each other, and each party's relations with and responsibilities to other relevant parties (see Johnson and Johnson 2005).

Building on Follett (1973), Kenneth Boulding (1990), Kelley (1997), and others (Gurr 1970, 2000; Rubin and Brown 1975; Tjosvold 1981, 1991; Blalock 1989; Rouhana and Fiske 1995; Alexander, Brewer, and Hermann 1999; Zartman and Rubin 2002; Van Kleef et al. 2006; Magee, Galinsky, and Gruenfeld 2007), the second dimension of our model represents the role of the relative distribution of power between parties. Power has been conceptualized in myriad ways (Fiske and Berdahl 2007) and is perhaps best understood as complex, fluid, multidimensional, and multidirectional forces (Lewin 1944; Foucault 1980; Fincham 1992; Coleman and Voronov 2003) that affect parties' ability to “make things happen” (Follett 1973: 99). For this model, we focus on parties' perceptions of relative power and define it specifically as the relative degree to which the parties believe they can affect each other's goals and outcomes (Thibaut and Kelley 1959; Depret and Fiske 1993).

Relative distribution of power constitutes the y‐axis of the model, with pure types of unequal distribution of power (A over B) at the top of the y‐axis, the opposite types of unequal distribution of power (B over A) at the bottom of the axis, and various types of relatively equal distributions of power represented along the middle of the y‐axis. Parties' relative degrees of influence can be affected by a wide variety of factors, including formal hierarchical power, wealth, social status, access to weapons, age, location in social networks, expertise, access to information, health and physical strength, endurance, number and influence of allies, charisma, and so on.

The top of the y‐axis of the model represents situations in which A has relatively high power and therefore unilateral capacities to affect the goals and outcomes of B (e.g., typical CEO–employee relations at work), and the bottom of the axis represents situations in which A is in relatively low power and B has unilateral capacities over A (e.g., a situation in which the employee comes to have the upper hand over the CEO). Again, along this continuum, we have various forms of relatively equal, bidirectional power distributed between A and B, with a situation in which the parties are closer to parity falling at the center of the dimension.

Our third dimension extends the work of George Homans (1958), Richard Emerson (1962), Peter Blau (1964), Kelley (Thibaut and Kelley 1959; Kelley and Thibaut 1978), and others who explored the role of power dependence (Ng 1980; Mannix, Thompson, and Bazerman 1989; Gerhart and Rynes 1991; Mannix 1993; Pinkley, Neale, and Bennett 1994; Kim 1997; Kim and Fragale 2005) as well as the work of those in the dual‐concern tradition (Thomas 1976; Pruitt and Rubin 1986; Pruitt and Kim 2004). This dimension involves the degree of importance of the relationship between the parties as well as the degree of their goal linkage (total goal interdependence) in a situation of conflict. It constitutes the z‐axis of the model, with high degrees of goal interdependence between the parties in conflict located at the front of the z‐axis (strong goal linkages and/or high proportions of linked goals), low degrees of interdependence located at the rear of the z‐axis (no, few, or weak goal linkages), and moderate degrees of goal interdependence located along the middle of the z‐axis. Relationships can vary in terms of the number of interdependent goals between the parties, the importance or strength of these goals, and the degree to which the links between goals are temporary or stable. This dimension represents the total degree of goal interdependence between the parties, both in general in the context of their overall relationship and in specific in relation to the goals involved in the specific conflict. Higher degrees of total goal interdependence will lead to greater levels of relational importance than lower degrees of goal interdependence.

These three basic parameters constitute the core of our situated model. They can be used to represent the basic relational context in which people experience conflict. Thus, conflicts that appear to be similar because parties' perceptions of incompatible activities are the same (you and I desire the same office space) may be experienced in fundamentally different ways depending on the three parameters in the model (our mix of cooperative or competitive goals, who has more power and how great the disparity is, and how important we each consider our relationship to be). The task, then, is to consider how different values for each of these parameters combine to produce qualitatively different social conflict experiences and outcomes.

We propose that when people perceive they are in conflict, three basic features of social relations (nature of interdependence, relative distribution of power, and degree of total interdependence) combine to affect qualitative differences in their experiences of the conflict. These differences can elicit distinct conflict orientations, which comprise disputants' perceptions, emotions, values, and behaviors in the conflict (see Proposition One). Further, we argue that social conflicts that arise in more task‐oriented situations will tend to induce more efficient, economically rational orientations, whereas conflicts that are more social‐emotional and less instrumental will tend to induce more socially rational and less economically efficient orientations (see Proposition Two).

When parties' experiences of their place within these different types of situations persist over time (e.g., when someone is stuck in a low‐power situation in a competitive conflict with his or her boss), it will tend to strengthen that orientation, which can become chronic (see Proposition Three). When orientations become chronic, they become more automatic and thus may be employed even when they are inconsistent (ill‐fitting) with the demands of the particular situation. Generally, more adaptive orientations to conflict — those that allow for parties to change their orientations and behaviors to satisfy goals in a manner more congruent with the actual situation — will lead to greater general satisfaction with conflict processes and outcomes over time (see Proposition Four).

We further argue that when social relations are characterized by persistent conflict between parties, they often internalize synchronized patterns for responding to the conflict (see Proposition Five). Finally, differences between cultural groups in attitudes toward interdependence and the equality–inequality of power distribution can affect the relative strength and availability of parties' conflict orientations for those parameters (see Proposition Six). Below, we elaborate on the components of the model and its propositions.

The three dimensions of social relations described above constitute a basic framework for the study of social conflict (see Figure One) that can be seen as a “stimulus field” (Kelley 1997) for conflict: a perceiver's representation of his or her external world or environment. Harold Kelley (1997: 143) characterized a stimulus field as “something that is neither objective nor subjective — or is both, if you like. The stimulus portion of the term carries an allusion to something objective but available to the senses — something ‘out there’ to be labeled and thought about and that may or may not be in view (in the visual field or in the mind) at any particular time. The field portion of the term implies a map‐like topography that spreads out before us and that, to varying degrees and on different occasions, is visible to us and available to our thoughts.” The model can thus be used to characterize how a disputant perceives and contextualizes any given conflict — within the context of the relationship in which it is occurring — at any point in time.

Thus, the stimulus field is a metaphor that represents how a disputant perceives the relational context of conflict in terms of:

1. differences between the parties' common and competing goals;

2. differences between the amount of power that the disputants have to affect each other's goals and outcomes; and

3. differences in how each party perceives the links between the parties' goals and the general importance of the conflict and the relationship.

Differences on the three dimensions may be situational (such as objective differences in resources or formal authority between parties that affect relative power) and/or they may reflect differences between each party's chronic PO to conflict (such as a tendency to view all conflicts as threats). Whether an individual's stimulus field is determined primarily by the particulars of the current situation, by the broader relational context, or by his or her predisposition to view the situation in a particular manner (chronic orientation) is predicated on the perceived relative strength of each of these factors (Deutsch 1982; Kelley 1997; see Mischel 1977).

For theoretical purposes, it would be useful to begin our discussion of the stimulus field in terms of its most extreme regions (pure competitive–pure cooperative, high power–low power, high interdependence–low interdependence, 2 × 2 × 2 = 8). Theory would predict, however, and research has shown that under conditions of very low degrees of goal interdependence (no, few, or weak goal linkages between parties), the importance of conflict engagement tends to dissipate (Deutsch 1973; Kelley 1997) and disputants' conflict orientations and behaviors become more uniform (Kugler and Coleman 2010). Thus, the four regions of the stimulus field operating under conditions of low interdependence tend to collapse to one (see Figure One). In other words, if our goals are relatively unconnected and our lives mostly independent from one another, then the relative power and type of interdependence between us loses relevance and, only slightly perturbed, we will tend to continue on our separate ways. Therefore, our current discussion focuses primarily on the five more extreme regions of the stimulus field:

• Region One: the party has high relative power in this situation, the goals are shared or complementary, and the parties have a highly interdependent relationship;

• Region Two: the party has high relative power, the goals are incompatible, and both parties are highly interdependent;

• Region Three: the disputant is in the low‐power position, the goals are not contradictory, and their relationship is highly interdependent;

• Region Four: the disputant is in the low‐power position, both parties' goals are contradictory, and the parties are highly interdependent; and

• Region Five: conflicts in which parties are not highly interdependent.⁷

We propose that parties will employ distinct POs to conflict according to which of these five regions of the conflict stimulus field they find themselves in. We further propose that these orientations help determine which perceptions, values, emotions, and behaviors the disputant will find to be relevant in a given conflict (see Deutsch 1982, 1985, 2007, 2012; Kelley 1997; Van Lange et al. 1997). In other words, the different regions of the stimulus field will tend to evoke very different conflict orientations, which influence:

1. how conflicts are perceived (as mutual problems or win–lose challenges or some combination of both);

2. how it feels to be in the situation (relatively comfortable versus anxiety provoking);

3. what is likely to be valued in the situation (solving problems and sharing benefits with other parties versus conquering them); and

4. how to best respond to the conflict and achieve one's goals (through respectful dialogue and problem solving versus forceful domination or submission to power; see Figure Two).

Thus, our first proposition (Proposition One) is that parties will use distinct conflict orientations depending on which regions of the conflict stimulus field they find themselves in. These orientations temporarily affect parties' perceptions, emotions, values, and behavioral response options in conflict.

Of course, the specific nature of the orientation associated with each region (conflict situation) will depend on cultural, social, developmental, and personality factors. The parameters that place the conflict in a particular region of our metaphorical map do not determine specific responses to conflict, but rather tend to orient disputants much like improvisational scripts, framing the disputants' general response to the conflict. The impact of these orientations can be exclusionary because they can influence which behaviors the disputant sees as not appropriate to a particular situation, as much or more so than which responses he or she would deem acceptable.

Accordingly (see Figure Two), we could expect Region One (high‐power position, congruent goals, high level of relationship interdependence) to encourage a benevolent orientation to conflict: a cooperative orientation and a willingness to work toward mutually beneficial outcomes and engage in constructive processes of dialogue, persuasion, and joint problem solving.

Mary Parker Follett (1973) suggested that people with this orientation employ a type of joint, coactive, noncoercive power, while Albert Bandura (1999) described it as a collective agency, and others have seen it as cooperative or positive power (Deutsch 1973; Tjosvold 1991). David McClelland found this type of orientation to involve “socialized power,” defining it developmentally as “the most advanced stage of expressing the power drive in which the self drops out as a source of power and a person sees himself as an instrument of a higher authority which moves him to try to influence or serve others” (1975: 20). Recent empirical studies provide evidence of this orientation in leaders, finding that under certain conditions high power can promote more other‐oriented behavior, social responsibility, and stewardship (see Handgraaf et al. 2008; Wade‐Benzoni et al. 2008). In addition, several studies in the organizational realm have demonstrated the strong effects of cooperative interdependence in fostering more constructive conflict dynamics between managers and employees (Tjosvold 1981, 1985a, 1985b; Tjosvold, Johnson, and Johnson 1984; Tjosvold, Coleman, and Sun 2003; Coleman 2004b). Cooperative goals, when compared with competitive and independent goals, were found to induce “higher expectations of assistance, more assistance, greater support, more persuasion and less coercion and more trusting and friendly attitudes” between superiors and subordinates in conflict (Tjosvold 1997: 297).

In contrast in Region Two (relatively high power, incompatible goals, high relationship interdependence), we could expect dominance: a more domineering, exploitive orientation to conflict in which parties primarily value winning (by defeating their opponents), maximizing their own outcomes, and using tactics of force, control, deceit, and oppression to achieve their aspirations. This orientation is consistent with the popular definition of power in social science as “an ability to get another person to do something that he or she would not otherwise have done” (Dahl 1957: 158). Dominance‐displaying responses are commonplace and typically the first course of action taken by many people in conflict, particularly by men (Terhune 1968; Sidanius and Pratto 1999), by those in high‐power positions (Rubin and Brown 1975; Sidanius and Pratto 1999; Gurr 2000; Zartman and Rubin 2002; Kim, Pinkley, and Fragale 2005; Galinsky et al. 2008; Magee and Galinsky 2008), and by people in organizations in general (Ury, Brett, and Goldberg 1988; Costantino and Merchant 1996).

In Region Three (low‐power position, complementary goals, high interdependence) we would expect parties to be orientated toward support: an approach to conflict in which parties have come to value the constructive leadership, decisions, and expected benefits bestowed by those in (cooperative) higher‐power positions, and seek to follow and assist them (see McClelland 1975). “Support,” as used here, typically involves obtaining assistance and support from others, often through a dependence relationship, to achieve one's goals. William Zartman and Jeffrey Rubin (2002) documented a variety of support tactics used by low‐power parties in international negotiations for “borrowing” power from other parties, the context, the conflict processes, and third parties (see also Salacuse 1999, 2002). They argue that disputants should employ these methods skillfully to elicit minimal resistance from higher‐power parties (see also Deutsch 1973). This orientation can be effective for achieving goals when in the lower‐power position, but may be particularly problematic for individuals who have insecure attachment adaptations (see Rusbult and Van Lange 2003).

We would expect parties whose orientations are represented by Region Four (low relative power, incompatible goals, high interdependence) of the model to be oriented toward appeasement, characterized by feelings of victimization and an attitude of forbearance in which parties seek to avoid harm, seek opportunities to escape or resist (if conditions change), and ingratiate themselves, suppress their own needs, and engage in self‐blame, denial, and, only when possible, sabotage. They would typically tolerate and placate a dominating high‐power disputant in order to avoid making matters worse. This kind of prolonged powerlessness can have dire physical and psychological effects (Sashkin 1984) and can lead to rigidity, hostility toward those in lower‐power positions, and, ultimately, irrationality and violence (Kanter 1977). This orientation has been somewhat understudied in relation to conflict, but has been found to be quite common in situations of ongoing competitive conflict in organizations in which disputants are in positions of relatively low power (Coleman et al. 2010). For example, in a study conducted in Japan (Ogasawara 1988), executive secretaries who worked for mean or incompetent bosses were found to have developed an extensive array of passive–aggressive tactics for seeking revenge on their superiors in a manner that allowed them to maintain deniability and keep their jobs.

We expect parties who fall into Region Five (low degrees of interdependence where relative power and type of interdependence have thus become irrelevant) to be oriented toward autonomy, to seek to exit the conflictual relationship, and to meet their needs through alternative means (ignore the conflict, withdraw from the relationship, turn to negotiation alternatives, etc.; see McClelland 1975). They simply seek to establish autonomy in order to achieve their goals unilaterally. Scholars have referred to this approach as having “power to” or “power from”— the disputant has enough power to achieve his or her objectives without being unduly constrained by someone or something else (see Coleman 2000). This is the orientation to conflict characteristic of power‐dependence theory, in which power is derived from having a good BATNA or through other means of reducing one's dependence (or increasing the other's dependence) in a particular relationship (Emerson 1962; Pinkley, Neale, and Bennett 1994; Kim, Pinkley, and Fragale 2005).

Finally, we predict that in contrast to the five outlying regions of the stimulus field that represent more extreme orientations (Region One through Region Five), those who hold less extreme orientations, which Lewin (1944) referred to as “cognitively unstructured situations” (e.g., characterized by equal power, mixed motives, and moderate party interdependence), will evidence a much broader range of values, feelings, and responses and will display no clear cognitive‐behavioral syndromes other than those dictated by strong local conditions (such as a strong work group culture) or chronic individual differences.

Recent studies testing this model in the context of organizational conflict (Coleman et al. 2010, 2011) have provided support for these distinct region‐orientation associations. The findings from focus group, critical incident, correlational, and experimental studies have found that when participants were presented with the same conflict (in terms of incompatible goals and issues), they described markedly different experiences, perceptions, emotions, values, and behavioral intentions across the five regional conditions. When faced with a Region One scenario (high power over their counterpart, compatible goals, highly interdependent relationship), participants described having a more benevolent orientation to conflict than they and others did when presented with conflicts in the other four regions. In a Region One scenario, participants said they valued taking responsibility for the problem and listening to the other party, and expressed genuine concern for their lower‐power counterpart.

In contrast, in Region Two scenarios (high power over the counterpart, contradictory goals, highly interdependent relations), participants were more likely to report anger and to take a threatening and confrontational approach to the other party, with heightened concern for their own authority and goals (dominance). In Region Three scenarios (parties had low relative power, goals were complementary, and relationships were highly interdependent) parties displayed more support than they did in the other regions, seeking respectful clarification of roles and responsibilities, working harder but reporting anxiety and confusion about the conflict situations. This was in contrast to the reactions of parties faced with Region Four scenarios (low relative power, contradictory goals, high relationship interdependence), who experienced higher levels of stress, fear, and anger and reported more tendencies to either do nothing in the sitution, vent only to peers, try to barter, or, if necessary, resort to sabotage, blackmail, or whistle blowing.

Parties in Region Five scenarios (low interdependence) had a less intense experience of the conflict and preferred to simply move on or exit the conflict (autonomy). Finally, mixed, nonextreme scenarios (parties had equal power, goal compatibility and incompatibility, and moderate levels of relationship interdependence) offered by far the broadest range of behaviors, feelings, and values and evidenced no coherent syndromes.

Myron Wish, Morton Deutsch, and Steven Kaplan (1976) identified another fundamental dimension of interpersonal relations: whether relationships are primarily task oriented or social‐emotional (Parsons 1951). In general, task‐oriented relationships have been characterized as more impersonal, objective, universalistic, and emotionally detached (Deutsch 1985). In contrast, social‐emotional relationships have been characterized as more personal, subjective, particularistic, and emotionally involved. This contrast has been an important research topic in the fields of leadership and management (Fleishman 1953; Bales 1958; Stodgill, Goode, and Day 1962; Blake and Mouton 1964) and conflict and negotiations (e.g., Guetzkow and Gyr 1954; Cosier and Rose 1977; Jehn 1997; De Dreu and Weingart 2003).

For the purposes of the current model, we are interested in how this dimension of social relations interacts with the dimensions in our basic conflict stimulus field to affect conflict orientations and responses. Following Paul Diesing (1962) and Deutsch (1985), we propose that differences along this dimension will significantly affect what disputants see as reasonable or rational orientations and behaviors in conflict. Situations in which people are primarily concerned with effective task accomplishment will tend to induce more economically rational orientations in which efficiency, utility, and maximization of goals are highly valued, typically at the expense of such social concerns as inclusion and solidarity. Such situations will tend to induce more efficient conflict orientations like dominance, appeasement, and autonomy. In contrast, situations in which the parties are more interested in social relations will tend to induce orientations and behaviors that are more socially rational such as benevolence and support, often at the expense of efficiency and utilitarian goal achievement.

Thus, our second proposition (Proposition Two): social conflicts that arise in relationships that are primarily task oriented will tend to induce more economically rational conflict orientations (dominance, appeasement, and autonomy), whereas conflicts that occur in primarily socially oriented relationships will tend to induce more socially rational conflict orientations (benevolence and support).

Recent research on relational accommodation in negotiations offers some preliminary support for this proposition (Curhan et al. 2008). This study showed that when negotiating dyads had highly relational goals, their negotiations produced less efficient economic outcomes but greater relational capital than did those dyads whose goals were more task focused. The direct relationship between task‐social relations and our model's conflict orientations has yet to be investigated, however.

The relative strength and salience of different conflict orientations will be influenced by combinations of social (local rules, roles, and customs), developmental (e.g., stages of moral, socioemotional, and cognitive development), and individual (authoritarianism, social dominance orientation, attachment style, etc.) factors (see McClelland 1975; Deutsch 1985; Rusbult and Van Lange 2003). Fundamentally, however, we suggest that it is the degree and quality of parties' direct and indirect (social modeling) experiences with each orientation that are most likely to determine the particular strength of that orientation and thus the character and contours of the various orientations available to parties in different conflict situations (see Higgins 1996). If individuals remain in a given situation for extended periods of time and achieve their goals through tactics associated with that particular conflict orientation (benevolence, dominance, support, appeasement, or autonomy), we could expect that orientation to become more pronounced over time. For instance, if someone were subjected to the whims of a brutal, humiliating boss for a long time, we would expect that person to develop a strong chronic orientation toward appeasement and relatively weak ones for benevolence or dominance.

Once an individual has developed a strong propensity for a particular conflict orientation (e.g., dominance), it can become difficult to change, even when it fails to satisfy his or her goals, the intensity of the conflict dissipates, or social conditions change (see Coleman et al. 2007). When this occurs, the orientation is said to have become chronic (see Higgins and King 1981; Andersen et al. 2007). More chronic orientations will often become automatic and may be employed even when they are inappropriate for a particular situation (Barge 1996). Thus, the transition to a new job with a more benevolent boss could be psychologically and behaviorally challenging for the employee described above, and she or he will need some time to make an adjustment. Ultimately, a person will adapt his conflict orientation according to the feasibility of the response associated with that orientation (Do I have the capacity to act in such a manner and what will the consequences be?) unless the orientation has become excessively chronic (see Pruitt and Kim 2004).

Thus, our third proposition (Proposition Three) is that when a party finds himself or herself in a particular region of the conflict stimulus field over time, he or she will develop a strong (chronic) PO consistent with that region.

Research on high‐power/low‐power party dynamics provides strong evidence that conflict orientations can become chronic. Joe Magee and Adam Galinsky (2008) have cited a litany of research detailing the psychosocial transformations that take place in organizations that can encourage more powerful parties to develop chronic dominance orientations and to strive to retain and acquire power. Case study research on state‐level international negotiations also provides strong support for the view that high‐power parties often become comfortable with dominance orientations and find it difficult to employ other strategies when power shifts and conditions change, and that low‐power parties, too, can become accustomed to their role (Zartman and Rubin 2002). In addition, decades of research using the prisoner's dilemma game show strong consistencies in people's pro‐self versus pro‐social orientations across a variety of social manipulations (Van Lange et al. 2007).

Various approaches to the study of POs have stressed that even though different orientations may be useful in particular situations, problems typically arise for people when their orientation (such as support or dominance) becomes fixed or when an individual's chronic orientation(s) is inappropriate for a specific situation (McClelland 1975; Deutsch 1985; Kelley 1997). From this perspective, psychosocial flexibility and the ability to identify and respond to relevant changes in the environment are critical, particularly over time or when conflict situations are in flux. For instance, research has shown that displaying higher levels of integrative complexity (which requires cognitive flexibility in problem identification) is associated with more effective and adaptive behavior, particularly in situations that are confusing or ambiguous (Suedfeld and Piedrahita 1984; Winter 2007).

People often hold strong chronic preferences, however, for particular orientations to conflict and find it emotionally distressing when situations require a different approach (McClelland 1975; Rusbult and Van Lange 2003). We stress that each of the different orientations outlined in our model has its particular utilities, benefits, costs, and consequences, depending on the psychological makeup of the parties, the orientation of other parties, and the nature of the situations faced. In fact, all of the orientations, when chronic, can have negative consequences (or pathologies) (see Deutsch 1985; Johnson and Johnson 2005 for a discussion of the pathologies associated with chronic cooperative orientations). Ultimately, as Deutsch (1982, 1985, 2007, 2011), McClelland (1975), and Zartman and Rubin (2002) have suggested, what is particularly useful in evolving situations of conflict is the capacity to adapt: to move freely between various orientations and employ their related strategies and tactics in a manner that helps to achieve one's short‐ and long‐term goals. Thus, we propose that more adaptive orientations to conflict are likely to lead to higher levels of goal attainment and greater general satisfaction with conflict processes and outcomes over time.⁸

Thus, our fourth proposition (Proposition Four) is that more adaptive orientations to conflict (those that allow for movement between different POs and behavioral strategies to satisfy goals in a manner congruent with the demands of the situation) will lead to higher levels of goal attainment and greater general satisfaction with conflict processes, relationships, and outcomes over time.

Research has offered support for the positive effects of adaptivity in conflict. Case‐based research on interstate negotiations found that parties tended to be more effective in negotiations to the extent that they were able to adjust their orientations and behavior to the relative (and relevant) power of the other side (Zartman and Rubin 2002). In a correlational study (Coleman, Kugler, and Mitchinson 2009), investigators found that more adaptive individuals (individuals who saw utility in employing all five orientations when necessary) had greater levels of satisfaction with conflicts in general than less adaptive individuals. This study also found that more adaptive individuals learned more from conflicts and had more global perspectives on conflict, focusing more on both long‐term and short‐term goals than less‐adaptive individuals. A third study, conducted through the collection of critical incidents of actual conflicts in participants' lives, found that people who were able to employ orientations and behaviors that were more congruent with the situation (more appropriate) expressed significantly more satisfaction with the conflict processes and outcomes, their relationships with the other parties, and their own behavior in those conflicts (Coleman and Kugler 2011).

Relationships characterized by persistent conflicts between parties that remain relatively stable for long periods of time (such as those between many labor and management negotiators or between many romantic partners) will tend to result in the internalization of strong orientations, expectations, and scripts for interaction dynamics between the parties (Zartman and Rubin 2002; Deutsch 2006b). In other words, just as there is a tendency toward consistency and coherence between different cognitions psychologically (Festinger 1957; Heider 1958) and between specific types of social relations and associated POs (Deutsch 1982; Kelley 1997), we suggest that there is also a tendency for consistency and coordination in the longer‐term conflict dynamics between people (see Vallacher and Nowak 1994; Nowak, Vallacher, and Zochowski 2002; Tiedens and Fragale 2003; Vallacher and Nowak 2006).

Social coordination between interacting parties involves the synchronization of partners' respective dynamics (Nowak, Vallacher, and Zochowski 2002). In fact, research has shown that when people synchronize their physical movements (marching bands, military units, dancers, singing groups, etc.), it tends to induce other types of coordination and group cooperation (Wiltermuth and Heath 2009). When conflicts persist and parties engage in repeated interactions under similar conditions, we can expect to see coordinated interpersonal dynamics emerge. These coordinated relational dynamics may become chronic so that virtually any encounter between parties, even two encounters of a very different nature (e.g., negotiating a contract versus attending a funeral of a mutual friend), results in the resurgence of the same social dynamic.

Thus, our fifth proposition (Proposition Five) is that social relations characterized by protracted conflict between parties will lead to the internalization of synchronized patterns for responding to the conflict. In their analysis of international conflicts, Zartman and Rubin (2002) found evidence of this type of coordination dynamic between high‐ and low‐power negotiators. When high‐ and low‐power relationships were ongoing and well established (e.g., the United States and Mexico), the parties tended to know their relative power roles and play them in a complementary manner. They write, “In asymmetrical negotiation, strong and weak work together, organizing themselves around predictable moves and responses” (Zartman and Rubin 2002: 285).

Even if important changes in the relationship occur (such as power shifts), the strength of the relational orientation may be enough to maintain the original dynamic or at least to delay significant changes in the dynamic (see Jost, Banaji, and Nosek 2004; Vallacher et al. 2010). In effect, once a conflict has become internalized, the component interpersonal behaviors unfold automatically. Different relational orientations, however, will vary in terms of their relative stability. In general, social conflicts that arise in formal role relationships will tend to be more stable than those that occur in more informal role relationships (Tajfel 1981; Galinsky et al. 2008).

Finally, we propose that cultural groups and subgroups (ethnic, religious, national, organizational, etc.) will differ in terms of the relative strength and availability of their orientations for interdependent versus independent conflict orientations, egalitarian versus unequal power orientations, and cooperative versus competitive conflict orientations. These differences will be commensurate with the relevant differences in cultural group value orientations identified previously in research on independent and interdependent self‐construal (Markus and Kitayama 1991), degrees of interdependence (Triandis and Gelfand 1998), types of interdependence (Tjosvold 2008), and power orientations (Hofstede 1980; Triandis and Gelfand 1998; Sidanius and Pratto 2001).

Thus, we could expect to see cross‐cultural group differences in the general availability, strength, and accessibility of the different conflict orientations that are consistent with the dominant value orientations of the groups. In other words, members of cultures identified as high power‐distance cultures will have a stronger preference for orientations in high‐power positions (benevolence and dominance) and low‐power positions (support and appeasement) than members of lower power‐distance cultures.

Members of cultures or groups who place a higher value on cooperative interdependence in relationships would be predicted to display stronger propensities for benevolence and support orientations than groups measured as more competitively interdependent, who would display stronger orientations for dominance and appeasement. In addition, members of groups that place a higher value on highly interdependent (collectivist) relationships and who define themselves in more interdependent terms would be more likely to display strong orientations characterized by high interdependence than groups measured as valuing low interdependence (independence), who would be predicted to generally display more autonomous conflict orientations.

Thus, our sixth proposition (Proposition Six): cultural group differences in value orientations toward different types and degrees of interdependence and toward the equality–inequality of power distribution will affect the relative strength and availability of conflict orientations for corresponding regions of the conflict stimulus field.

A recent study comparing American and Korean managers (Kim et al. 2011) provides qualified support for Proposition Six. The study found that Korean workers who reported employing more cooperative strategies in work conflict (benevolence and support) were significantly more satisfied with both conflict and work relationships over the long term and that being cooperative in conflict was even more important than being adaptive to changes in social conditions. Researchers reported the exact opposite results for the sample of American managers. American managers reported being more satisfied when they employed more adaptive, situationally congruent strategies, and adaptivity was found to trump cooperation in terms of work conflict satisfaction. This proposition requires much additional empirical investigation and validation.

The model we have presented builds on some essential features of social relations identified by prior research and theorizes how they interact to influence constructive and destructive dynamics in conflict. Our model helps to integrate disparate and even contradictory findings from decades of prior research to enhance our understanding of how three key variables — relative power, interdependence, and relational importance — affect conflict dynamics. The value of this model is not the identification of new factors and variables. Rather, our model shows how an analysis of those factors deemed the most essential is sufficient to capture the complexity of conflict in a wide range of interpersonal relations.

The situated model also highlights the importance of adaptivity in constructive conflict resolution. Instead of advocating a set of specific predispositions or conditions designed to promote positive conflict processes, our model stresses the necessity of adapting flexibly to new situations in a manner that helps to achieve important goals. Conflicts can be constructively managed when the disputants are able to adjust their orientations, strategies, and tactics as the evolving situation requires.

Research has found that although many negotiators and leaders tend to get stuck in one approach to negotiating conflict (often domination), our more effective leaders and negotiators are more nimble (Hooijberg and Quinn 1992; Zartman and Rubin 2002; Lawrence, Lenk, and Quinn 2009). They read situations carefully, consider their short‐ and longer‐term objectives, and then employ a variety of different strategies in order to increase their chances of success (Dörner 1994).

Thus, according to our model, effective leaders and negotiators should develop their capacities and skills for:

• Dominance— employing power, information, and authority to demand, incentivize, threaten, coerce, expose, and publicly shame opponents when absolutely necessary;

• Benevolence— modeling exemplary, collaborative, win–win leadership by listening carefully to the needs and concerns of opponents, finding common ground on the priority objectives, and uniting parties around a common vision and purpose;

• Support— reaching out to the other side, allies, and other stakeholders to persuade, seduce, barter, beg, and ingratiate in order to mobilize them and secure their support;

• Appeasement— learning to tolerate attacks, inflammatory rhetoric, and hyperbole of opponents in the short‐term, give in to them on their key demands, suck up to them as much as possible, and quietly lay in wait for conditions to change and opportunities to present themselves to blithely sabotage them and derail their agenda; and

• Autonomy— spending time and energy developing a good Plan B, in which it is still possible to achieve principal goals unilaterally.

This was the approach to conflict modeled and described by Nelson Mandela. In his autobiography, Long Walk to Freedom, Mandela (1995) described how he adapted his approach to conflict according to changes in his circumstances. The book details Mandela's long journey through low power as a young native black African (support orientation in his dealings with a benevolent father), his movement toward autonomy via education and legal training, and then his use of various forms of dominance (legal power, nonviolent resistance, violent militancy) during the years of the African National Congresses' antiapartheid struggle. In time, he was forced to revert back to low‐power and appeasement strategies while in prison for twenty‐seven years but eventually rose to equal power (during negotiations with the apartheid government) and then eventually to high‐power benevolence when elected president and beyond. Mandela's default orientation was mostly collaborative (he claims to have always sought peaceful negotiations with the Afrikaner government), but he was very capable of employing other orientations and tactics. He learned and adapted as conditions changed, but kept his sights on his long‐term goals (a fair and free South Africa).

A few caveats are in order. First, the parameters we emphasized in this article hardly reflect all the influences on conflict scenarios. Conflict, like any type of social relation, is a complex phenomenon in which a wide variety of variables play different roles at different times. Our concern was not to paint a complete picture, but rather to offer a blueprint that can serve as scaffolding for subsequent research that may identify other fundamental parameters relevant to conflict. Second, the model is intended to incorporate the insights of prior research and to conceptualize how the resultant model represents important features of constructive and destructive conflict. The model is thus both heuristic and integrative. But the value of the model will be ultimately judged by its verification in subsequent research. A few of the key propositions we presented have been translated into verifiable hypotheses and tested employing various methodologies, including focus groups, surveys, and experimentation, but more work is needed.

The situated model offers a preliminary perspective on some of the more basic dynamics of social conflict, such as the way in which the three parameters of the model interact to affect conflict orientations, the importance of adaptivity to long‐term conflict management and satisfaction, and how protracted conflicts can lead to the internalization of relational orientations that can resist change. As such, the model provides a solid step forward in our understanding of conflict dynamics in the broader context of social relations. But much work lies ahead to better elaborate on and specify how the parameters of our model operate to affect conflict dynamics over time. The dominant focus in social relations research on main effects and short‐term effects of variables and relative neglect of complex interactions and long‐term temporal trajectories severely limits the utility of our research for informing real‐world applications (Kim, Pinkley, and Fragale 2005; Fiske and Berdahl 2007; De Dreu and Gelfand 2008).

The development of this situated model of conflict provides us with an important platform to investigate the dynamical properties of the model. We suggest that such research should be informed by the many new ideas and tools coming out of complexity science, and in particular the study of dynamical systems (Vallacher and Nowak 1994, 2007; Nowak and Vallacher 1998; Vallacher et al. 2002), which is particularly well suited for advancing and formalizing research on the temporal flow of conflict. Rather than treating thoughts, feelings, and actions in conflict as static phenomena, the dynamical perspective enables researchers to investigate these features of experience as they unfold over time in accordance with specifiable patterns (see Gottman et al. 2002; Coleman 2006; Nowak et al. 2006; Coleman et al. 2007; Liebovitch et al. 2008; Nowak et al. in press).

In addition to the systematic testing of the five propositions presented here, another phase of this research will be to conduct studies of the parameters and their interactions over time and to employ computer simulations to forecast the long‐term consequences of particular configurations of the model's parameters. The findings from such modeling will then need to be compared with data from real‐world conflict dynamics, which will allow us to assess the validity of the assumptions underlying the model. This approach is necessary to investigate the more dynamic aspects of the model, such as the role of initial conditions, nonlinearity, and radical (catastrophic) changes in the initiation, escalation, and constructive management of social conflict. Although several methodologies that are appropriate for addressing these issues exist, the development of new dynamical methodologies may be called for (see Nowak and Vallacher 1994; Gottman et al. 2002 for illustrations). Such approaches present new challenges for those trained in traditional social science methodologies and ripe opportunities for the next generation of conflict scholars.