Definitional clarity is something of an impossible dream in science. Key constructs and hypothetical variables require explicit definitions, yet in many cases more complex terms cannot be defined to the satisfaction of all interested parties (so-called “primitive” terms).
Even the most important word in the field of group dynamics–group–is difficult to define: at least, the field has yet to reach consensus on THE definition to use to define what is and what is not a group.
One particularly intriguing sub-argument within this definitional discussion is the debate about size, and the status of the two-person group–the dyad–within the concept of “group.” Whereas some, such as Moreland and Levine (2012) do not consider dyads to be groups, others (e.g., Forsyth, 2010, Williams, 2010) assume two people can be considered to be a group–albeit, a very small one.
Moreland and Levine point to a number of reasons why dyads do not belong in the larger category of group.
1. Dyads seem to be more ephemeral than larger groups. That is, they both form and dissolve more quickly.
2. Dyads are much more emotionally involving than larger groups. Certain phenomena that are important in dyads (e.g., love, sex, jealousy) are rarer in large groups, if they occur at all.
3. Some phenomena that occur in larger groups cannot occur in dyads, because dyads are too small. These phenomena include group socialization, majority/minority relations, and coalition formation.
4. Even when a phenomenon occurs in both dyads and larger groups, it may operate quite differently in each, and so it is mistaken to draw conclusions about one by studying the other.
Williams (2010) disagrees, and offers a number of reasons why dyads are best considered to be groups, but his core justification is that some of the most fascinating of group processes–social facilitation, inclusion/exclusion, and social loafing–occur in both dyads and groups. Yes: (Williams, 2010, p. 273):
dyads are special groups that show greater or lesser processes than their larger sized counterparts. And yes, sometimes group phenomena are unlikely or impossible in dyads. But by excluding dyads from our consideration of group research, we neglect a great deal of research that informs group processes and dynamics and, just as important, we undersell our vibrancy and impact on the field.
As William’s position implies, what matters most is the use of scientific methods to reliable, law-like generalizations about the behavior of individuals when with others. If the processes that occur in dyads are so unique that they require researchers develop theories specific to dyads, and test them only in dyads, then the “we don’t need no stinkin’ dyads” position (Moreland, 2010, p. 264) is reasonable. However, if basic laws of group processes apply to groups that range in size from 2 to many, and if these theories’ assumptions can be tested in groups that similarly range in size, then dyads are appropriate targets of investigation. Williams, who studies the impact of exclusion on interpersonal relations, also notes that rejecting dyads from the group family seems oddly exclusionary: if dyads aren’t groups, and they aren’t individuals, then who will study them? Is an entire field of dyadism needed?
Groups come into existence when two or more individuals become linked in a relationship of some kind, where relationship implies interdependence and influence. These relationships, or “social ties,” vary considerably in terms of their durability, strength, and intensity. Two objects can be tied together in many ways–with threads, with ropes, string, chain, with plastic tiedowns, a steel weld–and so can group members.
Granovetter, in his classic 1973 analysis “The Strength of Weak Ties,” identifies a number of factors that influence the nature of these ties between group members: duration, emotional intensity, intimacy, and exchange, as well as directedness (reciprocal or one-directional) and valence (positive or negative). These factors generally combine to influence the overall strength of the relationship, where strong relationships are ones that generate (a) enduring, repeated, frequent interactions, (b) strong feelings of attachment and emotional contagion, (c) the exchange of personal information in copious quantities, and (d) interdependency. But, he cleverly notes that weak ties–once that typify the links between people who would describe themselves as “acquaintances” rather than “friends,” also substantially influence group dynamics.
First, weak ties are sources of new information. When individuals are linked through strong ties, they share the same experiences and alliances, and so their knowledge base is shared. If one individual is removed from such a tight-knit group, the impact is relatively negligible, in terms of information, because their knowledge was shared rather than unique. Weak ties, in contrast, connect individuals who hardly know each other, and also have less in common. In consequence, they provide more varied types of information, which Granovetter illustrates with example of job search. He discovered that many people who are seeking employment learn of positions not from their close friends, but from acquaintances. Their close friends can only tell them about jobs they are already aware of, whereas acquaintances know about jobs they had not discovered on their own.
Second, weak ties join groups together. Individuals are members of many different groups, which are joined together by weak rather than strong ties. For example, A, B, C, and D may be all very good friends, but A may also be acquainted with X–who is good friends with Y and Z. The weak tie between A and X links together the two groups of A-D and X-Z.
Barabasi, in his book Linked, provides some backstory to Granovetter’s publication. Granovetter, it seems, submitted his strength-in-weak-ties for publication while he was still a graduate student at Harvard. Granovetter had developed the idea after listening to a lecture by Harrison White, who did early work on social networks. Inspired, Granovetter carried out a field work in Newton, Massachusetts, where he discovered that most people seeking jobs learned about positions from acquaintances rather than friends. But when he sent the paper to American Sociological Review for review and possible publication in August of 1969, he got less than favorable evaluations: one reviewer said it should not be published for an “endless series of reasons” (Barabasi, 2003, p. 42). Granovetter, after recovering from the sting of rejection, recouped and rewrote, and submitted a revised paper to the American Journal of Sociology–where it was published in 1973. Google indicates this paper has been cited 19,134 times!