THIS CHAPTER WILL DISCUSS

1. Theories that are relevant to the concept of power.

2. The "dimensions" of power, and research showing how these dimensions work.

3. Theories concerning how coalitions can form within groups.

4. The ways in which communication can affect people's judgments about power.

5. How beliefs about gender and race influence power, and ways to counteract these beliefs.

In sociological approaches to group interaction, the concept of power is central. Max Weber, a founding father of modern sociology, is a key figure in this area of study. His work has influenced most of the theorists who have examined power in groups, and his definition of power is important for our discussion.

Weber's concept focused on what happens when two people get together. He saw power as emerging out of this social relationship. In his view, power is the probability that one person in the relationship can carry out his or her will despite the other's resistance.

The word "resistance" is important in Weber's definition. Resistance suggests conflict. Weber's concept implies that whenever conflict occurs in a relationship, power becomes an issue. Therefore, conflict and power are closely related.

As an example, we can look at the relationship between Cyndy and Doug. They get together to write a report, but they disagree about the content. Finally, Cyndy decides what they will write, despite Doug's protests. In this situation, according to Weber, Cyndy has power over Doug.

As we can see, Weber's definition of "power" closely relates power and conflict. Thus, examining the idea of power from the relational perspective is profitable. This perspective focuses on the forces and relationships that lead to conflict and power in groups.

Theorists using the structural perspective have also contributed to our knowledge about power. They have examined how the structure of relationships affect the power of group members.

We will begin with a look at these two perspectives, the relational and the structural. We will review their theories and accompanying research. After this, we will describe how coalitions form within groups, and we will examine the theories and research surrounding this topic.

Last, we will discuss how people express power relationships through verbal and nonverbal communication. We will examine how this affects people's judgments of one another, and how gender and race influence the process.

The relational perspective ties in well with Max Weber's definition of power. Weber saw power in terms of conflict. The relational perspective allows theorists to examine how conflict and power relate. In this way, it provides the basic approach to power in both interpersonal and group settings.

One of Kurt Lewin's students, Cartwright (1959), provided a field theoretical approach to power that was based on Lewin's ideas.

As we described in Chapter 3, field theory holds that various forces act on a person and move that person around his or her life-space. Some forces come from within the individual, such as the desire to reach a goal. Other forces are not from within but come from people affecting the person's life-space.

Sometimes these forces conflict. According to Weber's definition, when conflict occurs, power is involved.

Power in Field Theory

For example, Bernie wants to become a rock star, but his mother wants him to be a dentist. These forces conflict, and power comes into play.

Bernie may have some desire to please his mother, but following her wishes keeps him from reaching his own goal. What should Bernie do? According to field theory, Bernie's response dictates who has power in the relationship. To the extent that Bernie considers following his mother's wishes, she has power over him. If Bernie refuses to consider doing as his mother says, however, he has denied her power over him.

In general, if Bernie's own forces are weaker than his mother's forces, she has power over him. She has successfully affected Bernie's potential movement in his life-space.

Control in Field Theory

According to field theory, however, power does not necessarily mean successful control. As you can recall, a person's movement in his or her life-space is the result of all the forces acting on the individual.

In the example above, even if Bernie's mother does have power, Bernie will not automatically become a dentist. Maybe his father wants him to be a doctor. Bernie's actual movement depends on the strength of all of the forces acting on him.

Thus, in field theory, we must distinguish between "power" and "control." Power is the ability successfully to influence another person's potential movement in a life-space. Control is the ability to dictate a person's actual movement. This distinction arises because several people may have power over a person at a particular time.

Another theory from the relational perspective comes from Thibaut and Kelley (1959). Their approach to power is based on an analysis of game matrices. As you can recall, we used these game matrices earlier to represent interdependence and conflict.

Thibaut and Kelley looked at relationships between two people. They believed that such relationships involve two types of power: "fate control," and "behavior control."

To describe the differences between fate control and behavior control, we will first concentrate on only one person's control over the other. We will use game matrices to illustrate these concepts.

Fate Control

Figure 5.1 is an example of fate control:

FIGURE 5.1

		Steve
			Yes		No
Ron
	Yes	4		1
	No	4		1

In this game matrix, Steve can determine the number of points Ron gets from the game, no matter the number of trials or what Ron does. If Steve moves "yes," Ron gets four points whether he moves "yes" or "no." If Steve moves "no," Ron gets only one point no matter what he does. This is fate control because Steve has total control over the outcome of the situation.

Imagine that Steve and Ron are working on a report. Let us replace the words "yes" and "no" on the game matrix with two options for the content of the report. "Yes" now stands for a section on environmental laws, and "no" stands for a section on governmental process. The numbers indicate the amount of credit Ron will get for his contribution to the report. He will get either four points or one point, based entirely on what Steve chooses to write. What Ron writes does not matter. He has absolutely no control over how well the paper will do. Steve has fate control over Ron, because Steve's choice of topic completely determines how many points Ron will get.

Behavior Control

Figure 5.2 is an example of behavior control:

FIGURE 5.2

		Linda
			Yes		No
Debbie
	Yes	4		1
	No	1		4

In this situation, Linda does not have complete control over Debbie's outcomes. If Linda moves "no," for example, Debbie may get four points or one point, depending on her own decisions. This relationship is different from the relationship between Steve and Ron.

Linda does have some control, however. She can influence Debbie's decisions. For example, the two women are playing the game, and Linda decides that she wants Debbie to move "no." How can she make this move desirable for Debbie? By moving "no" herself every time she plays. Over a series of trials, Debbie learns that it is best for herself to move "no." When she does so, she wins more points than when she moves "yes."

Because Linda can influence the moves Debbie makes, she has behavioral control over Debbie.

Imagine that Linda and Debbie are writing a report together. Once again, as with Steve and Ron, the move "yes" stands for the topic of environmental laws, and the move "no" stands for governmental process. In the previous example, what Ron chose to write did not matter. Steve had all the control. This time that is not the case.

In this situation, Debbie's choice is important. It will help determine her points. Linda, however, has a great deal of control as far as being able to influence which topic Debbie chooses. For example, Linda wants Debbie to write about environmental laws. What does she do? She says that she will write about environmental laws too. Now Debbie can decide for herself how many points she wants. Will she write a section on governmental process and get only one point, or will she join forces with Linda on the topic of environmental laws and get four points? Most likely, Debbie will agree to write about environmental laws, as Linda wanted her to do.

Fate Control Affecting Both People

The person who controls the situation does not always have a stake in the outcome. In the above game matrices, neither Steve nor Linda could win any stated number of points. What happens when they can? Let us examine this situation and expand our analysis to include both people playing the game at the same time. By doing so, we can reach further insights into the relational perspective's concept of power.

Let us begin with fate control and examine the two matrices in Figure 5.3:

FIGURE 5.3

A)

		Steve
			Yes		No
Ron			3		3
	Yes	4		1
			3		3
	No	4		1

B)

		Steve
			Yes		No
Ron			4		4
	Yes	4		1
			1		1
	No	4		1

In each matrix, the number of points that Ron can win remains the same. Also, in both matrices, Steve has the same amount of fate control over Ron. When we look at the points that Steve can win, however, we see that the two games are different. What do these differences mean?

Matrix A. In Matrix A, the outcomes for Steve are the same in every circumstance. No matter what either Steve or Ron does, Steve will get three points.

Thus, if the two men were writing their report, Steve would get the same credit regardless of the topic. Also, as before, the topic that Steve chooses will control the credits that Ron gets, no matter what Ron chooses to write. We can see that, as in the previous example, Steve has complete fate control over Ron. Does Ron have any fate control over Steve? No, he does not.

You may note that if Steve chooses "yes" in the game, Ron will end up with more points than Steve. The score will be four to three, in Ron's favor. This does not mean, however, that Steve lacks power because he ends up with fewer points. Steve still has all the power in the relationship because his choice entirely determines whether Ron gets those four points. The value or size of an outcome that a player wins does not determine his or her power. Instead, power belongs to the person in the relationship who determines how much each person gets. History is full of stories of people who were not the richest in a kingdom but who had the most power because they could influence the king more than anyone else.

How would Steve handle the power that emerges from a situation such as in Matrix A? Most game theorists would say that Steve would have no preference for either the "yes" or the "no" move. Game theorists believe that people consider only their own outcomes in making their choices. In other words, they believe that people have "individualistic orientations." Because Steve will get three points no matter which move he makes, they would say he would not prefer one over the other. Thus, game theorists would predict a 50 percent chance that Steve will allow Ron to get the greater credit.

Game theorists are wrong about this prediction, however. As we discussed in Chapter 4, research about conflict implies that Steve is likely to move "no" to get more points than Ron. We will discuss this idea further when we look at research involving game matrices.

Matrix B. Now let us turn to Matrix B. In this situation, Steve still has complete fate control over Ron, but Ron can also control the number of points that Steve gets. If Ron chooses "yes," Steve gets four points. If he chooses "no," Steve gets only 1 point. As we can see, in this situation Steve and Ron both have an equal amount of fate control over the other.

Imagine that they are writing their report under these conditions. Each can now control the credit that the other gets. Ron decides to write about governmental process ("no") and wants Steve to write about environmental laws ("yes"). Steve does not agree to this plan because Ron would get four points and Steve would get only one. Instead, Steve retaliates by saying that he was going to write about governmental process too, and thus each would only earn one point. Each has enough power to retaliate in such a way if he does not receive the credit he wants. In the long run, we would expect them to reach an understanding. Both would write about environmental laws, and thereby each would earn four points.

Behavioral Control Affecting Both People

We can also find interesting implications when we look at circumstances of behavioral control in which both people have a stake in the outcome. This is illustrated in Figure 5.4.

FIGURE 5.4

C)

		Linda
			Yes		No
Debbie			4		1
	Yes	4		1
			1		4
	No	1		4

D)

		Linda
			Yes		No
Debbie			1		4
	Yes	4		1
			4		1
	No	1		4

In both these matrix situations, Linda and Debbie have an equal amount of behavior control over each other. In Matrix C, each can exert behavior control over the other to get maximum payoffs. Through successive trials, they can learn that it is best if they both choose "yes" or both choose "no."

Perhaps you have noticed that Matrix C is analogous to the example of a "cooperative social situation" that we discussed in Chapter 1. It was part of our introduction to the relational perspective.

Matrix D, in contrast, is different. The players are unable to find a "win-win" outcome. Whenever one makes a move to increase her number of points, the other player's points decrease. For example, if Linda moves "yes," Debbie will move "yes" to have the most points. She will hence have four points and Linda will have only one. If Debbie moves "yes," Linda will move "no" and win. As we can see, each player's move influences the other's choices. In this way, the players exert equal behavior control over each other because each makes her move based on what the other has done.

Again we can look back to Chapter 1 for something similar to Matrix D--the "competitive social situation."

Overview of Game Theory

What do all these game matrices have in common? In all of them, one person's choices or outcome depends upon the other person's moves. This, in short, is Thibaut and Kelley's view of power. Power exists when one person's actions depend on the actions of another person.

What about when both people's choices or outcomes depend on the actions of each other? In this case, we can say that each has some power over the other. The specific term for this kind of relationship is interdependence, in the sense that we used that word when we defined the relational perspective in Chapter 1.

To switch this reasoning around, we can say that whenever two or more people are interdependent each has some power over the others. Interdependency gives everyone some power.

With this in mind, it should be no surprise that power is a basic concept for theorists who have adopted the relational approach. Interdependence is a core concept for this perspective, and interdependence leads to power. Power hence is also a core idea in the relational approach.

Further Research Using Matrix Games

As we have seen, matrix games are helpful for researchers using the relational perspective. They illustrate relationships and show how people interact within these relationships. We have seen how researchers can set up various game matrices. In some, one player has complete fate control over the other; in others, one player has complete behavioral control over the other. In still others, both players have fate control or behavioral control over each other.

It follows, then, that once researchers understand the power relationships that matrices can create they begin to use matrix games to study how people use power.

Prisoner's Dilemma Game

We can interpret much of the research that we reported in Chapter 4, on conflict, as also being relevant to the issue of power. For example, players in the Prisoner's Dilemma Game have both fate and behavioral control over one another. Consider the Prisoner's Dilemma matrix in Figure 5.5:

FIGURE 5.5

		Person A
			Yes		No
Person B			5		9
	Yes	5		0
			0		1
	No	9		1

Let us carefully walk through the process to verify the kind of control that each player has.

First we will look at Person A. Does this player have fate control or behavioral control over Person B?

Imagine that Person B chooses "yes." In this case, B would get fewer points if Person A chooses "no" than if A chooses "yes." Imagine instead that B chooses "yes." Again, B would get fewer points if A chooses "no" than if A chooses "yes." In essence, no matter what Person B does, Person A can influence the number of points B receives. In either case, if A chooses "no," B will get fewer points than if A chooses "yes." In this way, Person A does have some fate control over Person B.

If we now examine Person B in the same way, we can see that this player has the same amount of fate control over Person A. No matter what Person A does, Person B can influence how many points A receives.

As we can see, both players have some fate control.

What about behavioral control? Again, let us look at Person A first. Can Person A influence the moves that Person B will make? The answer is yes. For example, if Person A wanted Person B to move "no," Person A could get this result. Over a series of trials, if Person A continually chose "no" Person B would soon learn that it was advantageous also to choose "no." When A chooses "no," B gets more points by choosing "no" than by choosing "yes."

The same reasoning holds true if we look at Person B. Both players have behavioral control in this matrix situation.

How do players respond to such a game matrix? How do they utilize the fate control and behavioral control that they have over each other?

Individualistic strategies. Game theorists assume that players use "individualistic" strategies. Each player tries to do his or her best. The main focus is on a person's own score and not on how the other player is doing. If players do use such strategies, behavioral control is the overriding concept. The implication is that Prisoner's Dilemma Game players will use their behavioral control over each other to compromise, and both will choose "yes." In other words, rather than wanting to use fate control to affect how the other player does, each player wants the best outcome possible for herself or himself. The players use influence only so far as it helps themselves, rather than hurting others. Over a series of trials, they learn that the only way they will each get the most points is by both choosing "yes."

Use of fate control. Research suggests, however, that players do not always use individualistic strategies. Many times players concentrate on how the other player is doing. They don't simply want to get the most possible points for themselves. They also want to affect the other person's outcome.

The research that we discussed in Chapter 4 illustrates this idea. It implies that players often try to use their fate control over one another. Each wants to get more points than the other by choosing "no." Players don't only want to do well themselves, they want to "win." The way that players use gates in the Acme-Bolt Trucking Game is a similar use of fate control to try to "win" the game.

In short, when players are given fate control, they are likely to use it against one another.

Apparently, game players expect this. They believe that, when their opponents are in a position of power relative to themselves, the opponents will use the power for their personal advantage.

We can illustrate this idea with the matrix in Figure 5.6.

FIGURE 5.6

		Bill
			Yes		No
Gretchen			4		3
	Yes	3		0
			3		2
	No	5		2

In this matrix, Bill has more fate control over Gretchen than Gretchen has over him.

Research suggests that Gretchen will expect Bill to use his power and choose "no" so that she will end up with the least number of points.

Unexpected variations of fate control. What happens, however, when a powerful player does not use fate control over the other participant? What would happen if Bill surprised Gretchen and chose "yes"?

Swingle (1970) decided to investigate this idea. He suggested that Gretchen would respond to such a move by playing aggressively and choosing "no." In this way, she would end up with more points than Bill. Swingle saw two scenarios that might explain such a response from Gretchen.

The first scenario is that Gretchen comes to see Bill as incompetent. Gretchen wants to win the game. She believes, accordingly, that Bill does too. The more powerful Bill is, the easier it is for him to win. The more powerful that he is, the more Gretchen will decide that he is stupid if he does not use his power to win. Because she believes that he wants to win, she feels that the only reason he does not win is that he does not know how to use his power as well as he should. If she believes Bill is stupid, she will play aggressively without expecting him to retaliate.

The second scenario is that Gretchen may anticipate that the more powerful Bill is, the more cautious he will be. Bill will be careful because he does not want to appear to be a bully. Hence, Gretchen can feel some safety while she plays aggressively. We can see this second scenario at work in the interaction of nations. Powerful nations often find it difficult to respond forcefully to attacks by smaller countries. If the powerful nations do respond, they are likely to lose face in the eyes of the world.

Were Swingle's two scenarios correct? He tested his expectations in a series of Prisoner's Dilemma trials. Research participants played confederates who cooperated 98 percent of the time. How did the participants respond?

The participants were most aggressive when the confederate was in a more powerful position. They used intermediate levels of aggressiveness when the confederate had the same amount of power, and surprisingly, they were least aggressive when the confederate was less powerful.

As you can see, Swingle's results reinforce his scenarios. It seems that Gretchen would probably play aggressively if Bill had more power than she and that Bill would perhaps play cautiously if he had more power than Gretchen.

As we discussed in Chapter 1, structural theorists differ from relational theorists. Researchers using the structural perspective go beyond relational ideas to discuss the social patterns that emerge from interaction.

As you will recall, these patterns include norms and roles and other structures that govern human relationships. Structural theorists believe that as relationships develop a process takes place. When people begin a relationship, they behave a certain way with each other. Over time, they start to form expectations about how they will act toward each other in the future. These expectations, in turn, eventually become evaluations of how they ought to act when they are together. Thus, what people expect changes into how they feel people should act.

How do structural theorists approach the idea of power? From their viewpoint, the above process takes place. For instance, one person regularly displays more power than the other in their interpersonal interactions. Over time, both people come to expect that power difference as part of their relationship. Eventually, such a power relationship takes on a degree of permanence.

In the process outlined above, the power relationship emerges over time. This is not always the case. Instead, group structures may exist that define from the beginning how power ought to be in a relationship. For example, a group may have an elected or assigned leader. It may also be part of an organizational structure, and one or more of its members may be higher up in the chain of command than others in the group.

Using these ideas, structural theorists have created theories about power.

A Cost-Reward Theory

Social exchange is a cost-reward theory that is relevant to power. This theory proposes that we interact through the exchange of "resources." Our goal is to advance our self-interest. We barter, each hoping to get what he or she wants most out of our exchanges. According to Foa and Foa (1974), we can classify the resources that we find attractive to exchange in the following six categories:

1. love (including affection and comfort)

2. status (including respect and esteem)

3. services (basically, activities that help other people)

4. goods (tangible products)

5. information (including advice and opinion)

6. money

Cultures tend to value exchanges differently. In the United States, we usually look favorably on exchanges within a category, such as love for love or information for information. We also accept some exchanges that cross category lines, such as money for goods or services. Conversely, we tend to frown on some other exchanges that cross categories, such as love or status for money.

Despite these social values, however, it is conceivable that people can satisfactorily exchange resources among any of the categories. In the end, the people making the exchange must decide if they are satisfied. Social exchange theory maintains that a satisfactory exchange occurs when each participant gains a profit. According to Homans (1961, p. 62), the "open secret" of exchange is for each person to give something away that is more rewarding to the receiver than it is costly to the person giving it away. Both people profit when this happens.

Blau's Social Exchange Theory

Blau (1964) proposed a view of power that follows from social exchange ideas. He believed that power becomes important in a relationship when three conditions apply. Let us use the letters A and Y to indicate two people in a relationship, to illustrate these three conditions.

First, A needs a resource from Y that A cannot find elsewhere. Second, A has nothing that Y wants in exchange. Third, A cannot force Y to give up what A wants.

When these three circumstances apply, according to Blau, A has no choice but to subordinate to Y and follow Y's wishes.

How does this fit in with social exchange theory? The idea is that A does have one resource that Y might want to have. That resource is power over A, the power that each person has over himself or herself. If A has nothing else to offer Y, and he or she wants what Y has badly enough, A can offer this power as a resource to exchange.

In a basic sense, we could give the example that Robin has food and Sandy is hungry. If Sandy has nothing to give Robin in exchange for the food, she can offer power over herself as an incentive for Robin to give her food.

In another example, Sandy wants very much to be Robin's friend, but Robin is indifferent. To spend time with Robin, Sandy may have to give Robin power. She may have to be willing always to do what Robin wants.

Two implications are significant in Blau's point of view. First, power is always one-sided. People do not like to give up power over themselves. Therefore, if they have any other resource that they can exchange, they will do so. When this happens, neither person has power over the other. For example, if Sandy had a skill at math and Robin needed her help, Sandy would not have to give up power over herself to be Robin's friend. Instead, Sandy could offer Robin help with math, and they would spend time together based on that exchange of resources.

Second, power is only truly important when one-sided exchanges in a relationship are normal rather than occasional. In this way we consider Blau's idea a structural theory. Power becomes central to the structure of a relationship only when exchange is generally one-sided. For instance, Robin will have power only if she usually has less interest in their relationship than Sandy does. If at times Robin has interest, she loses her power.

We have mentioned how Thibaut and Kelley's game theory has aspects of social exchange theory. When we look at the two above implications from Blau's approach, however, we can see how different the two approaches are. Blau's is clearly a structural theory, whereas Thibaut and Kelley's theory is relational. Thibaut and Kelley would not agree with either of the significant implications from Blau's theory. In contrast to Blau, they think that people can simultaneously have power over one another, and they think that short-term power differentials between people are significant.

The theories we have discussed so far did not specifically apply to decision-making groups. The expectation states approach to power (Berger, Fisek, Norman, & Zelditch, 1977) does apply to these groups.

According to this approach, when group members must make a decision, they want to identify the member who is most competent to make that decision. Once they have identified him or her, they are willing to let that person have power over the decision-making process.

Qualified Group Members

The theory's initial claim is that members try to find the most qualified person to make the decision. For example, a group must solve a series of mathematical puzzles. Members discover a math professor in the group. According to expectation states theorists, group members will defer to the math professor as they solve the puzzles. Hence, the math professor gains power within the structure of the group.

Competent Group Members

What if there is no qualified person in the group? Expectation states theory claims that the group does not give up looking for a person they can consider competent. They go through a similar process even when the group has no members with the needed skills. In this case, group members try to identify the person who is the most "intelligent" or "competent" among them. Once this is done, they allow that person to have power over the decision-making process.

Types of people and types of tasks. Groups can do this in several ways. One way involves associating people with tasks. Every culture does this. The belief is that a particular type of person is competent at a certain kind of task.

For instance, a group must make a decision concerning a mechanical problem. In this group's culture most people believe that men are better than women at mechanical tasks. In such a scenario, the group will tend to look for a man to control their decision unless a woman can show that she is skilled at mechanical tasks. What if this group needs to make a decision about a cooking task and its culture associates women with cooking skills? The same process is likely to result in a woman having the most power in the group as it examines the cooking task.

Types of skills and types of tasks. Another way a group will try to find its most competent member involves people's beliefs about which skills tend to go together.

For example, a group must make a series of decisions that concern word identifications. The group does not include a language expert; so members cannot get a qualified person to lead them. They also do not have any cultural beliefs that a particular type of person is good at this task. What can they do? One of their members is a math professor. The members may reason to themselves that her skill at math requires intelligence. They then tell themselves that a skill at language also requires intelligence. By associating these two skills, the group may decide to give her power over their decision making.

In other words, a "halo effect" (see Chapter 3, "Cohesiveness") takes place. If group members think that two skills go together, they will expect a person with one of the skills to have the other one also. This process can work similarly when group members believe that two skills do not go together. For example, many believe that people who are good at math are bad at art. Therefore, if a group had a decision relevant to art, a math professor would be the last person they would give power.

The expectation states theorists have performed many studies showing the above processes at work. We will describe a few later in this chapter.

Another structural approach to power has come from Bacharach and Lawler (1980). Their analysis is particularly relevant to groups that exist within organizations.

According to their theory, a person has power if he or she is able to make decisions that affect the activities of group members. For decision-making groups, Bacharach and Lawler see two types of power at work: authority and influence.

Authority exists when a group's structure gives certain members the right to make decisions for the group. For example, an organization might have something called the "remodeling group." This group must make a decision regarding the new employee lounge. The president of the company tells the group that Fred is their assigned leader and that whatever Fred decides will be how the group should plan the new lounge. In this case, Fred has authority. Other similar positions, such as organizational superordinates or elected leaders, have the same kind of authority.

Influence exists when members who do not have official authority are able to affect the activities of the group. People gain influence through persuading, suggesting, giving information, and the like. For example, in the remodeling group Sara and Mike have no official titles. The president of the company has given them no authority. The group's leader, Fred, however, listens attentively when Sara and Mike speak. Eventually, Fred's decision is based on some of what they have said. In this case, Sara and Mike have influence.

Amounts of Authority and Influence

One of the other differences between authority and influence involves the amount of each type of power that can exist in a group.

Authority exists in a set amount. If a group has an official leader, it has a given amount of authority. To give other group members more authority, one has to take away authority from the leader. Authority can never be more or less than its given amount; when one person gains it, another person must lose it.

For example, in "the remodeling group," Fred originally made all group decisions. He had all of the group's authority. Suppose, however, that the president of the company decides to change the group's decision-making procedure. The president now says that decisions will be made through voting or consensus. By doing this, the president has taken away a great deal of Fred's authority and given it to the group members.

Influence, however, exists in variable amounts. Group members could have no influence if a group is governed strictly by authority. On the other hand, the more that group members can affect a decision, the more influence they have. This amount can rise substantially.

When the remodeling group started out, for instance, Fred made all group decisions without consulting any of the other members. At that time little if any influence was at work in the group. Suppose the group changed its procedure so that Fred still made all the decisions but only after consulting with the others. He listened to their information and suggestions. With this change, the amount of influence in the group rose dramatically.

Legitimacy and Authority

Bacharach and Lawler saw legitimacy as important in determining who has power in a group.

In this part of their theory the researchers are concerned with authority, not with influence. Their belief is that legitimacy affects whether authority can work well in a group.

Bacharach and Lawler felt that authority is legitimate when group members believe that both the leader's position and his or her actions are appropriate.

Position. If group members do not accept the way a leader got his or her position, problems may arise. The members will not view this leader as legitimate, and they will be less likely to comply with his or her decisions. For instance, a group may feel that an assigned leader was chosen arbitrarily, rather than because of his or her abilities. In this case, members will probably have less respect for that leader's directives.

For example, the remodeling group may have heard that Fred frequently has lunch with the company president and that he knows nothing about remodeling buildings and making them comfortable for employees. In such a case, the members probably will think Fred's position as leader is not appropriate, because it came about because of friendship rather than his abilities. They will then think he is not a legitimate leader and will be less likely to accept his decisions.

Actions. A leader who takes actions that group members feel are not legitimate will find it difficult to use authority. It may be, for instance, that a leader demands privileges that other group members cannot get. Further, if the leader cannot explain any reason for these privileges, group members are likely to become disgruntled, and the leader will lose their support.

For example, Fred does know about employee lounges, but he behaves inappropriately at the meetings for the remodeling group. He comes late, falls asleep during the meeting, makes insulting comments, and so on. In this case, the group may decide he is not a legitimate leader and no longer follow his directions. Even though the group may approve of how Fred got his position, his actions have decreased their respect for him.

Legitimacy and Influence

Interestingly, Bacharach and Lawler did not look at the effects of legitimacy upon influence. Instead, they simply said that all influence is not legitimate.

They believe that influence is not legitimate because it runs counter to the "official" group power structure. For example, when group members such as Sara and Mike use influence, they circumvent Fred's official power. This behavior is not appropriate, according to Bacharach and Lawler. Therefore, it is not legitimate.

We consider this viewpoint inconsistent with Bacharach and Lawler's other reasonings.

We have seen how Bacharach and Lawler believed that appropriateness is the deciding factor when looking at legitimacy. We have outlined the process with regard to authority. Group members look at whether a leader's position and actions are appropriate. Using this criterion, they decide whether the leader's authority is legitimate.

We believe the same process, in reverse, can apply to influence. In the case of influence, the leader, not the group, evaluates what is appropriate. He or she can then decide whether the influence of group members is legitimate.

For example, Fred decides that suggestions from Sara and Mike are legitimate. He does so because they are long-time members of the remodeling group and because they always act responsibly. This legitimacy helps give Sara and Mike influence. Fred feels that suggestions from Terry are not legitimate, however, because Terry recently joined the group. In meetings, Fred can stop Terry from talking and can give other signals that show he does not find Terry's actions appropriate. In this way, Terry will not acquire any influence.

In addition to their other theories, structuralists have also been concerned with the basic dimensions of power. They have wished to understand what these dimensions are and how they affect groups.

What do we mean when we say "dimensions"? A dimension is part of a whole, something we can pull out of a structure and examine to see how it fits with everything else. As we all know, we can describe a certain room by saying that its dimensions are 12 feet by 10 feet. One dimension is 12 feet, another is 10, and together they make up the total room.

We can apply this idea to power, a many faceted concept. Structuralists wished to discover those facets and learn how they work.

Many theorists have proposed lists of the dimensions of power. We will focus on one particular list and relevant experiments.

The Five Dimensions of Tedeschi and Lindskold

Tedeschi and Lindskold (1976) proposed five dimensions: status, resource control, expertise, trustworthiness, and likableness. We will distinguish the five according to whether they affect authority or influence. The first two dimensions, status and resource control, relate to authority. People in positions of authority can use status and resource control to get subordinates to do what they want. The last three dimensions are connected to influence. People not in positions of authority can use their expertise, trustworthiness, and likableness to get others to follow their suggestions.

Study Methodology

Before we examine each dimension in detail, we will look at the types of experiments that researchers have used to study them. In this way we will have an overall idea of how each study worked. Later, we will refer to these study methods when we discuss each dimension.

The dimension of trustworthiness has been omitted from research. Trustworthiness is the extent to which members of a group believe that a fellow member is sincere and unselfish. Because this dimension of power has not been a topic of much research, we cannot examine it in detail in this book.

Modified Prisoner's Dilemma Game research. Tedeschi and his associates have done extensive research into the other four dimensions. Their studies were in relation to conflict situations, and they used a modified version of the Prisoner's Dilemma Game. In their game, a confederate threatened to take away points from a real participant if he or she did not make the moves that the confederate wanted. Tedeschi and his associates would establish a confederate's degree of power in terms of varying dimensions. We will discuss some of the ways in which they used the dimensions later.

As a further part of the study, the researchers manipulated the confederate's credibility by varying the number of times that the confederate would follow through on the threat.

Tedeschi and his associates discovered that their dimensions did affect power. When power was derived from these dimensions, it could increase participant compliance with the confederate. They also learned that threat credibility complicates the effects of power. The threat had to be credible to work effectively.

Expectation states research. Other research relevant to the dimensions of power includes several studies that expectation states theorists (Berger, Fisek, Norman, & Zelditch, 1977) performed. They tended to use a particular method that involved two participants who would sit in the same room but be separated by partitions. In this way, they could both know that they were in the room together but could not see each other. The researchers gave them a test for an imaginary skill, such as "contrast sensitivity." Using such a vague term, the examiners made the participants believe that they had either higher or lower levels of that skill.

Once this was done, the researchers gave both people a series of about 20 problem-solving tasks. Each task had two possible solutions. The researchers asked the participants to decide individually which option was right. This round of answers was to be "preliminary." The tasks were phony in that neither option was clearly correct.

After this individual test, the researchers said that they would show the participants each other's answers. They did not really do this, however. About 80 percent of the time the researchers showed answers that were simply the opposite of what the person had chosen. In this way, each participant believed that he or she generally disagreed with the other person being tested.

Finally, the researchers asked the participants again to make "final" individual decisions about the tasks. The point was to see how often the participants changed their choices between the "preliminary" and "final" series of answers. Researchers wanted to know how often people want to agree with one another.

We will now discuss how researchers used these studies to examine the dimensions of power that Tedeschi and Lindskold proposed.

Status

People gain status by means of their positions. These positions, in turn, come through organizations and society. When people believe that someone's position gives him or her legitimate power over others, that person has status.

Studies have shown that people in positions of authority influence others. For example, drivers usually follow the directions of a school crossing guard. They see the symbols of authority, such as a uniform, and they respect the guard's position and influence.

Various factors affect how much influence status really gives a person. We will focus on two, relevance and credibility.

Relevance

In the example above, the school crossing guard has influence when drivers see him or her in a street and children are nearby. In this situation, drivers will probably follow his or her directions because the guard's status as a director of traffic has relevance to the situation.

Yet, take the guard out of that situation, and things change. For example, we would not expect people to take the guard's advice about which foods to buy in a supermarket. The guard's status has little relevance in such a setting.

Several researchers have examined the ways in which relevance affects status and power.

When status is relevant. What happens when a discussion takes place in a setting in which certain members have status? What happens when the topic under discussion sometimes is relevant to the setting and sometimes is not? Bass and Wurster (1953) studied these questions.

Bass and Wurster examined status in a large corporation. They asked supervisors to participate in a decision-making group discussion. Observers then rated the ways in which members of the group influenced one another.

Overall, researchers found that the higher a supervisor's position, the more he or she attempted to influence other group members. The observers also rated how successful these attempts were. They found that the relevance of the topic was important. The groups discussed topics that were either relevant or not relevant to the corporation. When the topic had nothing to do with the corporation, the influence of higher-status supervisors was somewhat less than when the topic did relate to the corporation.

Thus, in the context of a corporation topic, the supervisor's status helped him or her have influence. When the context changed, the supervisor's influence declined.

When status is irrelevant. What happens when the status of an individual is not at all relevant, either to the setting or to the discussion topic? In such a circumstance, status has a limited effect. One test of expectation states theory (Berger, Fisek, Norman, & Zelditch, 1977) showed this to be true.

Members of the military participated in this test. Researchers told the participants that their partners were either higher or lower in rank than they were. Yet, the test was out of the context of the military, so each person's military rank was irrelevant.

What happened when participants thought they were lower in rank than their partners? These people changed their decisions in 25 percent of the cases when they thought they had disagreed with their partners. When people thought they were higher in rank, they changed their decisions 19 percent of the time. The groups were only a few percentage points apart in their rates of change. Status only slightly affected their decisions.

Credibility

Another factor that affects status is credibility.

Faley and Tedeschi (1971) used their modified Prisoner's Dilemma Game to study the effects of status and credibility on participants' moves.

Participants were ROTC members. Researchers started out by rating the participants' status according to their number of years in the ROTC. Participants had low status if they were in their first or second year of ROTC and high status if they were in their third or fourth year. Each participant played an "opponent" who was actually a confederate. Examiners told participants that their "opponent" was either a high-status or low-status ROTC member.

To make credibility a factor, the researchers attributed to the "opponents" three levels of credibility: high, average, and low. At high levels, the confederates followed through with their threats 90 percent of the time. At average levels, the confederates followed through 50 percent of the time, and at low levels, they followed through 10 percent of the time.

At the average and high levels of credibility, participants complied more with high-status "opponents" than with low-status ones. When the "opponent" had low credibility, however, participant compliance was low regardless of the confederates' status.

Credibility appeared to significantly affect whether status helped give people influence.

Resource Control

Resource control refers to the ability of a person in authority to reward and punish other members in his or her group.

When bosses can promote their employees or fire them, bosses have resource control. Most employees are far more likely to do what their bosses tell them under these conditions.

Bennis, Berkowitz, Affinito, and Malone (1958) looked at rewards in a hospital setting. They examined nurses who hoped for a certain level of reward from their supervisors. These rewards were in terms of salaries, promotions, and praise. The researchers found that the more the nurses received what they hoped for, the more time they spent doing what their supervisors wanted.

Lindskold and Bennett (1973) looked at both rewards and punishment. In their study, threats and promises were resources that a confederate could control. Lindskold and Bennett examined the effectiveness of these resources by manipulating which ones their confederates could deliver. For instance, sometimes the confederates could use only promises. At other times, they could use both promises and threats.

Lindskold and Bennett found that promises alone were not as effective as a combination of promises and threats. Confederates who could deliver both a promise and a threat got participants to comply more often than confederates who could deliver only a promise.

Expertise

Expertise refers to the knowledge or skill of a person in a relevant area.

How effective is expertise in helping to give people power? Overall, it is particularly effective when a person with expertise is dealing with people who believe they are not experts. A good example is the interaction between physicians and patients. Most patients feel they lack knowledge and skill about their health problems, and they comply willingly with their physician's orders. Another common example is the interaction between auto mechanics and customers. People are usually willing to comply with the demands of mechanics when it comes to their seemingly most prized possessions, their cars.

Researchers have studied how effective expertise is in this type of situation. Several tests of expectation states theory (Berger, Fisek, Norman, & Zelditch, 1977) are relevant to expertise.

The effect of designated skill levels. In one test, researchers told participants that they and their partners were either high or low in an imaginary skill called "meaning insight." Examiners then asked the participants to solve a series of "meaning insight" problems.

Researchers looked at what happened when they told participants that they had disagreed with their partners. They wanted to see how often these people changed their answers to agree with their partners.

Examiners found that when participants thought they had a lower skill level than that of their partners, they changed their answers 56 percent of the time. Conversely, when participants believed that they had a higher skill level than that of their partners, they changed their answers only 22 percent of the time.

Clearly, participants felt affected by their supposedly uneven levels of skill. Those who thought they had relatively low skill levels were much more likely to change their answers than those who thought they had relatively high skill levels.

The halo effect and skills seemingly related. Berger and his associates also showed the impact of halo effects in expertise.

In one test, researchers again told participants that they had different levels of a skill called "meaning insight." Next, they gave the participants a series of problems that they said related to another imaginary skill, "spatial insight."

Even when the researchers did not tell the participants that these two skills were related, the people in the study seemed to believe that a person with the skill of "meaning insight" would also do well at a task that needed the "spatial insight" skill.

Again, researchers looked at the times when participants thought they had disagreed with their partners. Participants who thought they had lower skills than those of their partners in "meaning insight" changed their decisions at a rate of 40 percent. With those who thought they had higher skills in "meaning insight" than did their partners, the rate of change was only 25 percent. This was true even though the problems supposedly related to "spatial insight" and not to "meaning insight."

Thus, the participants' beliefs about their expertise in a skill affected how they reacted to problems related to another skill. They were much more likely to change decisions related to "spatial insight" if they thought they were relatively lower in the "meaning insight" skill.

The halo effect and skills negatively related. In another study, Berger and his associates again gave participants ratings in "meaning insight." They then told participants that people who have high ratings of "meaning insight" have low ratings of an imaginary skill called "contrast sensitivity." They also said the converse of that was true, that people unskilled in "meaning insight" were highly skilled in "contrast sensitivity."

The researchers then gave the participants a series of "contrast sensitivity" problems to solve. As before, Berger and his associates looked at the times when people changed their answers to agree with their partners. Their results showed that the participants apparently believed what they were told.

Those who thought they had relatively higher levels of "meaning insight" seemed to believe they consequently had lower skills at "contrast sensitivity." These people changed their decisions 53 percent of the time. Similarly, when participants thought they were relatively lower in the "meaning insight" skill, they apparently believed this meant they were higher in the "contrast sensitivity" skill. They changed their decisions only 31 percent of the time.

Both studies show that the halo effect appears to work with expertise. When people believe that a person has a skill, they often take it for granted that he or she will or will not have another skill. What they believe depends upon how they think the two skills are related. They may feel that a person who has one skill should have another. Conversely, they may think that someone who has expertise in one area will definitely not have it in another. Halo effects appear to occur with regard to expertise.

Modified Prisoner's Dilemma Game studies. Tedeschi, Schlenker, and Bonoma (1975, Exp. 2) used their modified Prisoner's Dilemma Game to study how expertise and credibility functioned together. They wanted to see how these two factors affected participants' moves.

The researchers used confederates who were supposedly experts or nonexperts in aggression because they wanted an area of expertise that would seem relevant to conflict games. Tedeschi and his associates based the confederates' levels of expertise on their abilities in karate.

The study found that expertise on its own weighed heavily with the participants. If the confederates were experts, whether their threats were credible did not seem to matter. The researchers introduced the experts as people who had black belts in karate. Participants complied relatively often with these confederates, whether or not they carried out their threats.

Credibility did have a significant effect when the researchers gave it to the nonexperts. They introduced the nonexperts as those who were below average students. When these confederates carried out their threats, giving them credibility, the participants complied relatively often.

When confederates had neither expertise nor credibility, they had the least success. Participants complied least often with the wishes of nonexperts whose threats were never carried out.

Likableness

Overall, humans are more likely to do what someone wants if they like that person than if they do not like him or her.

Tedeschi, Schlenker, and Bonoma (1975, Exp. 1) used their modified Prisoner's Dilemma Game to study likableness and credibility. They examined how these two factors worked together to influence study participants.

At the beginning of the study, the researchers manipulated how much the participants liked or did not like their confederate "opponents" by using Byrne's method. As you can recall, we described in Chapter 3 how Byrne could increase a participant's liking for a fictional person. Byrne would first find out a participant's opinions. He then would say that the fictional person had attitudes that agreed or disagreed with those of the participant. In this way, the participant's liking for the "person" would increase or decrease.

Results showed that credibility was important. It affected whether likableness could give a person power.

Tedeschi, Schlenker, and Bonoma found that participants complied most with likable, credible opponents. Levels of compliance were always intermediate with an unlikeable opponent no matter how credible his or her threats were.

Interestingly, the opponents who were likable but not credible had the least amount of success. Participants complied least with these people.

Table 5.1 summarizes the five dimensions of power. It describes each dimension and indicates whether the dimension is related to authority or to influence.

Table 5.1	Dimensions of Power
Dimension	Type	Description
Status	Authority	Legitimate position in an organization or society
Resource control	Authority	Ability to reward or punish members of group
Expertise	Influence	Relevant knowledge or skill
Likableness	Influence	Liking for or desire to be liked by the person
Trustworthiness	Influence	Sincerity or unselfishness

We have examined ways in which relational and structural theorists have approached the concept of power. Now we will turn to another important aspect of power in groups, the formation of coalitions.

Whenever there is conflict among group members, the possibility always exists that coalitions will form.

A coalition occurs whenever two or more group members join to gain power over the other members of the group. Members of a coalition pool their personal resources. They expect that as a unit they will be able to control the group decision.

Up to now, theories and research studies about coalition formation have not related directly to decision-making groups. Instead, researchers have been more concerned with larger and less interdependent social collectivities.

Before we discuss specific theories, we will give an example of a typical situation that has interested coalition theorists.

At a nominating convention for a political party there are 90 delegates. Three candidates--Larry, Isabel, and Robert--are battling for the party's nomination. Forty delegates support Larry. Thirty delegates support Isabel, and 20 delegates support Robert. As one can see, no candidate has the majority. None of the delegates wants to change his or her vote; so the convention is deadlocked.

In response, representatives of the candidates meet in the proverbial smoke-filled rooms. They want to break the deadlock by forming a coalition of two candidates. A coalition of any two candidates will lead to a majority for the coalition candidate.

As in any negotiation, each person has something to offer and something he or she wants. For instance, the representatives for Robert know that they speak for the least popular candidate. They would be unrealistic to demand the nomination. Either Larry or Isabel will become the final coalition candidate. Robert's representatives, however, do have some things to offer and some things they want. They can offer the votes of 20 delegates, which will ensure the victory of either Larry or Isabel. They want jobs for Robert's supporters on the election team of the victorious candidate.

The other representatives of candidates have similar things to consider as they negotiate with one another. What will happen in this type of situation?

Scholars interested in coalition formation want their theories to account for two issues.

First, the theories should explain and predict how the members of the successful coalition will divide the rewards. Such a theory would tell us, for instance, how many jobs Robert's representatives will get for Robert's support.

Second, they want theories that can explain and predict which coalition will form in situations such as the one above. A good theory should tell us which two candidates will pool their delegates and win.

Essentially, the two questions are: How will those coalitions utilize their rewards? Which coalitions will form?

Vinacke and Arkoff's Game Theory

Vinacke and Arkoff (1957) proposed one early theory based on a claim that each member of a winning coalition is equally important because without each member the coalition would not win.

Utilization of rewards. For instance, three people are building a bridge. One person may contribute only the last 4 bricks to a structure that required 100. This person is not less important than the others. Without his or her 4 bricks, the bridge would not have crossed the river. Thus, when the time comes for rewards, according to Vinacke and Arkoff, all three people should share equally. For instance, all three should get the same share of a toll that they charge people crossing the bridge.

In other words, this theory holds that rewards are divided by a "norm of equality." The norm is that all members of the coalition have equal importance. Once this norm has been established, all behavior follows it.

Let us go back to our example of the political convention. Vinacke and Arkoff theorize that if Robert enters into a coalition with another candidate his supporters should receive half the jobs of the eventual election team.

This idea of the "norm of equality" answers the research question, How would the coalitions utilize their rewards? What about the other research question? Which coalitions do Vinacke and Arkoff think will form?

Which coalitions form. Their answer follows from their idea that members of a coalition are all equally important and thus will get the same reward no matter which coalition they are in if that coalition wins. People need not weigh which group will benefit them most. The person with the four bricks, for instance, need not wonder which group of bridge builders to join. As long as the group he or she joins gets the bridge across, the rewards will always be the same. Hence, Vinacke and Arkoff theorized that people do not care which coalition they join as long as that coalition wins.

Let's look at the candidates. Robert has enough votes to help either Larry or Isabel win the nomination. In such a case, this theory predicts that both coalitions are equally likely. Robert has no obvious reason to prefer one over the other because both would win. Further, for the same reasons, Larry and Isabel are equally likely to form a coalition without Robert. All three coalitions have a one-third chance of occurring.

Vinacke and Arkoff believed that their proposal was consistent with game theory. For that reason it has come to be known by that name.

Gamson's Minimum Resources Theory

Gamson (1961) made an alternative proposal from that of Vinacke and Arkoff.

Utilization of rewards. Gamson did not believe that coalition members give out rewards based on a norm of equality. Instead, he believed that a "norm of equity" is established in coalitions. A "norm of equity" occurs when the proportion of rewards that each person gets out of a relationship equals the proportion of resources that he or she put into it. For instance, if a person put only 4 bricks into a bridge that took 100 bricks to build, he or she should expect only 4 percent of the profits from a toll charged on that bridge.

Let us go back to the smoke-filled room and look at the nominating convention example. Gamson would say that each candidate would expect rewards proportional to his or her input. Hence, Robert's representatives should expect to receive a number of jobs on the election team proportional to the number of delegates they give to the winning coalition.

Which coalitions form. Gamson believed that people do not merely want to be in a winning coalition; they weigh other considerations as well. They wish to get as much of a reward for their part in the victory as they can. For instance, the person with 4 bricks has a choice between joining two teams of bridge builders. One team has designed a bridge that will take 100 bricks to build; the other team has designed one that takes only 50. In Gamson's view, the person with 4 bricks would want to join the team building the 50-brick bridge. By doing so, he or she can expect to receive 8 percent of the toll profits, because the 4 bricks constitute 8 percent of the total bridge. These profits would be higher than the 4 percent he or she could expect from the other team.

According to Gamson, in our nominating convention example each candidate's representatives would want to be in the coalition that gives them the greater proportion of jobs.

If Larry's representatives formed a coalition with Robert, the proportion of delegates is two to one. Larry has 40, and Robert has 20. Therefore, Larry's representatives would expect to get two-thirds of the jobs on the final election team. If, instead, they formed a coalition with Isabel's supporters, the proportion is 4 to 3. Larry has 40 delegates, and Isabel has 30. In this case, Larry's representatives could only expect four-sevenths of the final jobs. Clearly, Gamson would say, Larry's supporters would rather form a coalition with Robert than with Isabel. They would want to get the greatest rewards possible.

Now let us look at Isabel's representatives and consider what they would want. They would also want a coalition with Robert. With Robert, they would get three-fifths of the eventual jobs. With Larry, they would get only three-sevenths of those jobs.

So both Larry and Isabel want to form a coalition with Robert. Robert's supporters thus have the power to decide which candidate will get their votes. What would they want to do? They would want to form a coalition with Isabel. With Isabel, the representatives would get two-fifths of the election team jobs. If they joined Larry, they would get only one-third of the jobs.

Thus, Gamson would predict that Isabel and Robert would form a coalition. Isabel's representatives would get three-fifths of the jobs, and Robert's supporters would get two-fifths of them. Larry, who had come to the convention with the most delegates, would be the loser.

Several ideas follow from Gamson's reasonings. One is that a person with great individual power is not always the winner. We can see that in our example above. The candidate with the most individual power, Larry, cannot compete with the two candidates that have less individual power but a good reason to pool what they have.

Another implication of Gamson's theory is that the coalitions that do eventually form are the ones that can win by the smallest possible margins. The Isabel-Robert coalition will win with 50 votes to Larry's 40. Either of the other possible coalitions would have won by larger margins. Gamson would say that the coalition that forms is the one that wins using the smallest possible amount of resources. For this reason, we have come to call his proposal "minimum resources theory."

Study to Compare Game Theory and Minimum Resource Theory

We have looked at game theory and minimum resource theory. Which can best predict which coalitions will form? Vinacke and Arkoff (1957) performed a study to try to answer this question.

In their study, groups of three played the board game Parcheesi^TM. In this game, players roll a die and move a piece around a board. The first player who gets around the board is the winner and receives 100 points. The researchers assigned a weight to each player. The players could multiply the number that came up on the die by their weight and then move the piece that number of spaces. For example, if a player's weight were 4 and the number on the die were 5, the player multiplied 4 by 5 and moved 20 spaces.

Participants could form coalitions and move one piece. When they did this, they could add their weights together. For instance, if a player with a weight of 3 formed a coalition with a player that had a weight of 2, they would have a combined weight of 5. They could move their piece based on their throw of the die multiplied by 5.

Which coalitions form. To alter the situation, researchers gave different weights to players as the experiment progressed. They wanted to see which, if any, coalitions would form in varying circumstances.

One situation was analogous to our political convention example. In this case, players had the respective weights of four, three, and two. As you can recall, minimum resource theory predicts that the players with the weights of three and two would form a coalition. In contrast, game theory predicts that players would have no clear preference, and each of the three possible coalitions would form one-third of the time.

Which coalitions formed? The results supported Gamson's minimum resource theory.

Vinacke and Arkoff examined 90 cases. Out of these 90, the players with the weights of three and two formed a coalition 59 times. Clearly, the players preferred one coalition over another. This finding is consistent with minimum resource theory and not consistent with game theory.

Other situations also supported Gamson's ideas. They were more consistent with minimum resource theory than with game theory. For example, in one case the players had weights of three, two, and two. The people with weights of two formed a coalition in 64 of the 90 cases.

Utilization of rewards. Vinacke and Arkoff's study next examined how the winning coalitions would utilize their rewards. They told these winners that they had to decide how to divide among themselves the 100 points they got for winning.

Minimum resource theory predicts that the winners would use the norm of equity to divide their points. Hence, in the three-two coalition, the player with the weight of three would get 60 points and the player with the weight of two would get 40 points.

Game theory, instead, predicts that players would divide the points according to the norm of equality. Each player would get half the winnings, or 50 points.

The actual results of the study were about halfway between these two predictions. When players with weights of three and two formed a coalition, the player with the weight of three generally ended up with about 55 points. The player with the weight of two usually got about 45 points.

Komorita and Chertkoff (1973) attempted to explain this finding. They wondered how we could expect this outcome to follow from the negotiations between the two players. They reasoned that both the norm of equality and the norm of equity play a part in the negotiations.

In their view, the player with the greater weight would argue for the norm of equity because this would give him or her the most points. In fairness, however, this player would realize that he or she would not have won the game without the other player's contribution. The player with the lesser weight, on the other hand, would argue for the norm of equality. He or she would want the most points possible. This player, however, also would realize that the other player made a larger contribution to their victory.

The players would negotiate these points, each arguing for the norm of his or her preference yet each realizing that perhaps adjustments have to be made. In the end it would be natural for them to split the difference between the possible outcomes.

Basic Assumption of Theories

Both game theory and minimum resource theory have one assumption in common: that people form coalitions in response to the power each has and the rewards that each can get in a given situation. It seems straightforward. People seek the most efficient means to get the most rewards possible. Is this always true?

If the assumption is true, it implies that, if relative power among people changes, the coalitions among them will change in response. For instance, if each player's weight in the Vinacke-Arkoff game changes or if the number of delegates changes in our example of the political convention, the coalitions will alter.

Studies of the Basic Assumption

Studies have been performed to see if this occurs. Researchers have looked at coalition formation in games that are played repeatedly over time.

Lieberman ( 1964) conducted a study using a 40-trial game. Players could constantly change coalitions if they wished. They were able to negotiate between each trial by writing notes. Lieberman found that in some cases coalitions changed rapidly, with players deceiving and double-crossing each other regularly, particularly in the early trials of the game.

In other cases, however, players began to ignore the short-term advantages of switching coalitions. When a particular situation changed, they did not change their coalition partners. Instead, they began to stick with one partner through thick and thin. When these players negotiated, their notes often stressed the importance of trust and dependability. Some participants later told Lieberman that they came to realize the long-term advantages of entering into a coalition with a person who appeared to be trustworthy. They wanted someone they believed the third player could not "buy off." These players began to act more "trustworthy" themselves. They wanted others to perceive them as worthy of joining a coalition.

The basic assumption of game theory and minimum resource theory may not always be true. People sometimes look at issues such as long-term stability when they form coalitions. In these cases, they may not always follow the most efficient path to get the greatest short-term rewards.

Table 5.2 compares the predictions made by game theory and minimum resource theory. The predictions include which coalition is expected to form and how winnings are divided among coalition members.

Table 5.2	Predictions of coalition formation theories
Theory	Which Coalition Will Form?	How Will Winnings Be Divided?
Game theory	All winning coalitions equally likely	Equally among members
Minimum resource	Coalition that wins by smallest possible margin	Proportionally to what each contributes (equity)

Other Limitations

Even considering the issue of long-term stability, proposals such as minimum resource and game theory are limited. They basically see coalition formation as an issue of power and possible rewards, and this is not always the case.

Many other factors enter into coalition formation. For example, people usually form coalitions with people they like more often than with those they dislike. In decision-making groups in particular, members are often divided on substantive issues. They form coalitions based on agreements and disagreements concerning the issues. Relative power advantages and possible rewards usually do not prevail. They will not lead people who have strong disagreements to form coalitions.

Thus far in this chapter, we have discussed many facets of power. We have shown how people may gain and use power. We have shown how inequities in relationships between people can lead to differences in the amount of power that each has. You can recall, for example, the situation in which one person desires a relationship much more than the other does. In this case, the inequities in attraction cause power differences. We showed many other examples of relationships that can lead to power differences.

Further, this chapter has discussed how the structures that arise out of relationships can reinforce power differences. One example was the group of company employees in which a supervisor has the ability to give credible threats and promises. The corporation's structure reinforces the supervisor's power.

We have also described the way in which power affects coalition formation. Differences in power can help determine how people organize into coalitions. We saw an example of people playing a board game and looking at how much "weight," or power, each had before determining which coalitions to form.

One topic we have not discussed is the relationship between communication and power. It has been suggested that communication and power directly affect one another. We have seen how power relates to relationships, structures, and coalitions. Now we will examine how power relates to communication.

Researchers have proposed the idea that people with unequal amounts of power communicate differently. These differences are both verbal and nonverbal.

In other words, there appear to be "powerful" and "powerless" styles of communication. Further, people notice these differences in style. They seem to base their reactions and judgments about others partly on the style of communication.

If someone communicates in a "powerful" manner, others often judge that person to be "powerful." The way they then treat him or her follows from that judgment. Thus, powerful people tend to reinforce their position by the way they communicate. Their communication style often causes others to defer to them, and this, in turn, helps them maintain their power.

Verbal Communication

What is each style like? To answer that question, we will first look at the characteristics of verbal communication that are associated with power.

Characteristics of the "Powerless" Verbal Style

To understand the differences between "powerful" and "powerless" communication styles, we will focus on the verbal behaviors of the "powerless" speaker.

In general, it appears that people who are relatively powerless tend to speak as if they were unsure of themselves. They exhibit this idea in various ways. When compared with people who are relatively powerful, these "powerless" speakers tend to:

1. Use more "hedges"--explicit indications that they are not confident about what they are saying. These include phrases such as "kinda" or "I guess.

2. Ask more "tag questions"--questions added to the end of declarative statements. Tag questions make people look as if they need approval regarding what they are saying. An example is, "The weather's nice today, isn't it?"

3. Utilize rising intonations more often at the end of declarative statements. This makes their statements sound like questions.

4. Use more polite forms of speech. Relatively powerless people tend to make requests rather than give orders. They will say "please" and "thank you" more often, and they will curse less than relatively powerful people.

Study of effectiveness of verbal styles. Erickson, Lind, Johnson, and O'Barr (1978) studied the effectiveness of these verbal styles. They found that people did respond to the differences between "powerful" and "powerless" speech styles. They granted more power to the "powerful" communicator.

In the study, participants watched videotapes of a staged "trial." In each trial, a confederate posed as someone who was giving testimony after witnessing an accident. In all cases the confederate's testimony had the same factual content. In some videotapes, however, the confederate spoke in a "powerful" style and in others he or she used a "powerless" manner.

The results showed that the styles had an effect. The participants evaluated the "powerful" speaker as being more credible and attractive than the "powerless" one. When they saw a "powerful" witness, they also judged the plaintiff as deserving a relatively larger amount of compensation for the accident.

Nonverbal Communication

Now let us turn to the nonverbal behaviors of the different communication styles.

Researchers have performed many studies about the relationship between power and nonverbal communication styles. Most relate to situations in which people with different levels of power speak with one another.

Harper (1985) reviewed this research. The list we have compiled from his work does not focus on the "powerless" style, but outlines how the "powerful" speaker behaves in comparison with someone using a "powerless" style.

According to Harper's work, the research found differences in the following areas:

1. Body position and movement. Relatively powerful people tend to have a more relaxed body posture. They also position their bodies and limbs in asymmetrical ways. They usually make more arm movements and gestures. Their body positions tend to take up more room. For example, they may stand with arms akimbo.

2. Relative body position and distance. Powerful communicators are more likely to face away or even turn their backs on powerless people. Relatively powerless people, in contrast, usually face more powerful people directly. In addition, powerful speakers tend to approach the powerless more swiftly and to touch them more often. Powerless people, on the other hand, usually give more distance to powerful people.

3. Speech qualities. Relatively powerful people often speak for longer periods of time, and they speak more loudly and quickly. They also usually have a lower voice pitch than that of relatively powerless people. Powerful communicators tend to interrupt more, hesitate less, and laugh less than the powerless.

4. Facial expressions. Those who use the powerful style usually smile less and look less often at people who are not so powerful. The powerless, in contrast, smile more often and look at the powerful more often. The powerful tend not to alter the amount of time they gaze at someone, regardless of whether they are speaking or listening. Powerless people, on the other hand, alter the amount of time that they gaze at the powerful. When they are listening, they tend to pay attention to the powerful person's words by gazing at him or her a great deal of time. Yet, when they speak, they often do not look at the "powerful" person as much.

We will discuss this last finding in greater detail later in this chapter.

Earlier, we stated that people notice the differences between the "powerful" and "powerless" communication styles. Do individuals, however, react consistently when they see these styles? For the differences between the two styles to be important, people must notice them and react to them consistently.

A study by Lee and Ofshe (1981) suggests that people do react consistently.

In their study, participants read a description of a personal injury case. Based on this reading, the participants decided on a financial award to give to the injured person.

Next, the participants viewed one of three videotapes. Each videotape showed a group of three people discussing the case. The group's discussion followed a script. In each tape, two members argued for a financial award of $15,000 while a third person wanted an award of only $2,000.

Beyond this, the researchers controlled the communication styles that the group's members used. In one tape, the third member, who wanted the lesser award, behaved in a "powerful" style. He or she spoke loudly and quickly, with no hesitations, had a relaxed posture, and gazed steadily at the other members. In another videotape, this third person used a "powerless" communication style. He or she spoke softly and slowly, with many hesitations and ''filled pauses," such as "uh" and "ya know." He or she also had a rigid posture, with eyes directed downward. In the third videotape, the third group member behaved in a neutral manner.

The reasonings and arguments of the third person were always the same, regardless of style. The researchers wanted to see how the different communication styles of the third group member influenced the participants.

Each participant saw one of the three videos. Afterward, they each made a final decision about the financial award for the injured person. The researchers wished to discover whether the participants changed their decisions because of the videotapes. If they did, the study looked at whether the arguments of the third group member influenced that change.

It turned out that the "powerful" communication style had a consistent effect.

After observing the tapes that showed a "powerful" third group member, the participants changed their awards to be an average of $4,778 closer to what the third member argued was appropriate. When they saw the "neutral" third member on tape, the study participants still moved closer to the amount the third person argued for, but less so. They changed their minds to be an average of $2,426 closer to the award that the third person wanted.

In contrast, when participants saw the "powerless" third member on tape, they changed their minds to disagree even more with the powerless person. Their awards changed to become an average of $2,843 further away from what the third person argued was just.

Thus, the more "powerful" the communicator's style, the more participants changed their judgments to agree with his or her opinion.

It appears, then, that people with more "power" act in a more "powerful" manner. They do this in both verbal and nonverbal ways. Further, people who use a "powerful" manner tend to influence those around them.

These findings in and of themselves are significant for decision-making groups. They point out one way in which members can eventually control their group's discussions and influence its decisions. Such knowledge can be useful. For instance, if a group realizes that one member is using all the verbal and nonverbal behaviors of a "powerful" communicator, that group can be aware of what is taking place. If the person's arguments are faulty, the group can work not to be swayed by the "powerful" manner.

These findings are also important in a less obvious way. The implications go beyond the decision-making process and affect other areas relevant to group discussion. We will now turn to one of those areas--the role of gender differences.

Do women and men differ in their behavior that is relevant to communication?

Researchers have performed many studies to try to answer this question. They have found that several gender differences appear to exist (see Baird, 1976, for a review).

Conformity Differences

The gender difference that most directly relates to power involves the behavior of conformity. Research has found that women appear to conform to men more often than men conform to women. A study by Tuddenham, MacBride, and Zahn (1958) forcefully showed this.

In the study, a group of five participants sat in the same room. There were many groups like this in the study, and they had various gender ratios: five men, three men and two women, two men and three women, or five women. In the room, partitions separated group members.

Each participant answered a series of questions. The researchers falsely told each participant that he or she was the last of five people to answer the questions. They further showed a list of responses that they said the others had given. These were actually phony answers designed to be extremely different from the typical responses of most people.

The researchers wanted to see how much the participants' responses approximated the phony answers. This would show that the false answers influenced the participants. The findings showed that the more men in the group, the more group members conformed. The more men and the fewer women, the more the phony answers influenced both men and women.

Verbal Communication Differences

There are probably gender differences in verbal communication. However, it is not entirely clear what these differences are. For a long time researchers have believed that men talk more in groups than do women. However, in a review of ten relevant studies, Anderson and Blanchard (1982) claimed to have found no consistent differences between women's and men's talk time. A second long-standing belief is that there are content difference between men and women in their group discussion. Women were believed to specialize in maintenance-oriented content and men in task-oriented discussion. Consistent with belief, Anderson and Blanchard's review found that the content of men's talk tends to be proportionally higher in task functions and proportionally lower in maintenance functions than women's talk in groups.

However, one must be careful in interpreting this finding. One study (Nemeth, Endicott & Wachtler, 1976) found no gender differences in the number of maintenance functions during group discussion. If this finding is accurate, it implies that any differences between women and men during group discussion is found in task work, not in maintenance.

Lakoff (1973) has hypothesized that "women's speech" differs in style from men's. She claimed that women are more likely than men to use rising intonation while making declarative statements. They are also more likely to end declarative statements with tag questions. Further, she felt that women tend to use more hedges than men do and to speak more politely than men. However, it is not clear whether or not Lakoff's hypotheses are correct.

Nonverbal Communication Differences

Finally, gender differences seem to exist in nonverbal communication. Women tend to look more at men than men look at women. Women also look at others more when speaking than while listening. In contrast, men gaze at others about the same amount of time while listening and while speaking. Women also tend to take up less space than men.

How do these differences in male and female communication behavior relate to power in decision-making groups? With a couple of exceptions, the differences between the communicative behavior of men and women closely parallel the variations between the "powerful" and "powerless" communicative styles. Many scholars claim a close connection at work here. They hypothesize that women generally communicate in a relatively powerless manner.

Why might this be? Many researchers say that, in comparison with men, women overall have less power in our society (see Henley, 1977, for a discussion of this issue relevant to nonverbal communication).

As is usually the case in social science, not all the evidence related to this issue is clear. For example, in studies comparing "women's speech" with men's speech," researchers have not consistently found evidence to support Lakoff's hypothesis that women do speak less "powerfully" than men.

We have seen that men and women, in general, apparently differ in their communication behavior. We have further seen that women's behavior can be similar to the "powerless" speaking style. What process could lead to this outcome? More specifically, what happens in groups? Let us examine how researchers have looked at these questions.

General Ideas in Society

Researchers have looked at the general ideas in our society to help them find out why women so often use a "powerless" communication style in groups.

Women undoubtedly have less power than men in influencing our society. Relatively few women advance to positions of power in business or politics.

Further, studies show that people apparently hold ideas that reinforce this situation. Such studies have found that, when all else is equal, people evaluate women less favorably than they evaluate men. This is true whether men or women are doing the evaluating.

For example, let us look at a study by McKee and Sherriffs (1957). They asked participants to judge the "relative overall general worth" of men and women. When the researchers did not allow the participants to judge men and women as equal, an overwhelming majority of them rated men as either "somewhat" or "a trifle" superior to women. This was true for 36 out of the 44 female participants and for 36 of the 42 male participants.

McKee and Sherriffs also asked one group to look at a list of adjectives and to rate each as to whether it was more characteristic of men or of women. The researchers then asked another group to look at the list of adjectives and rate each as to how favorable or unfavorable it was. Overall, the adjectives that the first group matched with men and women had relatively favorable ratings from the second group. All the participants, however, both men and women, were more likely to attribute favorable adjectives to men and unfavorable ones to women.

The McKee and Sherriffs study is now more than 35 years old. General attitudes toward women may have changed since then, but their study has important implications. It helped researchers examine how these attitudes could affect the power structure of groups.

Expectation States Theory and Competence

Let us assume that, overall, people evaluate women more negatively than men. They also judge women to be less competent than men. If this is the case, expectation states theory (Berger, Fisek, Norman, & Zelditch, 1977) can account for the- process by which women become less powerful than men in decision-making groups.

As you can recall, expectation states theory claims that group members go through a process when they face a decision. They try to select the member of their group that is the most competent to make that decision, and they give that person power over the group. To do this, group members first try to find a person with skills relevant to the task. If they cannot, they then give power to a member who is the type of person they generally consider to be "competent" or "intelligent."

Hence, it follows that if group members generally feel that women are less competent than men, women will find it difficult to gain power in a group.

Many studies have looked at this hypothesis. The results support the idea that competency is an important issue when it comes to women and power in groups.

Studies of sex-typed tasks. One way to examine the issue of competency is to look at how groups approach "sex-typed" tasks.

In our society, certain tasks are "sex-typed." In other words, our society generally believes that men or women have skills that are specifically relevant to that task. For example, most people would think that building a bookshelf is a "male" task. Cooking a meal, on the other hand, is something many feel is a "female" task. Groups usually give power to men when the group is performing a "male" task, and they usually grant power to women if the task is a "female" one.

Many tasks, however, are not sex-typed. Decision making is not sex-typed; our society believes that both men and women make decisions. What happens in these cases? Expectation states theory would predict that general feelings about the competency of each sex would be influential. Let us say that the group members look on women more negatively than on men. In addition, they believe women are less competent than men. To the extent that these statements are true, the group will give less power to the female members than to the male ones. Also, the female members will be less able to influence the group than the male members can.

Lockheed (1985) found evidence of these tendencies at work. She reviewed studies of influence in mixed-sex task groups. Lockheed looked at eight studies of groups with tasks she considered male sex-typed. In all cases, men had more influence than women. Next, she examined six studies of groups with tasks that she judged to be female sex-typed. Of these six, women had more influence than men in three, men had more influence in one, and two showed no gender difference in influence.

Finally, Lockheed reviewed 32 studies of groups working on tasks that she judged to be non-sex-typed. These were neither "female" nor "male" tasks. In these 32 cases, men had more influence than women in 20. Women had more influence in only one study. In the other 11 studies, no gender difference occurred.

Thus, Lockheed's findings support the idea that, in general, the average person feels men are more competent than women. People relate the characteristic of being a man to competent task performance. The average group member reacts to this idea by giving men, rather than women, power.

Yet, when the task was "sex-typed" as female, women were able to gain some power. The issue of competency swayed the group members and altered the power structure.

Specific shows of competency. Another way to look at the issue of competency is to look at how group members react to women who demonstrate that they have knowledge that pertains to the task at hand.

We have seen how general ideas about competency hurt women's chances for gaining power in groups. Hence, for women to gain power in their groups, some factor must negate the idea that they are less competent than men. We have already discussed one such situation--a group that is performing a "female" task.

What happens when women demonstrate competence in their group's task? Studies have found that in this circumstance women can gain power.

Bradley (1980) had groups of five people discuss concepts relevant to journalistic plagiarism. The concepts included fair use, common law copyright, and infringement. The groups were composed of four male participants and a confederate who in some groups was male and in other groups, female. The confederate either showed knowledge about the concepts under discussion or admitted ignorance of them.

After this initial discussion, Bradley had the participants read a report about a case of journalistic plagiarism. She asked them to make preliminary individual judgments about the proper penalty for the plagiarist. Next, the group discussed the issue, and the confederate argued for an especially severe penalty. After this discussion, the participants made final individual judgments about the penalty.

Bradley looked at the results to find out if the participants had changed their minds about the penalty. If their final judgments were more severe than their preliminary ones, the confederate had influenced them.

This study, like the ones relevant to "sex-typed" tasks, supported the hypothesis of expectation states theory.

Expectation states theory would predict that our society takes for granted that men are competent. Thus, the male confederate would always have the same ability to influence others, whether or not he had shown competency relevant to the task--in this case, knowledge about journalistic plagiarism. On the other hand, the theory would predict that our society takes for granted that women are not competent. Hence, a woman needs to show herself competent before she can have influence.

According to the theory, then, the female confederate would only be as influential as the male confederate when she had shown that she was competent, by displaying knowledge about journalistic plagiarism.

This is exactly what Bradley found in her study, that women had to show themselves competent before the group gave them influence.

It appears that, overall, women can gain power in groups by showing that they are competent.

Women and Communication Styles

We have seen how the issue of competency can affect how much power groups give to women. When a woman shows herself to be competent, she can gain power.

Can women have more influence in groups in other ways? Studies have shown that women can gain power by using "powerful" forms of communication.

Study of verbal communication style. Bradley (1981) performed a study that examined communication styles. It was similar to the one we discussed above, regarding journalistic plagiarism. In this study, however, the confederates either used or did not use hedges and tag questions during their arguments. As you can recall, such verbal tags denote a "powerless" communication style.

What Bradley found was consistent with the findings of her earlier study. The male confederate always had the same level of influence, whether he used 'powerful" or "powerless" language. The female confederate, on the other hand, was as influential as the male confederate only when she used "powerful" language.

Study of nonverbal communication style. We can find similar effects in nonverbal communication. Earlier in this chapter we described how the "powerful" tend to look at the "powerless" in a consistent manner. They gaze at others the same amount of time while they are speaking and while they are listening. The "powerless," in contrast, usually look at the "powerful" much more while they are listening than while they are speaking.

A series of studies by Dovidio and Ellyson (1985) related gender and power differences in eye gaze patterns. They found links between competency and the "powerful" communication style.

Dovidio and Ellyson had mixed-sex pairs discuss three issues. With the first issue, the woman was more knowledgeable than the man. With the second, the man knew more than the woman, and with the third topic the two were equally knowledgeable. The researchers measured the proportion of time that participants looked at each other during speaking and during listening. Table 5.3 shows the results.

Table 5.3		Proportion of Time Gazing
	Women		Men
Relative Knowledge	Speaking	Listening	Speaking	Listening
More	58%	56%	53%	54%
Equal	41%	77%	47%	53%
Less	40%	74%	29%	51%

When both members of the pairs had equal knowledge about the topic, the men acted "powerfully." They looked at the women about the same proportion of time when they were speaking and listening. The women, on the other hand, acted "powerlessly" in this situation. They looked at the men a greater proportion of time when they were listening than when they were speaking.

When one member of the pair had more knowledge about the issue than the other, however, he or she acted "powerfully." It didn't matter if that person was male or female. Similarly, when one person had less knowledge than the other, he or she behaved in a "powerless" manner. Hence, the person who felt competent acted "powerfully."

In summary, all else being equal, women act less "powerfully" than men. As a consequence, group members give them less power than men when they are in mixed-sex groups.

If women demonstrate their competence or act in a "powerful" manner, however, they can acquire power in groups.

We must add research that suggests that even competent and "powerful" women may not be able to close the "power gap" entirely between themselves and their male groupmates. Nevertheless, studies have clearly shown how women should behave if they wish to establish themselves as "powerful" members of their group. Their actions may not bring them all the power they want, but they will be more powerful than if they behaved in a "powerless" manner.

In the last section, we discussed research that showed that women generally act in a less "powerful" manner than men in group settings. Further, unless women act in a particularly "powerful" way, they tend to have less influence than men in groups.

We saw the explanation that researchers have proposed to account for this. Expectation states theory shows how the process may work. Many people believe women are generally less competent than men, and these beliefs carry over into the group setting. If the group expects that women are not competent in a task, the members do not give power to women.

It is not surprising that the same circumstances pertain to ethnic minorities. If people do not believe that ethnic minorities are competent, they will not give power to those minorities in mixed-ethnic groups.

How does ethnicity affect power in groups? There has been far less research on this topic than on the question of how gender affects power in groups. Further, the research has not led to clear-cut results. Nevertheless the information suggests that the problems that ethnic minorities face as members of groups have much in common with those that women face.

In some circumstances, evidence indicates that members of ethnic minorities act in a less "powerful" manner than non-minority members. Katz, Goldston, and Benjamin (1958) formed groups consisting of two black males and two white males and studied the amount the men talked to one another. They found that the white males consistently talked more than the blacks, particularly to one another. When the black males did speak, they generally directed their comments to the whites.

Can a "powerful" communication style help ethnic minorities, as it helps women, gain power in groups? Evidence indicates that this is true. Some research, however, shows that members of ethnic minorities who act in "powerless" manner are treated differently than members of the "majority who act in a "powerless" way.

Tuzlak (1989) performed a relevant study. It was similar to the Lee and Ofshe (1981) study that we discussed earlier. Participants read descriptions of personal injury cases and made an initial financial award to the injured person. They then watched a videotape in which someone argued for a particular award. Afterward, they made final decisions about what award to give.

In Tuzlak's study, the videotape showed either a black male or a white male arguing for a particular financial award. The men used either a "powerful" communication style or a "powerless" one. In the "powerful" videotape, the man had a relaxed and open posture, he looked at the camera, he had fluent speech, and he did not have "nervous" gestures. In the "powerless" videotape, the man slouched and had closed posture; he looked down, spoke hesitantly, and used "nervous" gestures such as scratching his head and playing with a pen.

The actual content of the arguments remained the same in all the videotapes. After seeing the videotapes, the participants made final financial awards to the injured person. In addition, they rated the male they had seen on his status, degree of dominance, competence, and likability.

As in the earlier study, Tuzlak looked at how much the participants changed their financial awards to move in the direction of the arguments on the videotapes. In this way, he measured how much influence each man had on the participants. Expectation states theory would predict that the black male would need to use the "powerful" style to have as much influence as the white male, no matter whether the white male used the "powerful" or the "powerless" communication style.

This did not occur, however. Participants changed their financial awards more in the direction of the argument if they saw a "powerful" male than if they saw a "powerless" one. It did not matter if the "powerful" man were white or black. The "powerful" style always had more influence than the "powerless" one.

Still, the participants gave, overall, the lowest ratings to the "powerless" black male. He received lower ratings than the "powerful" black and white men and the "powerless" white man. This finding implies that, although "powerless" whites may not be any more persuasive than "powerless" blacks, people often still rate them as more effective communicators than "powerless" blacks.

Weber, a founder of modern sociology, defined the term "power" as the probability that one-person in a social relationship can carry out his or her will despite the resistance of the other person.

Scholars from the relational perspective have proposed theories about power that follow from Weber's definition. Lewin was one such scholar. According to his field theory, one person has power over a second person if the first can affect the second's potential movement in his or her life-space.

Other scholars using the relational perspective are Thibaut and Kelley. Their game theoretical approach distinguishes between "fate control" and "behavior control." Fate control occurs when someone can affect someone else's outcomes in a given situation. Behavior control takes place when an individual can affect another person's behavioral choices in a given situation. Theorists from the structural perspective have also contributed to our understanding of power. Blau, a social exchange theorist, was one of these scholars. He claimed that one person has power over another when three conditions take place. The conditions are that the first person has something the second person (1) wants badly, (2) cannot get from anywhere else, and (3) has nothing to give in return. Unlike game theorists, Blau feels that power must be one-sided.

Another theory from the structural perspective is expectation states theory. It proposes that, when group members face a decision, they follow certain steps. These steps are to try to identify a group member who is competent to make the decision and then to give that person influence over the group's decision. If the group cannot find someone who has the specific skill necessary to make the decision, the members give power to the person they feel is the most "competent" in general.

A final theory from the structural perspective comes from Bacharach and Lawler. They distinguish between two types of power in decision-making groups, "authority" and "influence." Authority exists when a group's structure gives a member the right to make decisions for the group. Influence comes about when members of the group who do not have authority nevertheless have an impact on group decisions.

Structural theorists have also proposed typologies of the basic "dimensions" of power. "Status" is a dimension that refers to the authority that comes from the positions that people have both in organizations and in society. "Resource control" is a term for the authority that people possess if they can reward or punish others. "Expertise" describes the influence that someone can get because of knowledge or skill in relevant areas. "Likableness" refers to the influence that can come out of a desire to be liked by particular people. "Trustworthiness" is a term for the influence that can arise when others perceive that a person is sincere and unselfish. Studies based on expectation states theory support the idea that several of these dimensions are connected with power. Other research using a modified version of the Prisoner's Dilemma Game leads to similar conclusions. These latter studies, however, also point out how important credibility is in the use of power.

Coalitions form within a group when two or more members pool their resources to gain power over the other members. Theorists are concerned with two aspects of coalitions. They wish to explain and predict which coalition will form, and they want to explain and predict how the coalition members will divide any winnings from their actions. Game theory predicts that it is equally likely that any winning coalition will form. It further holds that coalition members will divide any winnings equally among themselves, using a "norm of equality." In contrast, minimum resource theory predicts that the coalition will form that would give each member the most possible winnings. The "norm of equity" determines the coalition for each member. This norm holds that the proportion of winnings equals the proportion of investments.

According to research, minimum resource theory appears to predict accurately which coalition will form. Members apparently split their winnings in a manner halfway between the norm of equality and the norm of equity. Other research findings are inconsistent with both theories. These include the finding that members' believing in the importance of trust and dependability appears to increase the stability of coalitions over the long run. Also, apparently unrelated factors, such as interpersonal attractiveness and agreement on relevant issues, affect which coalitions will form.

Communication contributes to the establishment and maintenance of power inequities among people. There appear to be "powerful" and "powerless" styles of both verbal and nonverbal communication. Based on the judgments of observers, research suggests that people who use a "powerful" way of speaking have more influence than those who use a "powerless" style.

Women generally have less power in groups than men. Some scholars feel that women, as a group, use less "powerful" forms of communication than men use. If this is true, it may account for some of the gender differences in power. It also appears that people in our society evaluate women as a group less highly than they evaluate men as a group. If so, expectation states theory could explain why women have less power than men in most group decision-making situations. An exception is a group making a decision about a topic that is sex-typed as "female" in our culture, such as cooking. Women can also increase their chances of gaining power in a group by showing competence and by communicating in a "powerful" style.

Researchers have examined the topic of ethnic minorities in groups less than they have looked at women in groups. Studies suggest, however, that ethnic minorities face problems in decision-making groups that are similar to those women face. One similarity is that, in some circumstances, group members that are ethnic minorities apparently act in a less "powerful" manner than non-minority members. Ethnic minorities, like women, can gain power by communicating in a "powerful" style. Evidence indicates, however, that people treat ethnic minorities who act in a "powerless" manner differently from those in the "majority" who act in a similar way.