The examination of characteristics relating to leadership, such as enthusiasm, verbal skills

Chapter 11 - Leadership: Communication Approaches

THIS CHAPTER WILL DISCUSS:


1. How all group members can perform leadership functions.
2. The ways in which leaders emerge from previously leaderless groups.
3. The importance of group members' perceptions about leadership.
4. How charismatic leaders affect group performance.
5. Gender differences in leadership.

INTRODUCTION

In the last chapter, we described the trait, style, situational, and contingency approaches to leadership. All four have been very influential, and have lead to expansions of our knowledge about such issues as what types of personality leads to people becoming leaders, what type of method leaders should use, and in what situations are particular types of personalities and particular types of leadership methods best.

However, none of the perspectives described in the last chapter take communication into serious consideration. Yet, in the context of group discussion, leadership occurs through the process of communication. Therefore, to have a complete understanding of how leadership works, we also need approaches that tell us what role communication plays in the leadership process. In this chapter, we will consider four approaches to leadership in which communication performs a critical role. We will also talk about whether there are gender differences in leadership.

THE FUNCTIONAL APPROACH TO LEADERSHIP

The functional approach to leadership first appeared at about the same period of time (the late 1940s and early 1950s) that scientists began questioning the trait and style approaches. As we discussed last chapter, some scientists responded to the problems with the trait and style approaches by adopting the situational and contingency perspectives.

As you recall, each of these approaches has a unique perspective on leadership. For instance, the trait approach is concerned with the permanent characteristics of a leader. The style point of view maintains that the method a leader uses is most important. The situational perspective looks for differences in leadership style across different circumstances. Finally, the contingency approach examines leadership by looking at the association of trait and situation. These four points of view all have one thing in common, however. All of them, in the end, are concerned with a particular person in a leadership role.

The functional perspective, in contrast, in concerned with the behaviors of a group. Specifically, the functional approach examines the behaviors that allow the group to reach its goals. In Chapter 8, we discussed how behaviors serve functions that help a group solve its task and maintenance problems. Certain behaviors serve as leadership functions. For instance, leadership functions include actions that serve to help the group perform its task and clarify the group's goals. They also include behaviors that help the group interact effectively and give members social support.

The functional approach maintains that a group need not have only one particular member who performs these leadership functions. Any group member can perform them. This means that any member can lead. If we accept the functional perspective, we have the potential to uncover which behaviors help lead a group toward its goals. These behaviors help the group if at least one member performs them.

For example, a group of mechanics volunteers time to help low-income families repair their cars. They call themselves the Saturday Morning Car-Tuners. For the group to perform their task, members must ensure that the group is organized. Everyone has to know what they are supposed to do to get the cars repaired. In addition, for the group to maintain itself, members must ensure that everyone in the group feels appreciated for their efforts. If not, members may leave the group, and it might not survive.

Bales: The "Equilibrium Problem"

All functionalists make a distinction between task and maintenance behaviors. Some functionalists go even further with the idea. These theorists claim that the demands of task and maintenance leadership are so unique that different people must assume leadership of each. These researchers might say, for example, that one person in the Saturday Morning Car-Tuners should lead the group in repairing cars. However, another person should lead in handling good communication and morale-building within the group. The group should not ask one person to be responsible for both areas.

In Chapter 8, we described the "equilibrium problem." This hypothesis ties in with our present discussion. As you recall, Bales proposed the idea of the equilibrium problem in 1953. He believed that the requirements of the task and maintenance functions in a group are opposed and that too much attention to either causes problems for the other. What the group needs to do is create equilibrium, or balance, between task and maintenance functions.

Also in Chapter 8, we noted one property that could help groups maintain their equilibrium: alternate group discussion between task and maintenance statements. A second property that can help is a division of the major leadership responsibilities. A group with this property will have one person responsible for performing the task leadership functions and another person responsible for the maintenance leadership behaviors. In such a group, one person will assume the role of task leader and will be responsible for making the group work on its task. Unfortunately, because the requirements of the task often conflict with the members' personal goals, this leader will cause resentment to build up among group members. For example, Fred, a member of the Saturday Morning Car-Tuners, has the personal aim of working only on challenging repair problems. He wants to sharpen his skills. The group goal, however, is to work on any car problem, even if it is minor. The task leader must ensure that Fred works on all types of car problems. Fred may become resentful, however, if he rarely gets to work on challenging jobs and instead has to work on any job in order to satisfy the group's task of helping everyone.

The group responds to this tension and resentment by generating a maintenance leader. The maintenance leader has the responsibility of draining the tension from the group. For instance, someone in Fred's group may come up to Fred and praise his work and pass on a compliment from someone Fred has helped. This person would be the maintenance leader of the Saturday Morning Car-Tuners.

Research

In Chapter 8, we discussed Bales's research, in which his groups solved "human relations" problems. After their discussion, the group members rank-ordered one another on five questions:

1 - The extent to which each contributed the best ideas for solving the problem.

2 - The extent to which each did the most to guide the discussion.

3 - The extent to which they liked each other.

4 - The extent to which they disliked each other.

5 - The extent to which each was the group leader.

This allowed Bales to determine for each group which member was the "idea person," which member was the "guidance person," and so on.

Part of the results of this research was reported in Bales (1953), with the rest to be found in Bales and Slater (1955). It revealed that in about half of the cases, members judged their group as having one member leading in "ideas" and a different member doing the most "guidance." In the other half of the groups, the same person was evaluated as performing the most of both task functions. Further, the person judged to be the "group leader" was the same person as the "guidance" specialist 78.6% of the time and the "idea" specialist 59.3% of the time. In either case, these task leaders tended to be the most talkative members of the group.

However, in more than 70 percent of the cases, the "best liked" group member was someone other than the "idea" or "guidance" leader. This member was usually the second of third most talkative member. Further, the "group leader" was also chosen as "best liked" only 14.3% of the time.

Bales believed that there was a clear distinction in his groups between the task leader or leaders (the idea and guidance specialists) and the maintenance leader (the best liked member). This distinction between leaders was not necessarily apparent during a group's first meeting. After the first meeting, members chose the same member as both the "best liked" and as the "idea" leader 64.4% of the time. Thus, in their view, both task and maintenance leadership functions were performed by the same person. However, in subsequent meetings, the functions of task and maintenance became progressively more divided. By the fourth meeting, the same person was evaluated as both "best liked" and best on "ideas" only 10.7% of the time. Similarly, the odds that a group member was chosen as both the "best liked" and as the "guidance" leader fell from 40.6% after the first meeting to 17.9% after the fourth.

Bales thought this occurred because of the tension brought about by the task leader's actions. Remember that according to the equilibrium hypothesis, task work causes tension among members. For Bales, it follows that the person who is believed to be most responsible for that tension, the task leader, would be blamed for that tension and disliked by the other members. In fact, the most talkative member of Bales's groups, who was usually rated either the "idea" or "guidance" leader or both, received the most votes for "disliked member." The group tended to like a different member the most. This person was usually either the second or third most verbally active and tended to be rated low on "ideas." We can consider this liked person the maintenance leader. Thus, Bales found that groups tend to distinguish between the two leaders.

Criticisms of research. Bales concluded from this research that different group members perform different functions. One person behaves as a task leader, and another takes over as the maintenance leader. As described earlier, Bales felt this occurs because the task leader's actions cause tension that can only be relieved by the actions of another person, the maintenance leader. There are however some problems with these conclusions. First, Bales's conclusions about leadership were based on members' ratings of one another. However, Bales did not examine the actual communication that occurred in his groups' discussions to see if, for example, the person judged as the "guidance" leader actually did the most guidance during the discussions. This failure leads to some ambiguity concerning the results. For example, Bales showed that the correlation between the person the group liked best and the person or persons the group considered idea or guidance leaders declined across meetings. But we do not know why. One interpretation of these results is that the group's most well-liked member did less task work as time progressed. A second idea is that the group's most well-liked person did very little task work all along, and as time went on the group liked its task leader less and less. Without an examination of who in the group performed which functions, we have no way of knowing which of these possibilities is correct.

In addition, certain theorists believe that the results of Bales's research were due to his participants being uninterested in the discussion and wanting to leave as quickly as possible. Why would this change Bales's results? By nature, decision-making tasks usually take a long time when members are interested in making a good decision. If we accept the idea that Bales's students did not want to spend time on the discussion, it makes sense that they would resent members who appeared interested in the task. Interested members would make the group take longer to complete the discussion. Thus, the groups did not like the task-oriented members very much.

This idea has led to a more general claim. Researchers have proposed the idea that the more task-oriented group members are, the more they like task-oriented people. Thus, it is not difficult for a task-oriented leader to also become the maintenance leader when the group is interested in its task. Turk (1961) conducted a study that supported this claim. In the study, nursing students participated in decision-making groups considering issues in which they were interested. After the discussion, they rated one another. In this case, the students liked their task leader.

It follows from this idea that it is more likely that people will split up the task and maintenance functions in low task-oriented groups, rather than highly task-oriented ones. For example, it may be that Fred's group of volunteer mechanics is not very interested in repairing cars. Perhaps they simply enjoy getting together to watch television in the garage. If this is true, the task leader may not be popular. The group may not like the head mechanic very much if he or she pushes the group to work on cars. Hence, the group needs another person to make membership enjoyable for the mechanics and handle the maintenance side of group activity.

The resulting corollary to the idea above is that highly task-oriented groups do not require two people to handle task and group maintenance behaviors. This, however, does not invalidate the distinction we make between task and maintenance functions. It also does not discredit the further distinction between two types of task functions, substantive (generating and evaluating ideas) and procedural (moving the group along the decision-making process). Each still has quite different requirements. For instance, it may be true that the head mechanic in Fred's group is the leader for both task and maintenance behaviors. However, he or she cannot lead both behaviors the same way. The head mechanic might need to be tough and organized when it comes to the task of repairing cars but joking and kind when it comes to giving praise to the members.

Benne and Sheats: Functional Roles

Back in Chapter 8, we discussed Benne and Sheats's (1948) essay listing the functional roles that group members can perform during discussion. This is perhaps the clearest statement of the functional approach to leadership. At that time, as described earlier, most researchers concentrated on the position of "group leader." In contrast, the authors argued that "leadership" consists of a set of functions helping groups perform their tasks satisfactorily and getting along well with one another. Members share leadership to the extent that they perform these functions. A group member is performing task leadership when she or he is performing either substantive roles, such as the initator, opinion giver, and elaborator, or procedural roles such as the coordinator, orienter, and procedural technician. Similarly, a group member is performing maintenance leadership when he or she performs roles such as the encourager, harmonizer, and compromiser.

As we discussed at that time, Benne and Sheats did not make any claims about which roles are the most important or when in the discussion process particular roles need to performed. We believe that there is perhaps a presumption that the more that members perform task and maintenance roles, the more successful the group will be. Other functional theorists would dispute this presumption. Rauch and Behling (1984) argued that when a group's task is very clear, the group does not need very much task leadership and gets upset if their leader is too task-oriented. Task performance suffers as a result. Further, they felt that while a moderate amount of maintenance leadership encourages group members and helps them perform their tasks better, a lot of maintenance leadership can be too much of a good thing. When group members perceive continued task leadership to be redundant, they can become dissatisfied.

Ancona and Caldwell: External Functions

No functional theorist doubts that we must make a distinction between task and maintenance functions in groups. Some functional theorists however believe there are other types of functions that are as important as task and maintenance. Ancona and Caldwell (1988) claimed that groups within formal organizations also have to be concerned with their relationships with other groups and individuals outside of the group. These other groups and individuals may be part of the organization, or they may be outside of it. In either case, there are leadership functions involved in maintaining these relationships. Ancona and Caldwell called these "external functions."

Based on extensive interviews with members of organizational groups, Ancona and Caldwell found four basic types of external functions:

1 - Scout activities. These are involved in bringing into the group both the information and resources that the group needs to perform its task. Scout functions include learning about the environment in which the group does its work, getting information that is relevant either to the group's current task or possible later tasks, and getting feedback about the group's performance.

2 - Ambassador activities. These are involved in getting information and resources from the group out to other groups or individuals. Ambassador functions include opening up channels with these outside parties, informing them about the group's progress on its tasks, coordinating with the outside parties when a task is being performed together with them, and persuading or motivating the other parties to do what the group wants them to do.

3 - Sentry activities. These are involved in controlling the amount or type of information and resources that come into the group. The group can either let information and resources enter the group as is, modify it in some way, or keep all or part of it out of the group.

4 - Guard activities. These are involved in controlling the amount or type of information and resources that leave the group. The group can either deliver it immediately, decide to wait to deliver it until some later time, or refuse to deliver it at all.

Ancona and Caldwell provide some hypotheses about when the group will perform more or less of these functions. At the beginning of the task, the group needs to perform scout and ambassador activities in order to determine the requirements of the task, establish relationships with important outside parties, and obtain needed resources. During the performance of the task, the group needs to perform sentry and guard activities in order to protect the group from becoming overloaded with unnecessary and distracting information and resorces. At the completion of the task, the group returns to scout and ambassador activities so as to deliver their completed work to the parties that need it and get feedback about their performance.

General Conclusions: Functional Approach

As described earlier, the functional approach to leadership provides a unique view of the leadership process. Other approaches view leadership in terms of a person taking on the leadership role. In contrast, the functional perspective accounts for leadership in terms of the behaviors that help a group perform its task, maintain its cohesiveness, and interact with its environment. As anyone can perform those behaviors, every member can participate in its group's leadership.

Having said that, it should be clear from the research findings we have discussed that the majority of leadership functions are usually performed by a minority of the members of the group. In fact, in a lot of cases, one member does the bulk of both the task and maintenance leadership functions. Sometimes that member has been assigned or elected to the leadership role, and so it is not surprising to find that member actually leading the group. However, in many cases, groups operate without an assigned or elected leader. When one member comes to perform the leadership functions in a previously "leaderless" group, we call that member the group's emergent leader. Bales's groups were an example of leaderless groups in which task and maintenance leaders emerged over time.

Research about the process by which leaders emerge from previously leaderless groups can be said to comprise its own perspective toward leadership. We will turn to an examination of this approach next.

THE EMERGENT APPROACH

The "Zeitgeist Theory"

Earlier, we described the philosophical view of history called the "Great Man Theory of Leadership." It implied that gifted people determine history. The emergent approach to leadership is a descendant of a proposal that conflicted with the "Great Man" point of view. It has come to be known as the "Zeitgeist Theory of Leadership." Zeitgeist is a German word meaning "the spirit of the time." This competing theory was that history determines leadership, not vice versa. Different places and times have unique requirements for their leaders. People can assume leadership positions only if their talents meet the requirements of a particular time period. If conditions change, there will be different requirements, and other people will become leaders.

According to the Zeitgeist Theory, Julius Caesar and Napoleon were not inherently great men. Instead, they were fortunate. Fate placed them in circumstances in which they had relevant talent to offer. Place them somewhere else in history, and they would not have been leaders. For instance, place them in early twentieth-century India. Would Napoleon's military talents have served the people of India best at that time? India was not looking to create an army and conquer foreign lands at that time. Instead, India needed a person whose ideas were very different from Napoleon's. As we know, Mohandas Gandhi was the leader whose talents were uniquely able to organize the people of India. He found a method whereby people without military strength could effectively work toward the changes they wanted.

Advance in Leadership Theory

As with the functional point of view, the emergent approach came to prominence in the late 1940s and early 1950s when the problems with the trait and style approaches became evident. As with the situational approach, advocates of the emergent perspective claim that the group's situation determines the group's leadership. However, whereas the situational approach is concerned with the style elected and assigned leaders should adopt in different situations, the emergent approach is concerned with predicting who will emerge as leader in a previously leaderless group and explaining why some people emerge rather than others.

Advocates of both the situational and emergent approaches understood that a situation consists of two parts. One part is the task that the group must perform. The other part includes the needs and desires of the group members themselves. Hence, they realized that group members themselves help to determine who their leader is. This concept was a significant advance over earlier approaches to the study of leadership. Both the trait and style approaches did see that leaders influence group members. However, both ignored the possibility that members may influence their leader in return. Scientists who held the emergent point of view came to an even further realization. They saw that even leaders with strong claims for power must receive their groups' approval in order to be effective. These claims for power include expertise, legitimacy, and the like. These realizations on the part of scientists who held the emergent viewpoint opened three new avenues of thought.

1. Leader emergence. The trait and style approaches limited their inquiries to the ways in which appointed leaders affected group behavior. In contrast, the emergent methodology shifted attention to the manner in which leaders emerge from previously leaderless groups. In other words, the researchers wanted to know the ways in which leaders gain influence over other group members.

2. Method for assuming leadership. Research was performed to discover any differences between the behavior of groups with assigned leaders, elected leaders, and emergent leaders.

3. Deviance and leader acceptance. Theorists attempted to describe how a leader can still be accepted by a group after he or she has performed deviant actions.

The research using the emergent approach involved all three of these issues.

Leader Emergence

As we have noted, the first research issue involved the ways in which leaders emerge from groups. To begin, it is clear that amount of communication is the most important factor in determining which group member will emerge. Most simply, the more a person communicates, the more likely it is that the person will emerge as leader. Reviews of literature by Stein and Heller (1983) and Mullen, Salas, and Driskell (1989) found that the amount that a person talks in a group correlates with group member judgments of task leadership in the range of .5 to .7. It should be noted that the correlation between amount of talk and maintenance leadership is far lower, at about .15. In the Bales groups discussed under the functional approach (Bales & Slater, 1955), the biggest talker was judged as "group leader" 50% of the time.

However, it stands to reason that what a person says should have as much impact on other group members' judgments of that person's leadership as how much that person talks. Surprisingly, many studies have found that the content of a person's talk has no effect on whether the person is judged to be a leader. However, most of these studies were performed in groups of students that were just formed into groups and were performing tasks that did not affect them, such as "Lost on the Moon" or a "human relations" problem. Other studies have been performed using groups that had existed for many weeks or months and were performing tasks that did affect them, such as groups of students who were working on a class project for which they would get graded. In these studies, not only how much a member talked but also the content of their talk was related to whether or not they emerged as a leader.

There is reason to believe that members are particularly likely to be viewed as leaders if they perform procedural functions during group discussion. Recall that in the Bales groups, the member judged as "guidance" specialist was also judged as "group leader" 78.6% of the time. Two other studies support this idea.

Baker study. Baker (1990) studied the first hour of discussion of eight groups of students in the process of performing two class assignments. She related the content of those discussions to judgments made by group members of who emerged as group leaders, who did not emerge as leaders but maintained high status within the groups, and who did not emerge as either leaders or high-status members.

Baker concluded that emergent leaders tended to be concerned with procedural matters during discussion. Sometimes they also contributed ideas to the group, but they tended to not offer their opinions very often. In contrast, those who had high status but were not considered the group's leader concentrated on substantive issues; they contributed ideas and opinions. Finally, members who were neither leaders nor high in status concentrated on opinion giving.

Ketrow study. Ketrow (1991) videotaped a three-member "group" trying to solve a problem about white-collar crime. Each member of the "group" had a pre-assigned role in performing their task. One member specialized in procedural work, one in substantive, and the third in maintenance. One hundred fifty research participants watched the videotape and then made judgments about the members of the "group." Of these participants, 128 chose the procedural specialist as the group leader. Although the substantive specialist was chosen as the most influential member by 95 participants, only 17 chose that person as the group leader.

Thus there is good reason to believe that verbal activity is related with emergent leadership. Of course, not every talkative person becomes a leader. As just described, the content of talk also has an impact on group member's judgments of one another's leadership. However, other factors may be involved. If verbal activity is associated with emergent leadership, perhaps nonverbal participation is too.

Baird study. Baird (1977) showed the plausibility of this proposal. He videotaped the interaction of 10 five-person groups. In doing so, he discovered that members' perceptions of leadership correlated with various nonverbal signals. With arm and shoulder gestures, the correlation coefficient was .45. In addition, the coefficient with vertical head movements was .27. There was a smaller correlation of .15 with shifts in posture, absence of horizontal or "no" head movements, and absence of frowns. Eye contact, smiles, and absence of hand or finger movement were also slightly related to leadership. These had a correlation of .09. Taken as a whole, it appears that groups judge people as leaders when their movement styles are active but not nervous. Also, group members believe that leaders give nonverbal signals that imply positive evaluations of the other group members.

Relevant Input Variables

The research we have discussed thus far shows that process (verbal and nonverbal communication) is related with output (leadership emergence). However, as we have discussed throughout this book, a valid theory of group discussion also needs to consider relevant input variables that affect group process. There are several input variables that can have an impact on the amount of leadership-relevant communication that occurs during group discussion. Next, we will consider three of these input variables; seating position, approval, and task-relevant knowledge.

Seating Position

Several studies have shown that seating position has an effect on emergent group leadership. For example, Strodtbeck and Hook (1961) performed research using 69 twelve-person mock juries. Those sitting at the ends of the jury table were the most talkative members of the groups, and were most likely to be chosen to lead the mock juries. Research has shown that people sitting in those positions also receive the most communication from other group members. It is clear that the ends or the "head" of a table are considered to be the "leadership seats" for a group discussion.

Seating position also has effects on the order in which group members speak. Steinzor (1950) examined the order in which members of two 10-member groups sitting in a circle spoke to one another. He found that, when a particular group member made a statement, the next statement tended to come from a group member sitting on the opposite side of the circle. Thus, comments are likely to go back and forth across opposite seating positions in the group.

Howells and Becker study. Howells and Becker used this idea in a study published in 1962. They reasoned that they could set up a "power differential" by placing fewer people on one side of a table than on the other side. They reasoned, as Steinzor had found, that comments go back and forth across a table. If this happens, the people on the less populated side proportionally receive and make more comments than those on the other side. The scientists formed 20 five-person groups. Two members, or 40 percent of the group, sat on one side of the table. Three participants, or 60 percent, sat on the other side.

If all else is equal in such a situation, we would logically expect that the number of leaders from each side should match their percentage in the group. Thus, among the 20 groups, 8 leaders should emerge from the less populated side of the table, matching the 40 percent. Similarly, 12 leaders, or 60 percent, should emerge from the more populated side. Did.this occur? No. Instead, only 6 leaders emerged from the more populated side. This was only 30 percent. In contrast, 14 leaders came from the less populated side. This was 70 percent, a great deal more than a logical prediction would have calculated.

Thus, the results supported the contention of Howells and Becker. They had been able to create a power differential. The less populated side did have participants who spoke more often than the other members. Consequently, the less populated side produced more leaders than a logical outcome would have predicted.

Thus, research has shown that certain seating positions can make it more likely that a group member will become a leader. Other studies, however, have shown that this effect is weak and can be canceled by more powerful factors. For example, Bass and Klubeck (1952) found that seating position had no effect on leadership judgments in discussion groups with preexisting status differences among members.

Approval

A second factor that can affect emergent group leadership is whether or not a member receives approval for what she or he says. A study by Pepinsky, Hemphill, and Shevitz (1958) shows this impact.

Pepinsky, Hemphill, and Shevitz study. Pepinsky, Hemphill, and Shevitz conducted a study with four-person groups. The groups task was to build "products" out of tinkertoys that could be sold at a "buyer's table." The group had to purchase the tinkertoys, build the products and sell them at a profit at the buyer's table.

Each group included two confederates who acted in either of two different ways. In one condition, the confederates waited until the real participants had taken some leadership actions and then disapproved of what they had done. For example, they would say that they did not like bossy people who told them what to do. In the second condition, the confederates showed approval of the real participants' actions. For example they would say that they liked people who could get things going and come up with a good plan. The researchers found that the participants whose actions had received approval from the confederates performed many more subsequent leadership actions than the participants whose actions who had been rejected.

Task-Relevant Knowledge A third factor that affects emergent leadership is whether group members have knowledge relevant to the group's task. Hemphill (1961) reported an unpublished study by Shevitz that demonstrated this effect. Shevitz organized three-member groups with two "experts" and a third, non-expert member. One of the experts was a member of the local Amateur Radio Association, and the other had experience in statistics courses. The group had two tasks; assembling the components of a circuit from a schematic diagram, and a statistics problem. The "radio" expert performed the most leadership acts during the first task, while the "statistics" expert did the most during the second.

Method for Assuming Leadership

Emergent leadership researchers studied a second issue. They wanted to know whether there are differences between groups whose leaders get their position in different ways. Does it matter whether leaders are assigned from the outside, elected by the members, or allowed to emerge from group discussion? Contrasting hypotheses could be proposed concerning this issue. It could be that groups with assigned or elected leaders perform better than groups with emergent leaders because emergent-led groups begin discussion without knowing who is in charge. It also could be that groups with emergent or elected leaders perform better than groups with assigned leaders because assigned-leader groups are not involved in the choice of leader and as a consequence are less satisfied with their leadership. Several studies have been performed that are relevant to this issue.

Mortensen study. Mortensen (1966) analyzed the leadership emergence process in six groups that met three times. Three of the groups had assigned leaders, and three did not.

The three assigned leaders performed the most leadership actions during the first group meeting. However, by the third meeting, only one on the three was still leading his or her group. In the other two, the assigned leader had been supplanted by another member to whom the group had responded more positively.

Mortensen concluded that the group's approval of an assigned leader is important. Assigned leadership can help a person quickly assume the leadership role, but only if the group approves of him or her. In contrast, assigning a leader can slow down a group and make it difficult for a leader to emerge if the group disapproves of the assignment. In this case, leadership emergence would have been faster if there had been no leadership assignment.

Goldman and Fraas study. Goldman and Fraas (1965) found that the way in which a group acquires its leaders is important to the group's performance. They assigned leaders to groups through four methods. One way was that the group elected a leader by vote. The second method was that the experimenters selected a leader based on the person's ability. The third way was that the scientists arbitrarily chose a leader, and the last method involved groups that had no assigned leader.

The researchers then had the groups play "Twenty Questions." They compared the success rates of each group. The researchers found that groups with elected leaders performed the best. The performance of groups that had leaders appointed based on ability was close to the best success rate. Leaderless groups were a distant third. The groups with the worst scores were the ones with arbitrarily appointed leaders.

Hollander and Julian study. Hollander and Julian (1970) decided to compare the performance of elected and appointed leaders. The researchers told groups and the leaders of the groups that one of two criteria had determined leadership. They led their participants to believe that either the group had elected the leader or the experimenters had appointed him or her. In either case, the leader supposedly had important abilities for the position.

Hollander and Julian found that "elected" leaders had slightly more influence on their group's opinions than "appointed" leaders. They also were far more willing to deviate from their group's ideas. However, their position was far more precarious. Members lost confidence in "elected" leaders if they failed or if the group considered them incompetent at their task. In contrast, "appointed" leaders had to both fail and have the group consider them incompetent before the group would reject them.

Larson study. Larson (1971) compared nine groups whose members agreed on whom they felt had emerged as leader with five other groups whose members did not agree on who led the groups. Larson called the first nine groups "stable leadership" groups and the other five "unstable leadership" groups. The stable leadership groups spent more time on substantive issues and less time on procedural and irrelevant topics than the unstable groups. Further, the leader initiated the most substantive ideas in the stable groups. In addition, the stable groups spent more discussion time, per idea, on the topic that the leader had initiated, in comparison with topics the other members initiated. Thus, it appears that groups with stable leadership were more prepared to take on the their task than groups with unstable leadership.

Looking at these studies as a whole, it appears that competence and group acceptance are the most important factors in how a group will behave with a certain leader. It does not necessarily matter whether a leader is elected, emergent, or appointed. A group can behave well and be successful under any of these conditions. As long as the group accepts the leader and he or she is competent, the group can function well with the leader.

Deviance and Leader Acceptance

As we have seen, the emergent approach emphasized the importance of a group's acceptance of their leader. Emergent leadership researchers studied a third issue. They attempted to describe how a leader can still be accepted by a group when he or she has performed deviant actions.

One of the most important influences on how much a group accepts any member of a group, including their leader, is the amount of conformity the member displays. When a group member regularly conforms to the group's norms, the other members come to think that the member has the group's best interests in mind. Therefore, the other group members will highly trust and accept the conforming member.

When a group is making a decision, however, an important part of its leader's job is to propose creative solutions.

Creativity and Deviance

Any act of creativity in a group setting goes against the norms of the group. As a consequence, by definition, creativity is deviant. Simply proposing a new course of action, for instance, is a deviant act. How can a group member be creative without having the group dislike his or her deviant ideas?

Researchers have long known that group members that create new proposals are highly respected by others in the group. However, the group respects these creative people only if they have shown enough respect for group norms in the past to have gained the group's trust. Experimenters have found that successful group leaders are both the greatest conformers and the most significant deviators when it comes to group norms. Hollander (1958) created a banking metaphor in an attempt to explain why this is so.

According to Hollander, a group member's positive status is a result of an accumulation of acts that the other members look upon highly. Most important among these valuable acts are those that conform with group standards. As a member conforms, he or she builds up a kind of "bank account" of status and trustworthiness in the eyes of the other group members. The account contains what Hollander calls "idiosyncrasy credits." The credits correspond to the amount of deviant behavior that the group will accept from a given member. A person "cashes in" the account when he or she wants to be deviant. The more credits in the account, the more a person can deviate without losing the group's trust.

For example, Jane is in a group of singers. She goes along with the group norms, singing what the leaders pick out and rehearsing as she should, for years. One year she wants to sing a certain song during a festival. The song does not appeal to the group, and they think that it would be strange to have it on their program. Jane very much wants to include it, and she stands her ground. Luckily, Jane has a lot of idiosyncrasy credits in the group. She has been a loyal and true member for a long time. People listen to what she has to say, even though she is being very deviant in her behavior. In the end, they decide that Jane must have a good reason for wanting the song so badly, and they do let her include it. Jane may have cashed in all her "credits," but the group listened to her and still trusted her when she was finished.

Thus, a person's conformity leads group members to trust him or her. They perceive that any deviance on the part of this person is sincerely intended to be in the group's best interest. However, a group can apply pressure on a deviant to conform if his or her account becomes "overdrawn." This can happen when the group members begin to doubt whether the deviant has allegiance to the group. For instance, Jane's "account" with her group may very well be overdrawn after her request for the song she likes. If she were to soon ask for another "odd" song, the group might reject her proposal and also pressure her to be a "good" member again. The successful leader is someone whose normal conformity is great enough to allow occasional but significant deviance. This leader is able to deviate when the situation calls for it and still keep the group's trust.

General Conclusions: Emergent Approach

As the research we described earlier shows, the emergent approach has had some success. It did leave behind some valuable insights into the process of leadership emergence. In addition, it was able to show how important it is that groups approve a leader. Rejected leaders have little legitimacy. The approach further revealed what can happen when leaders receive their positions in different manners. They can be elected, emergent, or appointed. The emergent research has shown how different circumstances can affect the success of leaders, based on how they have become leaders.

THE PERCEPTUAL APPROACH TO LEADERSHIP

One of the most important legacies of the emergent approach is the idea that a leader has successfully emerged from a group when the members of the group believe that person to be the group's leader. Thus, the emergent approach emphasizes the notion that group members' perceptions about leadership are critical to the process of leader emergence.

The perceptual approach to leadership grows out of that notion. It is an attempt to understand how the average group member judges the leadership abilities of other members. Hence, it focuses on an important aspect of the leader emergence process

The perspectives we have discussed differ mainly in only one point. This is in regard to which factors they claim are important. Respectively, we have seen approaches toward leadership that maintain that a leader's success is due to his or her traits, his or her style, or the situation of the group. Further, we have discussed viewpoints that claim that the most important factors affecting leadership are the behaviors of all group members or the interaction between a leader's traits and the situation of the group. Despite these differences, scientists who advocate these approaches are united in their belief that they can objectively define "leadership" as a variable.

Advocates of the perceptual approach do not share this belief. They believe that leadership exists primarily within the mind of each group member. Thus they believe that the way to study leadership is to examine what group members think leadership is. It is consistent with the perceptual perspective toward groups that we discussed in Chapter 1.

Impression Formation Process

The perceptual approach views leadership as an "impression formation process." We discussed the process by which personality impressions are formed in Chapter 3. Let's review it now. For example, Sue wants to form an impression of Charlie; so she watches how he behaves. Sue particularly notices attention-grabbing, or salient behavior, on Charlie's part. Sue then decides whether his character is responsible for his behavior or if he is just reacting to the situation. If she decides that Charlie is responsible for the way he acts, she forms an impression of him. This impression starts with a trait that describes the behavior she has observed. The impression then builds, by adding other traits that Sue believes are associated with Charlie's first trait.

"Halo" and "Horns" Effects

If Sue observes something she believes is a "good" trait, she will tend to believe that Charlie has more "good" traits. Scientists call this tendency the "halo effect." If, on the other hand, Sue thinks Charlie has a "bad" initial trait, she will build up her impression by assigning more "bad" traits to him. This is the "horns effect." Finally, Sue will judge whether she likes Charlie based on her impression.

Biases enter into our impressions because of the halo and horns effects. As we have discussed, once we assign a trait to a person, we tend to further assign related traits. We can do this because we have a preexisting idea of which traits are associated. Such a process of judgment often can be wrong. It is not always true that clumsy people are stupid, for example. For instance, in Chapter 3 we described how the halo effect often causes people to make "good" judgments about persons who are physically attractive. People tend to say that attractive people are intelligent, mentally stable, and so on, without seeing any relevant evidence that this is true. Clearly, attractive people are not always "good" in all ways.

Impression Formation Process and Leadership

The way in which we form impressions and make judgments about the leadership ability of our fellow group members works in the same way as any impression formation process. First, we watch the other members of our group to see which leadership behaviors they generally exhibit. We are particularly attentive to the most salient of those behaviors. Next, we decide whether each person's character is responsible for any leadership behaviors or whether the situation is responsible. If we decide that a person is responsible for his or her actions, we form an impression of the person as a leader. Our impression will include not only the behaviors we have actually observed but also the traits that we believe are associated with leadership.

For example, we see that Joan speaks her mind readily. We think that is what a leader should do. We then may come to believe that Joan also has integrity, spirit, organizational skills, and other traits that we think leaders should have, even if we do not see her exhibiting them. We may then decide that Joan is truly the leader of the group.

Behavioral Salience

We have noted that impression formation begins when an observer notices a particularly salient behavior. Behavioral salience is an important part of the process by which group members come to judge one another's leadership. In fact, some of the variables involved in leader emergence work the way they do because they make a person's behavior particularly salient.

Amount of communication. When someone is talking, we are likely to pay attention to them. Thus, the more a person talks during group discussion, the more attention they receive from the other members. Given this, it should not be surprising that talkative members are more likely to emerge as leaders than non-talkative members.

Further, anything that increases a talkative member's salience would increase the likelihood of that member emerging as group leader. For example, consistently with the material on deviance and minority influence we discussed in Chapter 6, a small subgroup within the larger group is likely to gain the attention of the entire group if they are deviating from the group consensus. Thus, the behavior of a talkative member of that subgroup would be particularly salient to the rest of the group, and especially likely to emerge as a leader. In their review of literature about the relationship between amount of talk and leader emergence, Mullen et al. (1989) found research evidence for this effect.

Seating position. One of the reasons that the ends or head of a table are the "leadership seats" is that a person who sits at one of those positions becomes perceptually salient to the other members of the group. Perceptual salience can increase the odds that a member will become group leader for two reasons. First, they are more likely to be in the normal line of sight of other group members. The other group members are then more likely to direct their attention to them. Second, placing group members where they will receive the comments of other members is likely to lead them to talk frequently. As discussed, frequent talking will likely result in other group members judging them to be leaders.

Implicit Leadership Theories

As we have described, if an observer decides that a person is responsible for their behavior, the observer begins to form an impression of the person. This impression begins with a trait that describes the observed behavior, and then grows through the addition of other traits that the observer believes are associated with the first trait.

But where do these beliefs about the association among traits come from? They come from what researchers call implicit personality theories. An implicit personality theory is a person's beliefs about what traits and behaviors go together. For example, people's implicit personality theories usually associate "good" traits, such as intelligence and attractiveness, together. This is the reason why "halo effects" occur in impression formation. Similarly, implicit personality theories normally link "bad" traits, such as clumsiness and stupidity, with one another. This is the source of "horns effects" in impressions.

Research has shown that people have beliefs about the attributes that the "ideal leader" should possess. We can call these beliefs implicit leadership theories. Our own research (Pavitt & Sackaroff, 1990) has shown that implicit leadership theories tend to include traits such as "forceful" and "enthusiastic." These theories also include behaviors such as "states the group's procedure" and "encourages group member participation." Thus, these are included in the list of attributes that most people feel that the "ideal leader" should exhibit.

Group members use these implicit leadership theories when they form impressions of one another's leadership. For example, a group consists of Leon, Marvin, and Bram. Leon believes that an ideal leader who "encourages group member participation" is also "enthusiastic." He saw those two attributes as "going together" in an ideal leader. This judgment carries over into how Leon judges his fellow group members. For instance, if he were to think that Marvin exhibited a high level of "encouragement" behaviors, he would also judge Marvin as "enthusiastic." If Leon made a negative judgment about someone, he would connect similar negative attributes. If Leon judged that Bram exhibited a low level of encouragement behaviors, he would also think that Bram had low levels of the trait of "enthusiasm."

Leon is basing one judgment on another. He first observes Marvin and Bram and judges how much they encourage group members. As we have noted, Leon then believes that "encouragement" and "enthusiasm" go together in an "ideal leader." Therefore, he uses this belief and his previous judgment to decide how to assign the trait of "enthusiasm" to Marvin and Bram.

Further, group members use their implicit leadership theories as a basis for evaluating one another's leadership skills. If we form an impression of a group member that is similar to our preconceptions of the "ideal leader," then we would judge that group member to be a good leader. Analogously, if our impression is very different from our implicit leadership theory, then we would evaluate that member to be a poor leader.

We can use our example of Leon, Marvin, and Bram again to show this. Leon thought that Marvin had high levels of "encouragement" and "enthusiasm" attributes. Thus, Marvin's attributes were similar to those in Leon's vision of an "ideal leader." Hence, Leon evaluated Marvin as a good leader. In contrast, Leon thought that Bram had low levels of the attributes of "encouragement" and "enthusiasm." Thus, Bram was unlike Leon's vision of an "ideal leader." Accordingly, Leon, judged Bram to be a poor leader.

Research Showing Impression Formation Process

There has been some research exploring the impact of the impression formation process on group members' perceptions about their group's leadership. Our own research is an example.

Pavitt et al. study. This research has been described in two different reports (Pavitt & Sackaroff, 1990; Pavitt, Whitchurch, McClurg, & Petersen, 1995).

In the study, we presented the participants with a list of eight behaviors and eight traits that are part of most people's implicit theories of leadership. The participants then judged the extent to which the "ideal leader" would exhibit the behaviors and have the traits on the list. A month later, we put the participants into groups of three to five members. The groups then performed a decision-making task. After they made their decision, the members rated one another using the list of attributes they had used one month earlier. They also evaluated one another's leadership skill.

The results showed evidence that the participants' preconceptions about the "ideal leader" carried over into other areas. One area involved the impressions that the group members had of one another. The research looked at which traits and behavioral attributes the participants saw as "going together." There was a strong relationship between what they judged as "going together" in one another and what they saw as "going together" in an ideal leader. The correlation was about .7. In other words, if a participant believed that "stating the group's procedure" and being "organized" go together for the "ideal leader," they are likely to think the two attributes "go together" for other members of their group. This finding suggests that the members used their preconceptions as a basis for forming their impressions. Further, the study results showed that the participants' preconceptions about the "ideal leader" also influenced their evaluations of leadership skill. There was a strong relationship between how similarly they judged the "ideal leader" and one another and how good a leader they evaluated one another to be. The correlation was once again about .7. In other words, the more they thought a group member's attributes were similar to the attributes of the "ideal leader," the better a leader they judged that member to be. This result implies that the group members used their implicit leadership theory as a basis for evaluating one another's leadership skills.

It is important to note that these were groups that were just formed the study. The same study was also performed with groups that had existed for several weeks before the members made their decisions and judged one another. Researchers found these groups different from the shorter-lived ones. They looked again at what attributes members saw as "going together" in ideal leaders and which attributes they put together when they judged one another. Researchers found that the relationship between these two judgments was not as strong in these groups as it had been in the shorter-lived groups. In the longer-lived groups, the correlation measurement for these two judgments was only about .3.

Therefore, in newly-formed groups, members' judgments of one another are strongly influenced by their implicit leadership theories, leading to halo and horns effects in these judgments. In longer-lived groups, members come to know one another as individuals, and their preconceptions have less effect on how they judge one another. Instead, they gain knowledge about one another's unique personalities, and this knowledge starts to influence their judgments more than their preconceptions.

In other words, over time Leon might not rely so heavily on his vision of an "ideal leader" when judging Marvin and Bram. He might not use it as much as a basis for making judgments. For instance, as he gets to know Bram better, Leon may come to judge that Bram has a high level of the trait of "enthusiasm" even if he still believes that Bram is low in "encouragement" behaviors.

Judgmental Biases

The same processes that result in halo and horns effects also cause biases in judgments about leadership. As we have just discussed, halo and horns effects occur because people have ideas about which traits go together. People who work in groups similarly have ideas of which behaviors are associated with leadership. These ideas bias their ratings of specific leaders. The halo and horns effects can be very strong. Studies have even shown that people can rate imaginary "leaders" the same way they rate real leaders.

Rush, Thomas and Lord study. For example, recall our discussion of the Leader Behavior Description Questionnaire (LBDQ) from the previous chapter. Rush, Thomas, and Lord (1977) asked a group of participants to imagine a department supervisor of a "large Midwestern corporation." They then asked them to judge this imaginary person and fill out the LBDQ. The results showed that their "imaginary" judgments closely approximated the types of ratings that people make when they fill out the LBDQ for real leaders. The general leadership styles of consideration and initiating structure were clearly distinguished in the "imaginary" ratings. This result occurred even though the participants had seen no real behaviors on which they could base their judgments.

This finding implies that imaginary judgments are based on our ideas of which attributes are associated with leadership. Therefore, judgments of actual leaders are similar to fictional ones. This could well mean that the way we judge a leader is, in great part, a result of how we think a leader "should" act. For example, we may see Jack do certain things that we think a leader should do. We fill out the LBDQ about him. The LBDQ asks questions regarding behaviors that we have not seen Jack perform. For instance, he has never encouraged teamwork. However, we believe that Jack would encourage teamwork if it were appropriate, so we fill out the LBDQ accordingly. Thus, the halo effect has made Jack into perhaps more of a leader than he really is.

Implications for Leader Emergence
All this research leads to the conclusion that we judge people as good or poor leaders to the extent to which their actions are and are not similar to our preconceptions of how leaders should lead. There are other implications of these studies as well. Earlier we described how behavioral salience helped to explain how some of the variables involved in leader emergence work the way they do. The same applies for implicit leadership theories.

Content of communication. At that time, we discussed how behavioral salience is an important reason why a group member's amount of talk is associated with their odds of emerging as a leader. However, content of talk is also associated with these odds, and the effect of implicit leadership theory is important here. Although a person who talks a lot will gain other group members' attention, if the content of that person's talk is irrelevant to the task, then the group members' impression of that person will not be similar to their beliefs about the "ideal leader," and they will judge the person as a poor leader. In contrast, a big talker whose discussion content is task-relevant should lead the other group members to form an impression close to the "ideal" and evaluate that member's leadership highly.

Seating position. We have noted that the ends or the head of the table are the "leadership seats" for group discussion. We have said that one of the reasons for this is that these positions are visually salient and thus provide for those sitting in them the attention of the rest of the group. Another reason is that people have developed expectations of where leaders are supposed to sit during group discussions. The ends or the "head" of a table are believed to be the "leadership seats" for a group discussion. These expectations work just like implicit theories in influencing people's judgments about group leadership. A study by Pellegrini (1971) shows this effect.

Pellegrini study. Pellegrini took five photographs of a group of five women sitting around a table. The members of the group rotated like a volleyball team, so that each woman sat in each of the five positions at the table for the different photographs. Next, Pellegrini showed research participants one of the five photographs and asked them to make judgments of each of the five women. If seating position had no effect on these judgments, they would be similar among the women sitting in each position. However, Pellegrini found that the women sitting at the head of the table was consistently rated highest in persuasiveness, dominance, and leadership.

General Conclusions: Perceptual Approach

It might seem that because the perceptual approach is not involved with recommendations for who should lead a group or how a leader should lead, that it has no practical implications for group discussion. This is false; the practical applications of the research using the perceptual approach are clear. Remember that if we see a person as a group leader, for whatever reason, we will treat him or her as a leader. Thus our perceptions about one another determine who emerges a leader in our group.

The process of impression formation leads to predictable biases as people judge one another. There are certain behaviors we associate with leadership. We expect a group leader to show these behaviors, regardless of whether they are correct in a particular situation.

In addition, we are biased toward perceiving a group member's leadership behavior as something that indicates a permanent leadership character in that person. Thus, if a member acts as a leader in one circumstance, we expect him or her to do so again in a different situation. If the leader does not act as we expect, according to our bias, we may unjustly become unhappy with his or her performance. It is important that we are conscious of the biases that are inherent in forming impressions of one another. We must constantly remind ourselves that these biases lead us to have expectations about our group leaders that nobody could realistically fulfill.

THE CHARISMATIC APPROACH TO LEADERSHIP

As you recall from the previous chapter, the trait approach to leadership began with the premise of the "Great Man Theory of Leadership." This theory claimed that there are certain people who appear to have special leadership skills that have seemed to set them apart from the average person. As discussed there, researchers' attempts to find traits that distinguished "born leaders" from everyone else were unsuccessful.

However, history is full of political and religious leaders such as Mohandas Gandhi and Martin Luther King Jr. who seem to fit the "Great Man" idea. They have called forth extraordinary levels of dedication and effort from their followers. Back in Chapter 5 we described the work of Max Weber, a founding father of sociology, in defining the concept of "power." Weber also worked on understanding the process by which some leaders can inspire such devotion and commitment from followers. Weber proposed the idea that some leaders are perceived by their followers to possess a trait he called "charisma." The term "charisma" was originally defined as the ability to perform religious miracles. Weber used the term to refer to leaders who appear to have exceptional, almost superhuman abilities to perform difficult feats and succeed in seemingly impossible situations.

In the past few decades, some theorists have extended the idea of "charismatic" leadership to the context of organizations. There have been well-publicized examples of organizational leaders who seem to arouse the same type of response in their followers as Gandhi and King did in theirs. Examples include Steven Jobs of Apple Computers, Lee Iacocca of Chrysler, and Mary Kay Ash of Mary Kay Cosmetics. Although there are differences among these and other cases, a few similarities are clear. In each case, the leader was able to inspire unusual effort and performance from their subordinates. Because of this inspiration, Jobs and Ash turned their ideas into corporations worth hundreds of millions of dollars, and Iacocca took a large corporation on the bring of insolvency and converted it into a large profit maker.

Many reports have been published listing the strategies and tactics of successful charismatic leaders. Sometimes these reports have implied that charismatic leadership is the key to organizational success. However, it should be clear from everything we have written about leadership that such sweeping generalizations are likely to be wrong. In many if not most cases, the success of charismatic leadership is short-lived. In fact, as we shall describe, the characteristics that can help make a person a charismatic leader often include the seeds of their eventual downfall. Lee Iacocca, for example, eventually became more concerned with improving his public image nationally than with the operations of Chrysler. As a result, Chrysler once again began struggling. Although Iacocca (and Chrysler) eventually righted themselves, there are many cases in which charismatic leaders have led once-successful corporations to ruin.

Having said this, research suggests that charismatic leadership can lead to organizational success. Bryman (1992) reported that measures of charismatic leadership tend to correlate with job performance at about .3. Although this is not a very high correlation, it must be considered that the success of charismatic leadership depends on the situation in which the leader and the members of the leader's group of subordinates are in.

In fact, a number of theorists have proposed very similar theories of charismatic leadership (see Bryman, 1992, for a good review). These theories are comparable to Fiedler's contingency theory of leadership in some ways. They propose that certain traits lead some people to adopt what could be called a "charismatic" style of leadership. There are certain situations in which this style leads to successful performance and some in which it backfires. However, this theory differs from Fiedler's in that communication is specifically described as part of the leadership process. Thus charismatic leadership theory fits the input-process-output model.

A Theory of Charismatic Leadership

Leader characteristics.According to most theorists, charismatic leaders have a number of traits in common. First, they are high in the "Nepo," or "need for power" characteristic we defined in Chapters 1 and 4. Second, they have great self-confidence. Third, they are unusually willing to take risks. Fourth, they have a strong conviction in the correctness of their beliefs.

Of course, having relevant traits is not enough. Advocates of the trait approach ignored the fact that leaders are most successful when performing tasks they are skilled in or knowledgeable about. Successful charismatic leaders generally have a great deal of experience in the organizational area in which they are managing. Lee Iacocca had many years of experience in the automobile industry before he took charge of Chrysler.

Finally, charismatic leaders have developed a "vision" of what their group or organization should stand for. This vision is often abstract, but must be simple enough for subordinates to understand. For example, Steven Jobs had a quest to improve education through making computers available to the average person. Mary Kay Ash wanted to help women learn to become independent and empower themselves.

Thus, charismatic leaders have particular traits, relevant experience, and a vision.

Leader style. Charismatic leaders perform a number of characteristic behaviors that we can call a "charismatic" style of leadership. First, they are able to successfully communicate the content of their vision and confidence in their subordinates' ability to attain it. They do this through both verbally and nonverbally. Verbally, charismatic leaders tend to use a number of different strategies when addressing subordinates. Shamir, House, and Arthur (1993) have listed some of the strategies.

1. They often refer to shared values and moral justifications. As discussed, Jobs described Apple's goal as bringing computers to schools and the average person. He also referred to Apple as if it were David taking on IBM as an evil Goliath. Thus the competition between the two multimillion dollar corporations was represented as a fight between good and evil.

2. They tend to refer to distant goals rather than near ones. For example, Jobs was more likely to discuss what Apple Computer could eventually do rather than what would occur tomorrow.

3. They tend to emphasize the collective identity of the subordinates. The needs of the corporation are always more important than the desires of its individual members for the charismatic leader.

4. They tend to express high expectations for the subordinates. Their task is described as very difficult, and success as requiring great effort.

5. They tend to show great confidence in the ability of their subordinates. No matter how hard the task as they have described, the subordinates have the "right stuff" to succeed.

Second, charismatic leaders deliver this content in an animated nonverbal style. They have great variety and dynamism in their voice, and use gestures, movements and eye gaze actively when talking.

In addition, the charismatic leader must "live" their vision. They must seem to practice exactly what they preach. If a charismatic leader expects their subordinates to put in sixty hour weeks, the leader should put in seventy. If the charismatic leader wants subordinates to look upon the organization as their family, they must take on the role of maintenance leader. If a charismatic leader wants the subordinates to trust their own abilities and judgments, the leader must trust the decisions and actions that the subordinates take.

Situational factors. This style of communicating and behaving appears to be successful only in the right situations. First, the circumstance should be particularly stressful or uncertain. This would occur either in a newly-formed organization or in an older organization that is having problems that threaten its survival. Second, the subordinates need to be creative and dedicated. What can happen is that subordinates of this type who have become apathetic under other leadership styles are inspired when a charismatic leader takes over. Third, the leader must have the power to actually follow through on their vision. For example, if a middle-level manager has certain goals that those higher up in the organization do not value, then the manager will not be able to see their ideas through.

Results.When a charismatic leader is able to communicate their vision and their confidence in their subordinates' ability to attain it, and when the situation is as we have described, the result in subordinates is

1. unusual confidence in and loyalty to leadership

2 strong identification with the organization and the leader's vision

3. high motivation and commitment

4 increased self-confidence.

This can result in tremendous effort on the part of the subordinates and great success for the organization.

However, when the situation does not meet the requirements we listed above, the charismatic leader will not be successful. Further, although many of the traits of the charismatic leader are important in helping the leader succeed, they can and often do lead to eventual failure. Their high self-confidence and need for power can lead them to be extremely autocratic and self-serving. They can have little tolerance for subordinates disagreeing with them, and often blame others for their own errors. In the long-run, subordinates often tire of the charismatic leader, lose their dedication, and may even work to undermine the leader's authority. Further, if the charismatic leader has superiors in the organization, they often tire of the leader's style and become unwilling to tolerate any mistakes on the leader's part. There have been many reported examples of charismatic leaders who were not at the top of their organization being forced to leave the organization after only three or four years.

Explanation for the Effect

In attempting to explain why charismatic leaders succeed or fail, theorists have been influenced by the contingency view that leader characteristics and situational factors are input variables affecting the outcome of job performance. They also have also considered the process variables of communication and behavior linking the input and output. In addition, they have described the importance of the subordinates' perception of the charismatic leader in a way reminiscent of the perceptual approach. In fact, they have used the same psychological factors in their explanations as the perceptual theorists.

Salience effects. As we have discussed, perceptual theorists stress the impact of the salience of a group member's behavior on the odds that they will emerge as a group leader. Analogously, charismatic advocates talk about the importance of salience in the process by which a leader becomes viewed by followers as charismatic. The charismatic leader's animated communication style and inspirational communication content serve to catch the subordinates' attention, as does their attempts to behave consistently with their vision.

Implicit theory effects. We have also described the role of implicit theories in the process by which group member form leadership-relevant impressions of one another. One theorist, Conger (1989), has particularly emphasized the importance of the impression formation process on subordinates' evaluations of their leader as charismatic. After viewing the leader's salient behavior, subordinates may then use their implicit leadership theories to develop a particularly strong and positive impression of the charismatic leader. Like the "ideal organizational leader," the charismatic leader may be viewed as confident, visionary, trustworthy, and expert. Analogously, if things start going bad, the subordinates might change their impression of the leader to include characteristics the "ideal leader" does not have. The charismatic leader might now be viewed as stubborn, dictatorial, and back-stabbing.

Although there is great agreement among scientists about this theory, much of it is speculative, and little of it has received research examination. A welcome addition to the work on charismatic leadership has come from two experiments testing it.

Howell and Frost study. Howell and Frost (1989) gave business students an individual task to perform in which each acted like a business executive making a decision. Two other confederates also "worked" on the same individual task side-by-side with the real participant. The task was described by a "leader" displaying one of the following leadership styles:

1 - Task-oriented. The leader described the task and told the participants to follow the instructions, show good judgment, and be methodical in their work. The task-oriented leader sat behind a desk to distance themselves from the participants, used a "moderate" tone of voice, intermittent eye gaze, and "neutral" facial expressions (no smiling or nodding).

2 - Maintenance-oriented. The leader described the task and told the participants to relax, work at their own pace, and not to worry about the outcomes. They sat at the front edge of the desk to lessen interpersonal distance, used a "warm" tone of voice, direct eye gaze, and "friendly" facial expressions (smiling and nodding).

3 - Charismatic. The leader described the task, emphasized the importance of the results to the future of the students' coursework, encouraged them to be creative, and expressed confidence in their ability. They alternated between sitting at the front edge of the desk and pacing around the room, used a "captivating" tone of voice, direct eye gaze and "animated" facial expressions.

Then the leader left and the task began. At one point, the leader returned and spoke again. The task-oriented leader reminded the participants to focus on the task, the maintenance-oriented leader thanked the participants for performing it, and the charismatic leader reiterated confidence in the participants' abilities to perform the task. At the end of the task, the participants reported their feelings about their experience on a questionnaire.

The findings showed support for the value of charismatic leadership. The charismatic leader condition was highest on task performance, with the maintenance leader condition lowest. Turning to the questionnaire measures, the charismatic and task-oriented leader conditions were highest on participants' satisfaction with their own performance and the charismatic and maintenance-oriented leader were highest on participants' satisfaction with the leader.

Kirkpatrick and Locke study. Kirkpatrick and Locke (1996) noted a weakness with the Howell and Frost study. Howell and Frost had the charismatic leader both exhibit a certain style of nonverbal behavior and express a particular type of communication content. The results showed that this combination led to more satisfactory results for the charismatic leader conditions than for the task-oriented or maintenance-oriented. However, there is no way of knowing whether it was nonverbal style or verbal content that led to these findings.

Kirkpatrick and Locke then performed a study to distinguish between these two possible factors. They asked participants to perform a routine assembly task, in which they inserted pages into notebook binders according to a set of instructions. Before the task began, a leader either expressed a "vision" (commitment to quality) and confidence in the participants' abilities, or merely gave factual information about how paper is made. Half of the "vision" and half of the "factual" leaders used the charismatic nonverbal style as Howell and Frost had defined them. The other half used Howell and Frost's task-oriented nonverbal style. Kirkpatrick and Locke found that of the two possible factors, only the verbal expression of vision had an effect on quality of performance. This effect was indirect, working through changes in participants' goals and self-confidence.

General Conclusions: Charismatic Approach

A lot has been written about the effect of charismatic leadership in organizations. Much of what has been written has been uniformly positive and overly optimistic about its effects on subordinates' attitudes, self-confidence, and performance. There is a danger that this attention will lead to charismatic leadership becoming the newest "fad" in organizations. Organizations are already advertising for "charismatic leaders" and trying to get their present managers to act more charismatically.

All this activity ignores the fact that charismatic leadership, like any other style, is only successful in certain situations. It also ignores the fact that "naturally" charismatic leaders only too often alienate both their subordinates and their superiors, and any gains in organizational performance are often short-lived. Further, there is no reason to believe that charismatic leaders have the best interests of everyone in mind. As much as Mohandas Gandhi and Martin Luther King Jr. fit the definition of charismatic leader, so did Adolph Hitler and Mao Tse Tung.

Having said that, it is clear that in the right circumstance, charismatic leadership can be a very successful method of management. As with the style and contingency approaches, the trick is for the organization to determine the situation they are in and to choose the right type and style of leadership for that circumstance.

PUTTING THEORY INTO PRACTICE

In this and in the previous chapter, we covered a lot of ground concerning leadership. We have described many theories, some of which are contradictory with others. The reader might imagine that it may be difficult to come up with practical suggestions from this confusion. It is true that one cannot provide a specific "How To Lead" guide. However, there are a few general recommendations that the leader should keep in mind.

First, leadership is behavioral. To lead a group is to perform certain leadership behaviors or to act in a particular leadership style. There are certain things that a person must do to be a procedural leader, a substantive leader, or a maintenance leader. Benne and Sheats (1948) have told us what those behaviors are. There are certain things a person must do to help their group interact successfully with outside groups and individuals. Ancona and Caldwell (1988) have told us what those behaviors are. There are certain things a leader must do to initiate structure or show consideration. Bass (1990) has told us what those behaviors are. To a large degree, these behaviors can be learned, and so anyone can become a leader.

Second, leadership is situational. There is no one best way to lead. But there are ways to lead that are best for particular types of tasks and for particular types of groups. Hersey and Blanchard (1969) and Vroom and Yetton (1973) have provided guides for helping us choose the right style for particular circumstances. The charismatic leadership theorists have told us when the charismatic style will help and when it will hurt the group.

Third, we must keep in the back of our minds the thoughts of those who have doubts about these ideas. For example, if the legacy of Bales's (1953) work is correct, then task-oriented leaders will be rejected by groups who do not care about their task. If Fiedler (1978) is right, then it may be very difficult for a person to learn how to perform a leadership style that is inconsistent with their personality. As the failure of the trait approach tells us, nobody is born to lead. It takes hard work, good sense, and sensitivity to be successful at leadership.

Before moving on to other topics, there is one more important issue about leadership to discuss.

GENDER DIFFERENCES IN GROUP LEADERSHIP

Gender and Power in Groups

In Chapter 5, we discussed ideas about power in groups. We outlined conceptions of how our society defines power and stated that, according to those conceptions, men tend to have more power in groups than women do.

Further, we described two reasons for this power inequity. These explanations are also useful as we approach the topic of gender differences in group leadership. Before we move on to that topic, let us review what we said about power and gender.

Perceptual
One reason for the power inequity in groups is perceptual. It involves how our society perceives men and women. Cultural stereotypes often cause both men and women to believe that women are, overall, less competent than men. Competency influences who gains power in groups. Hence, as expectation states theory explains, the common belief that women are less competent than men affects the power structure of groups. Members often give men power over their groups' decisions.

An exception often occurs when a group is performing a "female sex-typed" task, a task about which people believe that women have more relevant knowledge than men. For example, if a group is discussing how to sew a dress, the members often give power to women.

Behavioral
The second reason that women often have less power than men in groups is behavioral. As we discussed, our society has overall conceptions of a "powerful style" of communication. Women tend to act in a less "powerful style" than men in group settings. They tend to make fewer proposals, and interrupt less often than men do. They also tend to look at men more than men look at women when they communicate. In general, society judges that these behaviors are relatively low in power.

Hypothesis of Gender Differences in Leadership
How does power relate to leadership? We can assume that power and leadership often go together. We have seen gender differences when it comes to power in groups. Hence, it stands to reason that gender differences might also influence group leadership.

We shall discuss two overall hypotheses concerning gender and leadership in groups. With each, we will again focus on related perceptual and behavioral reasons.

Overall Preference for Male Leaders
First, it makes sense to hypothesize that, generally, women are less likely than men to emerge as group leaders. Again, there are both perceptual and behavioral reasons for this hypothesis.

The first reason is perceptual and follows from what we discussed under the "perceptual" approach to leadership. We described how people have beliefs about the traits and behaviors of the "ideal leader." We further showed that those beliefs influence how group members judge one another's leadership abilities.

Let us support that people have beliefs about how men and women are "supposed" to act in groups. People may believe that men are supposed to act the way that ideal leaders act. In contrast, they may feel that women are not "supposed" to act that way. If this were true, they would probably judge men to be better leaders than women. For example, someone believes that an ideal leader should act "competently." Further, this person feels that, overall, men act more "competently" than women. Thus, this person will look for a man to head the group rather than a woman.

Further, there is reason to believe that people do not expect women to assume leadership positions. Earlier in this chapter we discussed Pellegrini's (1971) study, in which participants were shown photographs of five women sitting around the table. The participants consistently rated the women sitting at the "head" of the table as higher in leadership than the other four women. Porter and Geis (1981) performed an instructive variation of Pellegrini's study.

Porter and Geis study. Whereas Pellegrini showed participants photographs of five-women groups, Porter and Geis showed their participants pictures of groups including either five women, five men, three women and two men, or three men and two women. In the groups with three women and two men, a women was sitting at the head and the men were next to her; the groups with three men and two women had a man at the head and the women next to him. Porter and Geis asked the participants to rate the members as Pellegrini had.

The results showed a strong tendency for sexual stereotyping. In the all-male, all-female, and majority-male groups, the person at the head received the highest ratings for leadership. However, in the majority-female groups, the men next to the women at the head received just as high ratings for leadership as did she. The most glaring results were found in the answers to the question of "who contributed the most to the group." In this case, the men in the female-majority groups were rated far higher than the woman at the "head." It is important to note that there were no differences in ratings between male and female participants. In other words, women were just as likely to rate the men higher than the woman at the "head" than were men.

These findings suggest that our implicit leadership theories include "male" as one of the attributes of the "ideal leader." This effect can overcome the tendency for us to see a person at the head of a table as the leader of their group.

The second reason that women are less likely to emerge as group leaders than men is behavioral and follows very clearly from issues we have discussed. The issue is how talkative group members are. Earlier in this chapter, we reported a strong association between how talkative group members are and whether they emerge as group leaders. The more a person talks, the more likely the person is to become a leader. In Chapter 5 we mentioned that women may talk less than men in groups. If this is true, it implies that women are less likely to emerge as group leaders than are men.

Types of Leadership
Our second overall hypothesis is in contrast to the first. We have proposed that, for both perceptual and behavioral reasons, people may believe that women act less like leaders than men do. Consequently, men most often emerge as group leaders. There are other possibilities, however.

People may believe that women and men act as different types of leaders. In this case, groups would not necessarily prefer to give men leadership roles. Instead, groups would decide which type of leader they need most and then perhaps search for a leader from one gender or the other, depending on their needs.

What is the reasoning behind this second hypothesis?

On a perceptual level, people tend to have stereotypes about men and women. Researchers have performed many studies of "gender-role stereotypes." In one case, Broverman et al. (1972) found that people generally believe that men possess traits such as being "objective" and "logical" and that men exhibit behaviors such as "separates feelings from ideas." In contrast, people feel that women's traits include being "sensitive" and "warm" and that female behaviors include such acts as "expresses tender feelings."

Looking at these beliefs, we can hypothesize that people would expect women to be maintenance leaders and men to be task leaders. We can carry this idea even further and say that group members look for different leadership styles from men and women. For instance, if both men and women were task leaders, people might expect men to lead in an autocratic style and women to lead in a democratic style.

Men and women could also exhibit behavioral differences in leadership roles, and group members could notice these differences. In Chapter 5 we mentioned that, proportionally, in groups women tend to be more maintenance oriented and men tend to be more task oriented. This does not necessarily mean that these differences continue to exist when men and women take on leadership roles; however, the possibility does exist.

Hence, as a result of both perceptions and behaviors, group members may believe that women and men act differently as leaders.

Research on Gender Difference in Group Leadership

Does research support either or both of these general hypotheses? Scientists have performed many studies concerning gender differences in group leadership. Eagly and her associates have reviewed these studies for overall trends.

Research on Gender and Leadership Styles
In the first review, Eagly and Johnson (1990) examined the question of leadership style. They looked at studies that investigated differences between the styles that women and men use in leadership positions.

Their findings were consistent with our second overall hypothesis. They discovered a tendency for female leaders to exhibit maintenance behaviors and for male leaders to exhibit task behaviors. In addition, Eagly and Johnson found that female task leaders are more likely to use a democratic style and that male task leaders are more likely to employ an autocratic style.

Research on Gender and Leader Emergence
In a second review, Eagly and Karau (1991) analyzed studies of gender and leader emergence. Their findings supported aspects of both our overall hypotheses.

In general, they found that men are more likely than women to emerge as overall group leaders and as task leaders. In contrast, women are more likely than men to become maintenance leaders.

Eagly and Karau discovered, however, that leader emergence depended on specifics beyond the question of maintenance and task needs. Although men are more likely to emerge as overall leaders, groups will consider several aspects before making their final decisions.

One aspect involves the "sex-type" of the task a group performs. Men more often than women become overall group leaders when their group performs male sex-typed tasks or non-sex-typed tasks. When the group performs female sex-typed tasks, however, women and men are equally likely to emerge as leaders. These findings are consistent with results for power that we discussed in Chapter 5.

Another aspect concerns whether tasks have objectively correct answers. Men are particularly likely to become overall leaders when their groups are performing problem-solving or judgment tasks.

Women, in contrast, are more likely to take on leadership roles when their groups face decision-making or negotiation tasks. We can find the reason for this discovery by looking at considerations we discussed in Chapter 9.

As we stated, problem-solving and judgment tasks have objectively correct answers. With these kinds of tasks, maintenance activities do not seem to help groups find answers. We have outlined the finding that women tend to be "maintenance" specialists. Therefore, groups could judge that women's contributions to these tasks are not particularly important, and they will not see women as potential leaders. In contrast, decision-making and negotiation tasks do not have objectively correct answers. For these tasks, groups must reach consensus. Maintenance behaviors are very helpful as a group works to achieve consensus. Thus, such groups will value women's maintenance contributions highly and will be somewhat more likely to judge women as leaders.

Finally, Eagly and Karau found that the longer the group meets, the less effect gender has on leader emergence. This finding follows from what we have learned about leadership from the perceptual approach. When a group is new, members are strongly influenced by their implicit leadership theories, which imply that a group's leader ought to be male. However, as the group develops, each member's actual behaviors come to have a greater effect on the member's impressions of one another's leadership. Therefore, who actually performs the leadership behaviors is more and more judged to be the group's leader.

Gender and Group Members' Evaluations of Leadership Performance
In a third review, Eagly, Makhijani, and Klonsky (1992) looked at studies of the ways that gender affects how group members evaluate leader performance. As earlier, their findings support aspects of both our overall hypotheses.

The review included two types of studies. In one type, male and female confederates used predetermined styles to lead groups. In the other, participants read scenarios that described women and men as leading groups in particular ways. In both cases the male and female leaders behaved in the same way. Thus, any differences in the ways in which participants reacted to the leaders would be due to gender.

Overall, the participants evaluated the women as slightly less competent leaders than the men. This judgment, however, depended on the leader's style. Evaluations of men and women were no different when both used democratic leadership styles. Participants assessed them as equally competent. When the leaders used autocratic styles, however, the participants judged the women as considerably less competent leaders than the men.

Why? As you can recall, earlier we discussed how leaders are "supposed" to act. People probably believe that men are "supposed" to lead autocratically and that women are "supposed" to lead democratically. This finding suggests that the participants were more rigid in their judgments of women than of men. It appears that men can "get away" with leading the way that women are "supposed" to lead but that women cannot "get away" with using the leadership style that men are "supposed" to use. In other words, male leaders can be more flexible than women without alienating the group members. In contrast, women have to "act like" women.

Gender and Leadership Performance
In the last review, Eagly, Karau, & Makhijani (1995) compared the actual performance of men and women as leaders. No overall difference for gender was found. However, there were many more subtle differences that imply sexual stereotyping. For example, men tend to be more successful as leaders when their subordinates were men, whereas women are more successful leading women. Women tend to be more successful in stereotypically "female" situations, such as leading in social service or educational organizations, while men excel in "male" circumstances such as the military. Finally, women are more successful when their job is seen as requiring primarily maintenance abilities, and men when task abilities are believed more important.

SUMMARY

The functional approach to leadership does not consider the leader as a particular "person." Instead, it is concerned with the types of leadership "behaviors" that must take place within a group for the group to reach its goals. Any group member can perform these behaviors, or "functions." However, only one or two people tend to do them. Traditionally, these functions have been distinguished into task and maintenance categories. Bales believed that task functions and maintenance behaviors are so diverse that different group members generally perform them separately. Other researchers disagree with Bales's idea. In addition, there is good reason to add a third type of function, relating with groups and individuals outside of the group.

Researchers worked to discover how leaders emerge from previously leaderless groups. This is the emergent approach to leadership. They discovered that a group's emergent leader is usually its most verbally and nonverbally active member. It has also become clear that different methods of choosing leaders can be equally successful. A group can acquire a leader through election or appointment or by allowing the leader to emerge. All of these methods can work well if the group accepts the leader, and the leader is competent. Acceptance of a group leader is partly based on the extent to which the leader has conformed to group norms in the past. Past conformity allows the leader to deviate, in an innovative way, on occasion without losing the group's trust.

The perceptual perspective is a unique approach toward leadership. It claims that leadership exists mainly in the minds of each individual group member. Studies have shown that people tend to rate both imaginary and real leaders in a similar fashion. Each group member has his or her own personal view of what traits a leader should have. Every member, in turn, attributes these traits to a person he or she considers a leader.

The charismatic approach has developed due to the observation that some leaders have the ability to inspire followers to unusually high levels of effort. It appears that these leaders have a set of traits that lead them to perform in a distinctive style. When in the right situation, this style can result in very dedicated subordinates and huge success. However, the style will backfire in other situations, and the effects of charismatic leadership often disappear over time.

Researchers have performed many studies concerning gender and group leadership. The findings suggest that women are more likely than men to emerge as maintenance leaders and that men are more likely than women to become task leaders and overall leaders. As task leaders, men tend to be autocratic and women tend to be democratic. The leadership styles that men use do not affect how people evaluate their competence as leaders; however, group members often evaluate women's competence negatively when women use an autocratic style. Finally, while there is no overall gender difference in actual leadership ability, men and women tend to be more successful leaders in different situations.