draft

Learning from Diversity: A Theoretical Exploration 529

Erica Gabrielle Foldy New York University

Learning from Diversity: A Theoretical Exploration

Public-sector organizations tend to be more racially and ethnically diverse than private-sector organizations, leading to the challenge of enhancing heterogeneous work group effectiveness. Recent work suggests that a group’s “diversity perspective,” or set of beliefs about the role of cultural diversity, moderates diverse group performance. One perspective, the integration and learning perspective, argues that heterogeneous groups function better when they believe that cultural identities can be tapped as sources of new ideas and experiences about work. However, simply holding the integration and learning perspective may not be sufficient. Research on general group learning has shown that it requires particular behaviors and cognitive frames. This article integrates recent work on diversity perspectives with long-standing research on team learning to propose a conceptual model of learning in culturally diverse groups. It suggests that both the integration and learning perspective and more generic learning frames and skills must be present.

Thinking and talking about diversity is ubiquitous in today’s organizations. While “diversity” may refer to many kinds of heterogeneity, one of the most challenging dimen- sions for many workplaces is cultural diversity, including racial and ethnic background (Williams and O’Reilly 1998). While organizations face a variety of difficulties related to managing cultural diversity, these challenges vary, to some extent, by sector. Public-sector organizations tend to have more diverse employee populations than organizations in the private sector, suggesting they have been more suc- cessful in recruiting employees of color, though often not in promoting them to higher levels (Riccucci 2002; Corn- well and Kellough 1994). However, their very success in diversifying their workforces poses a new challenge: How can public organizations ensure that heterogeneous em- ployees work well together?

This is a pressing question, especially given the ration- ales for broadening diversity in the public sector. Ospina (2001), for example, suggests two main reasons. First, cul- tural diversity enhances organizational performance by con- tributing to functional diversity, which, in turn, increases the prevalence of alternative perspectives and new ideas. Second, it enhances organizational legitimacy by creating an employee population that roughly mirrors the popula- tion it serves.

Erica Gabrielle Foldy is an assistant professor of public and nonprofit man- agement at the Wagner School of Public Service at New York University. Her research interests include identity and diversity in organizations, organiza- tional learning and reflective practice, and the interaction of individual, or- ganizational, and social change. She has published articles in several jour- nals and edited volumes and co-edited the Reader in Gender, Work, and Organization (Blackwell, 2003). E-mail: [email protected].

Articles from the Seventh National Public Management Research Conference

In 1991, an interdisciplinary group of scholars interested in management and organizations convened the first biannual National Public Management Research Conference. The conference was held at the Maxwell School of Citizenship and Public Affairs, Syracuse University. (Subsequent conferences were held at the University of Wisconsin, the University of Kansas, the University of Georgia, Texas A&M University and Indiana University–Bloomington). The Georgetown Public Policy Institute at Georgetown University hosted the seventh conference October 9–11, 2003. Public Administration Review agreed to publish articles from the conference. We would like to thank Professor William Gormley and his colleagues for their warm and generous hospitality.

Articles from the conference are published in two installments, the first of which is included in this issue. The second installment will appear in the November/December issue. We have selected articles that reflect the range and diversity of papers presented at the conference. Consistent with PAR publication policy, all articles were evaluated by external reviewers. —LDT

530 Public Administration Review • September/October 2004, Vol. 64, No. 5

However, both rationales may be undermined if diverse groups don’t work together effectively. First, research has established that simply creating diversity through hiring does not automatically lead to enhanced organizational per- formance. Research shows that diverse groups can be both more creative and effective (Watson, Kumar, and Michael- son 1993; Cox, Lobel, and McLeod 1991), as well as more conflictual (Jehn, Northcraft, and Neale 1999; Tsui, Egan, and Xin 1995). If heterogeneous work groups experience increased conflict and miscommunication, this may actu- ally lead to diminished performance. Second, while an agency’s workforce may represent its service population, if it doesn’t work together effectively, it will provide less- than-optimal services, thereby undermining legitimacy rather than enhancing it. Therefore, public-sector organi- zations must go beyond creating diversity to consider how to reap its positive benefits.

Some recent research focused on group diversity could provide a useful approach. Ely and Thomas propose the concept of “diversity perspectives” (Ely and Thomas 2001; Thomas and Ely 1996). They argue that diverse groups who hold the perspective that cultural identity is a resource for learning and growth are more likely to learn from difference, resulting in higher performance. But the authors do not elaborate the components of this perspec- tive or how such a perspective leads to learning in cultur- ally diverse groups.

Fortunately, the substantial literature on team or group learning offers a number of insights that could elaborate the concept of diversity perspectives and their contribution to learning (Edmondson 1999; Edmondson, Bohmer, and Pisano 2001; Argote, Gruenfeld, and Naquin 2001). This literature has made remarkable progress in delineating what enables group learning, including particular behaviors as well as frames or beliefs held by team members. Yet, this research has largely ignored the particular needs and dy- namics of culturally diverse work groups (exceptions in- clude Child and Rodrigues 2003; Taylor and Osland 2003; Gibson and Vermeulen 2003), despite the broad recogni- tion that such groups face both opportunities and challenges that homogeneous groups do not (Williams and O’Reilly 1998; Shaw 1981; Shaw and Barrett-Power 1998).

This article integrates insights from Ely and Thomas’s work on diversity perspectives with work on group learn- ing to explore the question: what enables learning in cul- turally diverse groups? This article proposes a model to answer this question. It begins by summarizing what we already know about group learning from the learning lit- erature. It then investigates culturally diverse groups, how they are different from homogeneous groups, and why perspectives about diversity affect their learning capacity. It then proposes and elaborates a model of learning in cul- turally diverse groups.

Group Learning While scholars have identified factors at the organiza-

tional, group, and individual levels that contribute to learn- ing (Dodgson 1993; Huber 1991; Levitt and March 1988), some argue that individuals are at the heart of group learn- ing. As Argyris and Schon point out, “We take individual practitioners as centrally important to organizational learn- ing, because it is their thinking and acting that influence the acquisition of capability for productive learning at the organizational level” (1996, xxii). Therefore, this article focuses on the beliefs and behaviors of team members which enhance team and organizational learning (Argyris, Putnam, and Smith 1985; Argyris and Schon 1996, 1974).

Argyris and Schon distinguish between Model I behav- iors and Model II behaviors (Argyris and Schon 1974, 1996). Model I behaviors, which are counterproductive to learning, include making untested attributions about oth- ers’ motivations and positions, not inquiring into others’ views, and offering no opportunities to test one’s advoca- cies. Model II behaviors, which enhance learning, include suggesting ways to test one’s advocacies and attributions and inquiring into why others view things the way they do. Implicit in this work is the individual’s capacity to surface and reflect on his or her embedded assumptions or mental models (Rudolph, Taylor, and Foldy 2001; Ayas and Zeniuk 2001). Table 1 compares Model I (low-learning) and Model II (high-learning) behaviors.

Work from this approach also suggests that particular beliefs or frames, in addition to skills or behaviors, are essential to learning. Model II (high-learning) frames in- clude “errors are puzzles to be engaged” and “role of learner as agent,” rather than Model I (low-learning) frames such as “errors are crimes to be covered up” and “role of learner as recipient” (Argyris, Putnam, and Smith 1985, 280) (table 2).

Research on these generic learning behaviors and frames demonstrates they can be quite effective, but little work has explored whether these approaches are suffi- cient for learning to occur in culturally diverse groups. Research on diverse groups suggests they face different challenges and opportunities than homogeneous ones, raising the possibility that generic learning approaches may not be sufficient.

Culturally Diverse Groups Cultural diversity refers to identities such as race,

ethnicity, nationality, religion, gender, and other dimen- sions of difference derived from membership in groups that are socioculturally distinct, that is, they “collectively share certain norms, values or traditions that are different from those of other groups” (Cox 1994, 5–6). Members of the same cultural identity group often—though not always—

Learning from Diversity: A Theoretical Exploration 531

have similarities of background and experience that shape the way they see the world.

Cultural identity groups also tend to be associated with power differentials, in that some groups have higher status and greater access to resources than other groups (Nkomo 1992; Omi and Winant 1994; Ragins 1997; Ridgeway and Berger 1986). While this is a broad generalization, in West- ern countries men tend to have more power than women, whites generally have more resources than people of color, and so on. When power differentials are roughly contigu- ous with identity groups, this may reinforce the bound- aries among the groups, making their group identities more salient (Lau and Murnighan 1998). Group identity is no longer based solely on cultural similarity, but on a shared status or shared interests.

Numerous studies have established that culturally di- verse work groups have different dynamics than homoge- neous groups (Williams and O’Reilly 1998). These groups

are broadly seen as having both increased challenges and opportunities: Diverse groups often experience miscom- munication and disabling conflict (Tsui, Egan, and Xin 1995; Shaw 1981), yet under the right circumstances, they can be synergistic and creative (McGrath 1984; Cox, Lobel, and McLeod 1991).

The factors underlying these different dynamics are nu- merous and complex, but four reasons stand out. First, a long tradition of research has demonstrated that individu- als are generally more comfortable when they are sur- rounded by people they perceive as more like them (Schneider 1987; Tsui, Egan, and O’Reilly 1992; Kanter [1977]1993; Brief 1998). We look for familiarity and simi- larity; we are reassured when others think, talk, and act like we do. These simple personal preferences have profound consequences for organizations and for work groups. “So- cial similarity, whatever criteria it uses, acts as a mecha- nism of exclusion or inclusion …” (Ospina 1996a, 141). Groups include those who feel familiar or safe and exclude those who don’t. Kanter, among others, has elaborated the implications of “homosocial reproduction” for organiza- tional stratification (Kanter [1977]1993). Demographic groups with power—often white men—tend to stay in power because they choose others like themselves as colleagues. Therefore, individuals from demographic groups with less power—often people of color and white women—continue to be shut out from the highest levels of the organization.

However, the powerful impulse toward similarity also has implications for group dynamics. It means that indi- viduals are inclined to seek groups made up of people like themselves. When individuals find themselves in diverse groups, despite their preferences, those groups are likely to feel less safe and less trusting. Less trust means that group members are less likely to give others the benefit of the doubt, leading to more conflict (Child and Rodrigues 2003). Such groups feel less familiar, meaning that group members are more likely to feel strange or to perceive oth- ers as strange. That dynamic contributes to less trust, and often more conflict.

Second, as suggested by the definition of cultural diver- sity, group members come with different life experiences that have shaped their values, approaches, and perspec- tives. Members of culturally diverse groups may be more likely than those of homogeneous groups to differ in how they define a problem, structure a discussion, view poten- tial solutions, or come to a decision. These differences of opinion represent a mother lode of creativity or a quag- mire of conflict, depending on how the group handles con- flict and differences (Chatman and Flynn 2001; Chatman et al. 1998; Jehn, Northcraft, and Neale 1999).

Third, group membership is associated with differing representation within the group. Members of groups in the minority, whatever that means in a particular context, are

Table 1 Comparison of Low-Learning (Model I) and High-Learning (Model II) Behaviors

Low learning 1. Make attributions about others

without testing them. 2. Advocate positions without

illustrating them or suggesting a way to test them.

3. Invoke abstract concepts that are impossible to disagree with.

4. Construct a situation as a dilemma or double bind; feel trapped.

5. Do not inquire into why you think, feel, or act the way you do.

6. Do not inquire into why others think, feel, or act the way they do.

High learning 1. Make private attributions

about others public; test them. 2. Illustrate and suggest ways to

test one’s advocacies.

3. When invoking abstract concepts, try to make them concrete and testable.

4. Make dilemmas public and discussable.

5. Publicly reflect on why you are thinking, feeling, or acting as you are.

6. Inquire into why others are thinking, feeling, or acting the way they are.

Table 2 Comparison of Low-Learning (Model I) and High-Learning (Model II) Frames

Low learning 1. Mistakes are crimes to be

prosecuted. 2. My role as a participant is to

be a recipient. 3. I don’t have anything to learn

from others in this group. 4. If I don’t have a solution, I

shouldn’t raise the problem. 5. If I feel uncomfortable in this

discussion, something must be wrong.

6. If I speak up, I will be criticized.

7. I don’t have any power or authority in this group, so I will be quiet.

High learning 1. Mistakes are puzzles to be

engaged. 2. I should be an engaged

participant. 3. I don’t know everything.

4. It is helpful to raise problems, even if I don’t have a solution.

5. Feeling uncomfortable in a discussion can be a sign that this is exactly where I should be.

6. I think I have something to contribute to the conversation, even though what I say may be criticized.

7. I can make a contribution here even though I don’t have any formal power or authority.

532 Public Administration Review • September/October 2004, Vol. 64, No. 5

more aware of their identity and of being different from the norm (Ashforth and Mael 1989; Kanter [1977]1993). Depending on the dynamics of the group, they may feel less comfortable or less welcome. Members of majority groups, who share ways of thinking and acting, may un- wittingly create environments that make it difficult for oth- ers to feel included.

Finally, diverse groups may also have different dynam- ics because of the power differences associated with cul- tural differences. The literature on group decision making and problem solving has demonstrated that more powerful members tend to talk more and have more influence on outcomes, contributing to less powerful members with- drawing and falling silent (Falk and Falk 1981). In cultur- ally diverse teams, members of dominant identity groups, such as whites, are often the more powerful members of such groups (Levin et al. 2002; Sidanius and Pratto 1999). Therefore, they may consciously or unconsciously act in ways that reinforce their dominance in their conversational styles, decision-making processes, social interaction, and so on (Ridgeway 1997; Smith-Lovin and Brody 1989; Elsass and Graves 1997). Like any less powerful group members, members of nondominant groups may also con- tribute to dysfunctional dynamics by withdrawing or by communicating largely with other members of their sub- group (Konrad 2003).

All of these dynamics suggest that generic learning frames and behaviors may be harder to establish in cultur- ally diverse groups. On an individual level, learning means making oneself vulnerable; it means admitting that one is dependent on others to grow and develop (Brown and Starkey 2000). As others have amply demonstrated, this is difficult enough in homogeneous groups and even more difficult in heterogeneous groups. For example, making mistakes may carry greater weight in heterogeneous groups, which means that admitting them will be all the more dif- ficult. This is particularly true for members of nondominant groups, who often feel their capabilities are in question, and therefore they may feel under greater pressure to per- form (Steele 1997). Carrying the frame that mistakes are opportunities to learn rather than crimes to be punished is more difficult under these circumstances. Maintaining a stance of inquiry may also be more difficult for all indi- viduals in diverse groups. The concept of “threat-rigidity” (Staw, Sandelands, and Dutton 1981) suggests that we are less open and flexible when we feel threatened or unpro- tected in some way. Working with others who are different from us can feel unsafe or unfamiliar, as opposed to the sense of comfort and fit that comes from homogeneity. For that reason, group members may be more likely to rigidly advocate their own positions rather than inquire about those of others.

Individual feelings of discomfort or threat create diffi-

cult group-level dynamics. Heterogeneous groups are less likely to achieve a broad sense of psychological safety (Edmondson 1999). In some cases, the difficult dynamics associated with cultural diversity may cause most mem- bers of a team to feel generally unsafe, whether they are members of majority or minority, dominant or nondominant groups. If dynamics are characterized by a lot of conflict, mistrust, and paralysis, then it is less likely that anyone will feel safe. In other cases, majority or dominant mem- bers may feel safe, while those in the minority will not because they feel uncomfortable or unwelcome (Earley and Mosakowski 2000; Elsass and Graves 1997).

If it is true that a learning stance is more difficult in diverse groups, then what could make the difference? What enables learning amid cultural diversity?

Diversity Perspectives Recent research suggests that a group’s diversity per-

spective is central to learning in culturally diverse groups. Ely and Thomas (2001; Thomas and Ely 1996) propose the idea of “diversity perspectives” as the key moderator of the relationship between diversity and performance. Diversity perspectives are a work group-level phenomenon; different work groups within the same organization can hold different perspectives. A diversity perspective is the way that group members think about the cultural differ- ences among them, whether they are important, and how they might be harnessed to further the group’s work.

A diversity perspective includes “the rationale that guides people’s efforts to create and respond to cultural diversity in a work group; normative beliefs about the value of cultural identity at work; expectations about the kind of impact, if any, cultural differences can and should have on the group and its work; and beliefs about what constitutes progress toward the ideal multicultural work group” (Ely and Thomas 2001, 234.) Such a perspective might be writ- ten in mission statements or diversity policies, but more often such explicit statements refer to the work group’s espoused theory rather than the theory in use actually held by group members (Argyris and Schon 1996, 13). A group’s diversity perspective is the approach, usually implicit, un- derlying the way it defines its tasks and goals and how group members interact.

Ely and Thomas (2001) identify three diversity perspec- tives. The discrimination and fairness perspective is con- cerned with the recruitment and retention of employees from protected groups. While its commitment to redress- ing past inequities is both essential and laudable, the per- spective presumes that cultural dimensions of diversity, such as race, nationality, or gender, don’t have important consequences for work practices, that nothing is to be gained by surfacing and engaging differences. The access

Learning from Diversity: A Theoretical Exploration 533

and legitimacy perspective challenges these notions to some extent. Organizations with this perspective celebrate cul- tural differences, but in simplistic and narrow ways. They are likely to bring in employees from nontraditional back- grounds to reach new clients and constituencies, such as hiring Hispanic employees to work in Spanish-speaking communities. There is little further investigation of the potential of diverse backgrounds and ideas.

Finally, the integration and learning perspective seeks to build, more deeply and comprehensively, on the varied skills, experiences, and ways of thinking of a diverse workforce. It suggests that organizations should “incorpo- rate employees’ perspectives into the main work of the or- ganization and to enhance work by rethinking primary tasks and redefining markets, products, strategies, missions, busi- ness practices and even cultures” (Thomas and Ely 1996, 85). Differences can be a source of growth, learning, and insight, but only if they are acknowledged and construc- tively explored.

To understand learning in culturally diverse groups re- quires integrating previous research on group learning with the literature on diversity perspectives.

Learning in Culturally Diverse Groups Figure 1 suggests a model for enabling learning in cul-

turally diverse groups. In order to learn, groups must have the integration and learning perspective, whether it is held implicitly or explicitly. They also must have generic Model II learning frames and behaviors, as illustrated in the work of Argyris and colleagues, Edmondson, and others (Ed- mondson 1999; Argyris, Putnam, and Smith 1985; Argyris and Schon 1996; Fisher and Torbert 1995). The model sug- gests that a group’s diversity perspective and learning frames and behaviors influence each other, resulting in a set of learning frames and actions that is specifically re- lated to cultural diversity. These frames and behaviors then enable learning in culturally diverse groups.

I will now explore each of these relationships in turn: How do diversity perspectives and learning frames and

behaviors influence each other to create specific learning frames and behaviors, and how do those specific frames lead to learning in multicultural groups?

Diversity Perspectives and Learning Frames and Behaviors

The model suggests that a group’s diversity perspective and its learning frames and actions influence each other, creating an overall environment that affects learning. If a group has both the integration and learning perspective and reflective learning frames and behaviors, then each will “activate” the other—that is, each adds an ingredient that makes the other come alive in a given group setting. The integration and learning perspective enlarges and contextualizes a group’s learning beliefs and behaviors; it allows them to address the particular challenges of learn- ing in culturally diverse groups. Just as important, the group’s learning beliefs and actions enable the group to enact its perspective on diversity: They provide the tools by which a group can mutually investigate their differences. I will elaborate each of these claims in turn.

The Integration and Learning Perspective Activates Generic Learning Frames and Behaviors. Why does this perspective activate learning in culturally diverse groups? Most importantly, and most simply, the integration and learning perspective is the only perspective that suggests cultural diversity is a source for learning, as indicated by its name. This approach contradicts the way that diversity is conceptualized in the other two perspectives.

Both the discrimination and fairness and the access and legitimacy perspectives share a basic color-blind stance, though they manifest it a little differently (Ely and Tho- mas 2001). The discrimination and fairness perspective argues that, although people may look different physiologi- cally, in fact people are just people and we are all the same in what really matters: how we think and what we do. While emphasizing our common humanity is important, especially when dealing with explicit racism and ethno- centrism, refusing to acknowledge cultural differences glosses over the very different histories of cultural groups in this country. It diminishes the legacy of slavery for Af- rican Americans and the impact of immigration and as- similation on white ethnic groups such as Eastern Euro- peans and Italians. It ignores the ongoing differences in experiences among cultural groups: for example, the fact that people of color are much more likely to experience discrimination, or that being bilingual allows deeper par- ticipation in multiple cultures.

Failure to acknowledge cultural differences also makes it impossible to consider how cultural background influ- ences our ideas and our contributions. It makes it difficult, for example, for an African American individual to con- sider that the way she approaches working with customers

Figure 1 A Model of Learning in Culturally Diverse Groups

Learning in culturally diverse groups

Integration and learning perspective

Generic high- learning frames and behaviors

Specific high- learning frames and behaviors related to cultural diversity

534 Public Administration Review • September/October 2004, Vol. 64, No. 5

might in some way trace back to experiences she had feel- ing unwelcome in some stores. It means that explicitly ask- ing about group members’ religious backgrounds as part of revamping human resource practices would be consid- ered unseemly or irrelevant, even though religious prac- tices may be affected by such policies as the holidays the organization chooses to close.

While the access and legitimacy perspective does sug- gest that cultural background matters, it limits it to very narrow spheres. It suggests that cultural background mat- ters only when members of a group are dealing with other members of their group: that the heritage of employees of Latino descent, for example, comes into play only when they have Hispanic customers. Members of nondominant groups have a special contribution only when dealing with other members of the same marginalized group. Therefore, it only makes sense to be aware of color within the group; color blindness is still the correct stance across groups.

The alternative—the integration and learning perspec- tive—conjectures that our cultural heritage brings a valu- able set of experiences that are broadly applicable to the way the organization does its work. The integration and learning perspective is quite explicitly not color blind: It acknowledges that our group identity says a great deal about our and our ancestors’ life experiences and informs who we are. Regarding the impact of cultural identity on work, this perspective argues that race, ethnicity, nationality, and other differences are extremely valuable resources because they often, though certainly not always, imply a different set of perspectives and cultural teachings. The integration and learning perspective differs from much of the cultural sensitivity literature (Adler 1991) in that it does not pre- sume to know, given a person’s cultural identity, what those experiences or teachings might be. It simply assumes that different life experiences are likely to bring different ways of thinking about work, and leveraging those different ways of thinking is likely to enhance effectiveness.

Learning Frames and Behaviors Activate the Integra- tion and Learning Perspective. While the integration and learning perspective is necessary to activate learning frames and behaviors, the reverse is also true. Without learning beliefs and actions, the integration and learning perspec- tive is simply an espoused perspective that will not mani- fest itself in the group’s work together. Learning beliefs and behaviors allow groups to undertake the difficult work of expressing and working with culturally based beliefs. Groups with the espoused integration and learning perspec- tive, but without a learning stance, might believe that cul- tural groups must be acknowledged and celebrated; that all groups have wisdom that can be tapped; and that creat- ing a multicultural rather than an assimilationist environ- ment is essential. However, they lack the learning tools necessary to create such an environment: capacities such

Table 3 Low-Learning and High-Learning Frames Related to Culturally Diverse Groups

Low learning 1. If I say something about race, I

may say something wrong or culturally insensitive.

2. As a white person, I don’t have anything to add to discussions about race.

3. As a person of color, I don’t have anything to learn from white people about race.

4. Racial issues are insoluble and I will make things worse if I say anything.

5. Talking about race makes me uncomfortable, so I don’t think we should do it.

6. People will think I’m too militant if I speak up.

7. There are no blacks in positions of power in this organization, so it’s clear nobody cares what we think and we have no authority—so I won’t say anything.

High learning 1. I should say something, even if

it might come off as prejudiced or racist, because it is an opportunity to learn.

2. Even though I’m white, I have something to contribute to discussions about race.

3. As a person of color, I could have something to learn from a white person about race.

4. Even though this is a racial issue—with a lot of history and complexity and no obvious solution—it still might be helpful to raise it.

5. Talking about race makes me uncomfortable—which could mean that is exactly what I should do.

6. I think this could be a helpful contribution, even though I may be considered too militant for saying it.

7. Even though there are no blacks in positions of power in this organization and I’m not convinced that anyone cares what we think, I will take the risk of speaking my mind.

as being able to surface and reflect on one’s embedded assumptions or to hold genuine curiosity on someone else’s point of view. In fact, an espoused integration and learning perspective could become a rhetorical device, a rigidly held belief used to promote a particular point of view and way of behaving, rather than a supple and flexible tool that al- lows new ways of working to come into being. In fact, any perspective—on diversity or any other topic—held in the absence of a learning stance can become an iron cage rather than an open door.

In summary, both the integration and learning perspec- tive and generic learning frames and behaviors must be present in order for learning to occur in culturally diverse groups. Each element activates the other; in other words, there is an interaction effect between the two elements.

Specific Learning Frames and Behaviors Related to Cultural Diversity. The interaction between the integra- tion and learning perspective and generic learning frames and behaviors enables the development of specific high- learning frames and behaviors related to cultural identity, which are crucial to developing a capacity for learning in culturally diverse groups. I suggest a number of these spe- cific high-learning frames and behaviors in tables 3 and 4, contrasting them with their low-learning equivalents. (I am not arguing that these specific frames and behaviors must be present; I am suggesting them only as examples.) I use race as an example of a cultural identity that may be sa-

Learning from Diversity: A Theoretical Exploration 535

lient in a given work group, but other dimensions of iden- tity could be substituted, though the specific content of the frames might differ.

How Specific High-Learning Frames and Behaviors Facilitate Learning in Culturally Diverse Groups

Ample research has demonstrated that particular frames and behaviors are more likely to enhance learning (Edmondson, Bohmer, and Pisano 2001; Argyris and Schon 1996; Rudolph, Taylor, and Foldy 2001). Frames such as “mistakes are opportunities for learning” and behaviors such as inquiry into others’ perspectives and feelings cre- ate an atmosphere of curiosity, engagement, and safety. Such an atmosphere allows individuals to loosen their at- tachment to particular approaches and try on new ways of thinking and acting.

I suggest that high-learning frames and behaviors related to cultural diversity play the same role in culturally diverse groups. These frames and behaviors make three important contributions beyond generic high-learning frames and be- haviors. First, these frames and behaviors manifest the idea, suggested by the integration and learning perspective, that cultural differences are an important resource for thinking about the way work is done. Encouraging people to draw on their cultural experiences will enrich, deepen, and broaden a team’s approach to its work. Second, these frames and behaviors suggest that cultural differences are not only significant, they must be discussed. Given the charged na- ture of race, religion, and other identities in American soci- ety (Nkomo and Cox 1996; Omi and Winant 1994), this is a big change in most group norms. But it is impossible to draw on cultural differences if they are kept veiled. Finally, these frames and behaviors hold everyone in a group re- sponsible for addressing difference. Everyone has some- thing to bring to the table, given their background. Every- one is expected to be open to what others bring and to consider how it can be incorporated into work practices. A generic learning stance argues that differences must be dis- cussed and that individuals are responsible for their own learning. These specific frames and behaviors explicitly make charged cultural differences part of the equation.

These frames and behaviors enable learning in two ways. First, they enable learning from difference—that is, they encourage explicit discussion of cultural differences and how those differences affect the way we think about the group’s work. This broadens the pool of options for struc- turing work, designing strategy, and carrying out other key organizational tasks. Second, they also enable learning across difference, meaning they facilitate learning about a host of things that may be unrelated to cultural differences. This is because general learning in culturally diverse groups may be hampered by the more difficult dynamics often faced by such teams, as outlined earlier in this article. Given the heterogeneity of perspectives, the sense of discomfort and lack of fit felt by employees when working with people unlike themselves, and the issues of subgroup dominance and marginalization, diverse groups tend to have more dysfunctional conflict than homogeneous groups. These specific learning frames and behaviors help to establish a sense of safety in a diverse group, allowing everyone to feel more open to new ways of thinking and acting.

Conclusion Many public-sector organizations have been relatively

successful at diversifying their workforce, but they are still struggling to reap the benefits of that heterogeneity (Ospina 1996b). This article suggests one approach to enhancing the effectiveness of culturally diverse work groups by focusing

Table 4 Low-Learning and High-Learning Behaviors Related to Culturally Diverse Groups

Low learning 1. Make attributions about others

without testing them: “X is just saying that because she knows nothing about the Latino community.”

2. Advocate positions without illustrating them or suggesting a way to test them: “This issue has everything to do with race.”

3. Invoke abstract concepts that are impossible to disagree with: “People are people. Everyone is the same under the skin.”

4. Construct a situation as a dilemma or double bind; feel trapped: “If I don’t raise this, I don’t think we can make any progress. But if I do raise this, people will think I’m too militant. I’m stuck.”

5. Do not inquire into why you think, feel, or act the way you do: “Even though X thinks I can’t deal with racial issues, this is just the way I feel and that’s that.”

6. Do not inquire into why others think, feel or act the way they do. “Jeez, Y always gets that scared expression when we talk about race. What’s his problem?”

High learning 1. Make private attributions about

others public; test them: “I realize I’m sitting here thinking you are saying this because you are unfamiliar with the Latino community, but perhaps I am being unfair to you.”

2. Illustrate and suggest ways to test one’s advocacies: “I think this issue is about race because it relates to our two depart- ments, one of which is mostly black and the other is mostly white. Perhaps we should talk to several members of each department …”

3. When invoking abstract concepts, try to make them concrete and testable:. “I think people are people, so I worry about dividing our constituency base into different ethnic markets. So perhaps we should do some research …”

4. Make dilemmas public and discussable: “I realize I feel caught in a bind. I want to raise an issue, but I’m afraid people will see me as too militant. But I think we’re stuck and this might help us get unstuck. So I want to take this risk …”

5. Publicly reflect on why you are thinking, feeling or acting as you are: “I wonder why I feel so strongly about this? Perhaps because I grew up in an all- white neighborhood and I don’t have much experience …”

6. Inquire into why others are thinking, feeling or acting the way they are: “Y, you have a worried look on your face. Are you feeling concerned?

536 Public Administration Review • September/October 2004, Vol. 64, No. 5

on how groups can learn from and across difference. It pro- poses a model that builds on work on group learning and diversity perspectives and suggests that two elements must be present: the integration and learning perspective on di- versity (Ely and Thomas 2001) and generic, high-learning frames and behaviors (Argyris and Schon 1974). These two elements each activate the other—turn them on, so to speak. The interaction between the elements results in specific high- learning frames and behaviors related to cultural diversity. Those frames and behaviors then catalyze learning in cul- turally heterogeneous groups. In addition to providing this framework, integrating the broader literature on team learn- ing with the construct of diversity perspectives enriches our understanding of both concepts.

Contributions to the Literature on Group Learning

Recognizing the particular challenges of culturally di- verse groups forces researchers to consider the social con- text in which learning takes place. Although the impact of organizational structure and culture on learning has been well researched (Popper and Lipshitz 1998; Nevis, DiBella, and Gould 1995; Levitt and March 1988), the role of the larger social environment remains largely invisible. But as embedded intergroup theory (Alderfer and Smith 1982; Alderfer et al. 1980) reminds us, groups in organizations are embedded in multiple contexts, not just departmental and organizational, but also societal. Therefore, group in- teractions and intergroup relations are not immune from larger societal dynamics. But most researchers on group learning conceptualize teams as acting in a social vacuum.

Teasing out the particular dynamics of multicultural groups requires placing those teams in a larger context and considering the impact of demography, history, privilege, and marginalization on their capacity for learning. The concept of diversity perspectives argues that diverse groups have a way of thinking about that broader context, even if their models are usually implicit and unacknowledged. Groups with a discrimination and fairness perspective be- lieve that, although societal inequities must be addressed through equal opportunity and affirmative action, cultural differences are not significant and should be muted. Groups with an access and legitimacy perspective believe that cul- tural background is important only when working with particular, usually nondominant, populations. Only groups with an integration and learning perspective believe that cultural experience is a resource for learning—that our experiences as members of social-identity groups should explicitly be brought into the workplace as material to en- rich our work. Therefore, the concept of diversity perspec- tives embeds learning in its social context and suggests that different approaches to that context will affect a group’s capacity to learn from and across cultural difference.

When we look at the experiences of culturally hetero- geneous groups, we learn more about the importance of power as it affects learning. Of course, all groups have to deal with power dynamics (Ridgeway and Berger 1986; Ridgeway 1997), and there is a growing recognition that power dynamics must be addressed in order for learning to happen. Edmondson (2002), for example, argues that whether groups are able to learn depends on whether power differences are exaggerated or minimized. In a study of 12 teams, she found that a subsection of teams were quite skilled at group reflection, but unable to take action when it might have threatened the authority of the group’s leader. This and other work (Contu and Willmott 2003; Lee 1997) highlight the role of power dynamics associated with or- ganizational level or position power (Finkelstein 1992). Consequences of learning that threaten the status of the superordinate may be suppressed in some way, damaging the integrity and reputation of the learning process.

However, little work has explored power inequities as- sociated with different cultural identity groups and their impact on learning. Multicultural groups will certainly have power dynamics related to organizational level, as do homogeneous teams, but they also have culturally as- sociated power dynamics. Certain subgroups may be rep- resented in greater numbers. Also, regardless of repre- sentation, certain subgroups may be dominant in some way while others are marginalized. Again, the notion of diversity perspectives provides a foundation for thinking about these kinds of power issues in groups. The integra- tion and learning perspective argues that cultural back- ground is connected to work. But to truly mine that re- source, groups must address obstacles to fully integrating the ideas and approaches of all of their members. This means that a group must somehow deal with power ineq- uities within the group.

Addressing the particularities of multicultural teams places learning in a broad social context that considers the consequences of cultural background and the particular power dynamics associated with cultural differences. This begins the process of grounding the concept of learning in a particular social and historical reality, rather than defin- ing it as disconnected from its source.

Contributions to the Literature on Diversity The concept of diversity perspectives is an important

advance in the broader diversity literature because it helps us understand why some diverse groups are more effective than others. Identifying the three perspectives provides a useful tool for work groups to think about their own ratio- nale for creating heterogeneity. However, as the construct’s authors themselves point out, “More theoretical and em- pirical development is needed to understand fully the inte- gration-and-learning perspective’s potential for connect-

Learning from Diversity: A Theoretical Exploration 537

ing organizations’ cultural diversity to their core work and work processes” (Ely and Thomas 2001, 270).

The literature on team learning provides a useful start- ing point for this exploration. What is the mechanism that connects diversity to work processes? According to the learning literature, it is learning itself, the process of syn- thesizing a variety of different perspectives into action- able ideas. This literature has traditionally focused only on functional or professional diversity and the expertise and knowledge they spawn (Edmondson and Moingeon 1998). However, it does suggest specific frames and be- haviors that enable groups to take advantage of that diver- sity, rather than being torn apart by it. The stance of curi- osity and inquiry implicit in those frames and behaviors is essential to thinking about any kind of difference, and there- fore is an important starting point for linking cultural di- versity to organizational processes. Grounding the con- cept of diversity perspectives in the established learning literature elaborates and enriches its connection to what we know about learning.

The increasing diversity of public organizations presents a challenge: Can heterogeneity be harnessed for increased organizational effectiveness? This article suggests that learning from and across difference could be one path to enhanced performance.

References

Adler, Nancy J. 1991. International Dimensions of Organiza- tional Behavior. Belmont, CA: Wadsworth.

Alderfer, Clayton P., and Ken K. Smith. 1982. Studying Inter- group Relations Embedded in Organizations. Administrative Science Quarterly 27(1): 35–65.

Alderfer, Clayton P., Charleen J. Alderfer, Leota Tucker, and Robert Tucker. 1980. Diagnosing Race Relations in Manage- ment. Journal of Applied Behavioral Science 16(2): 135–66.

Argote, Linda, Debroah Gruenfeld, and Charles Naquin. 2001. Group Learning in Organizations. In Groups at Work: Theory and Research, edited by M.E. Turner, 369–411. Mahwah, NJ: Lawrence Erlbaum.

Argyris, Chris, Robert Putnam, and Diana McLain Smith. 1985. Action Science. San Francisco: Jossey-Bass.

Argyris, Chris, and Donald A. Schon. 1974. Theory in Practice: Increasing Professional Effectiveness. San Francisco: Jossey- Bass.

———. 1996. Organizational Learning II: Theory, Method and Practice. Reading, MA: Addison-Wesley.

Ashforth, Blake E., and Fred Mael. 1989. Social Identity Theory and the Organization. Academy of Management Review 14(1): 20–39.

Ayas, Karen, and Nick Zeniuk. 2001. Project-Based Learning: Building Communities of Reflective Practitioners. Manage- ment Learning 32(1): 61–76.

Brief, Arthur P. 1998. Attitudes In and Around Organizations. Thousand Oaks, CA: Sage Publications.

Brown, Andrew D., and Ken Starkey. 2000. Organizational Iden- tity and Learning: A Psychodynamic Perspective. Academy of Management Review 25(1): 102–20.

Chatman, Jennifer A., and Francis J. Flynn. 2001. The Influence of Demographic Heterogeneity on the Emergence and Con- sequences of Cooperative Norms in Work Teams. Academy of Management Journal 44(5): 956–74.

Chatman, Jennifer A., Jeffrey T. Polzer, Sigal G. Barsade, and Margaret A. Neale. 1998. Being Different Yet Feeling Simi- lar: The Influence of Demographic Composition and Organi- zational Culture on Work Processes and Outcomes. Adminis- trative Science Quarterly 43(4): 749–80.

Child, John, and Suzana Rodrigues. 2003. Social Identity and Organizational Learning. In Blackwell Handbook of Organi- zational Learning and Knowledge Management, edited by M. Easterby-Smith and M.A. Lyles, 535–56. Oxford, UK: Blackwell.

Contu, Alessia, and Hugh Willmott. 2003. Re-Embedding Situatedness: The Importance of Power Relations in Learn- ing Theory. Organization Science 14(3): 283–96.

Cox, Taylor, Jr. 1994. Cultural Diversity in Organizations: Theory, Research and Practice. San Francisco: Berrett- Koehler.

Cox, Taylor, Sharon Lobel, and Poppy McLeod. 1991. Effects of Ethnic Group Cultural Differences on Cooperative and Competitive Behavior on a Group Task. Academy of Man- agement Journal 34(4): 827–47.

Dodgson, Mark. 1993. Organization Learning: A Review of Some Literatures. Organization 14(3): 375–94.

Earley, P. Christopher, and Elaine Mosakowski. 2000. Creating Hybrid Team Cultures: An Empirical Test of Transnational Team Functioning. Academy of Management Journal 43(1): 26–49.

Edmondson, Amy. 1999. Psychological Safety and Learning Behavior in Work Teams. Administrative Science Quarterly 44(3): 350–83.

———. 2002. The Local and Variegated Nature of Learning in Organizations: A Group-Level Perspective. Organization Sci- ence 13(2): 128–46.

Edmondson, Amy C., Richard M. Bohmer, and Gary P. Pisano. 2001. Disrupted Routines: Team Learning and New Technol- ogy Implementation in Hospitals. Administrative Science Quarterly 46(4): 685–716.

Edmondson, Amy, and Bertrand Moingeon. 1998. From Orga- nizational Learning to the Learning Organization. Manage- ment Learning 29(1): 5–20.

Elsass, Priscilla M., and Laura M. Graves. 1997. Demographic Diversity in Decision-Making Groups: The Experiences of Women and People of Color. Academy of Management Re- view 22(4): 946–73.

Ely, Robin J., and David A. Thomas. 2001. Cultural Diversity at Work: The Effects of Diversity Perspectives on Work Group Processes and Outcomes. Administrative Science Quarterly 46(2): 229–73.

Falk, Gideon, and Shoshana Falk. 1981. The Impact of Decision Rules on the Distribution of Power in Problem-Solving Teams with Unequal Power. Group and Organization Studies 6(2): 211–23.

538 Public Administration Review • September/October 2004, Vol. 64, No. 5

Finkelstein, Sydney. 1992. Power in Top Management Teams: Dimensions, Measurement and Validation. Academy of Man- agement Journal 35(3): 505–38.

Fisher, Dalmar, and William R. Torbert. 1995. Personal and Organizational Transformations. London: McGraw-Hill.

Gibson, Christina, and Freek Vermeulen. 2003. A Healthy Di- vide: Subgroups as a Stimulus for Team Learning Behavior. Administrative Science Quarterly 48(2): 202–39.

Huber, George P. 1991. Organizational Learning: The Contrib- uting Processes and Literatures. Organization Science 2(1): 88–115.

Jehn, Karen A., Gregory B. Northcraft, and Margaret A. Neale. 1999. Why Differences Make a Difference: A Field Study of Diversity, Conflict and Performance in Workgroups. Admin- istrative Science Quarterly 44(4): 741–63.

Kanter, Rosabeth Moss. [1977]1993. Men and Women of the Corporation. 2nd ed. New York: HarperCollins.

Konrad, Alison M. 2003. Defining the Domain of Workplace Diversity Scholarship. Group and Organization Management 28(1): 4–17.

Lau, Dora C., and J. Keith Murnighan. 1998. Demographic Di- versity and Faultlines: The Compositional Dynamics of Or- ganizational Groups. Academy of Management Review 23(2): 325–40.

Lee, Fiona. 1997. When the Going Gets Tough, Do the Tough Ask for Help? Help Seeking and Power Motivation in Orga- nizations. Organizational Behavior and Human Decision Processes 72(3): 336–63.

Levin, Shana, Christopher M. Federico, Jim Sidanius, and Joshua L. Rabinowitz. 2002. Social Dominance Orientation and In- tergroup Bias: The Legitimation of Favoritism for High-Sta- tus Groups. Personality and Social Psychology Bulletin 28(2): 144–57.

Levitt, Barbara, and James G. March. 1988. Organizational Learning. Annual Review of Sociology 14: 319–40.

McGrath, Joseph E. 1984. Groups: Interaction and Performance. Englewood Cliffs, NJ: Prentice Hall.

Nevis, Edwin C., Anthony J. DiBella, and Janet M. Gould. 1995. Understanding Organizations as Learning Systems. Sloan Management Review 36(2): 73–80.

Nkomo, Stella. 1992. The Emperor Has No Clothes: Rewriting “Race In Organizations.” Academy of Management Review 17(3): 487–513.

Nkomo, Stella M., and Taylor Cox, Jr. 1996. Diverse Identities in Organizations. In Handbook of Organization Studies, ed- ited by S.R. Clegg, C. Hardy, and W. Nord, 338–56. London: Sage Publications.

Omi, M., and H. Winant. 1994. Racial Formation in the United States. New York: Routledge.

Ospina, Sonia. 1996a. Illusions of Opportunity. Ithaca, NY: ILR Press.

———. 1996b. Realizing the Promise of Diversity. In Hand- book of Public Administration, edited by J. Perry, 441–59. San Francisco: Jossey-Bass.

———. 2001. Managing Diversity in the Civil Service: A Con- ceptual Framework for Public Organizations. In Managing Diversity in the Civil Service, edited by UNDESA-IIAS, 5– 10. Amsterdam: IOS Press.

Popper, Micha, and Raanan Lipshitz. 1998. Organizational Learn- ing Mechanisms: A Structural and Cultural Approach to Or- ganizational Learning. Journal of Applied Behavioral Science 34(2): 161–79.

Ragins, Belle Rose. 1997. Diversified Mentoring Relationships in Organizations: A Power Perspective. Academy of Manage- ment Review 22(2): 482–521.

Ridgeway, Cecilia L. 1997. Interaction and the Conservation of Gender Inequality: Considering Employment. American So- ciological Review 62(2): 218–35.

Ridgeway, Cecilia L., and J. Berger. 1986. Expectations, Legiti- mation, and Dominance Behavior in Task Groups. American Sociological Review 51(5): 603–17.

Rudolph, J.W., S.S. Taylor, and E.G. Foldy. 2001. Collaborative Off-Line Reflection: A Way to Develop Skill in Action Sci- ence and Action Inquiry. In Handbook of Action Research, edited by P. Reason and H. Bradbury, 405–12. London: Sage Publications.

Schneider, B. 1987. The People Make the Place. Personnel Psy- chology 40(3): 437–53.

Shaw, James B., and Elaine Barrett-Power. 1998. The Effects of Diversity on Small Work Group Processes and Performance. Human Relations 51(10): 1307–25.

Shaw, M. 1981. Group Dynamics: The Psychology of Group Behavior. New York: McGraw-Hill.

Sidanius, Jim, and Felicia Pratto. 1999. Social Dominance: An Intergroup Theory of Social Hierarchy and Oppression. Cam- bridge: Cambridge University Press.

Smith-Lovin, Lynn, and Charles Brody. 1989. Interruptions in Group Discussions: The Effects of Gender and Group Com- position. American Sociological Review 54(3): 424–35.

Staw, B.M., L.E. Sandelands, and Jane Dutton. 1981. Threat- Rigidity Effects in Organizational Behavior: A Multi-Level Analysis. Administrative Science Quarterly 26(4): 501–24.

Steele, Claude M. 1997. A Threat in the Air: How Stereotypes Shape Intellectual Identity and Performance. American Psy- chologist 52(6): 613–29.

Taylor, Sully, and Joyce S. Osland. 2003. The Impact of Inter- cultural Communication on Global Organizational Learning. In Blackwell Handbook of Organizational Learning and Knowledge Management, edited by M. Easterby-Smith and M.A. Lyles, 212–32. Oxford, UK: Blackwell.

Thomas, David A., and Robin J. Ely. 1996. Making Differences Matter: A New Paradigm for Managing Diversity. Harvard Business Review 74(5): 79–90.

Tsui, Anne S., Terri D. Egan, and Charles A. O’Reilly III. 1992. Being Different: Relational Demography and Organizational Attachment. Administrative Science Quarterly 37(4): 549– 79.

Tsui, Anne S., Terri D. Egan, and K. Xin. 1995. Diversity in Organizations: Lessons from Demography Research. In Di- versity in Organizations: New Perspectives from a Changing Workplace, edited by M. Chemers, S. Okampo, and M. Costanzo, 191–219. Thousand Oaks, CA: Sage Publications.

Williams, Katherine Y., and Charles A. O’Reilly III. 1998. De- mography and Diversity in Organizations: A Review of 40 Years of Research. Research in Organizational Behavior 20: 77–140.