Introduction

In their seminal work on collaborative writing, Ede and Lunsford (1992) distinguished between two modes of collaboration: the masculine hierarchical and the feminine dialogic. They defined the hierarchical mode as being “carefully, and often rigidly, structured, driven by highly specific goals, and carried out by people playing clearly defined and delimited roles” and the dialogic mode as being “loosely structured,” the people within the group having fluid roles and “the process of articulating goals . . .[being] often as important as the goals themselves and sometimes even more important” (p. 133). Considering these definitions of the masculine and feminine modes, we see that the masculine focuses on completing the task while the feminine focuses on the process of completing the task. Arguably, a collaborative group needs to focus on both.

Defining success for a dialogically collaborative group as one that forms a “collective mind” as they complete a task, Jones (2007) asserted that a collaborative group must both “connect” by developing social cohesion and exert the effort to complete a task. In order for a task to be a genuinely collaborative creation, in other words, the group must become a single entity greater than the sum of its parts and capable of creations more profound than any of them could have created individually. In becoming a single entity, the group must negotiate with one another as they encounter both social and task related conflicts along the way. To negotiate, group members must “confront the dissonance” within the group’s collaboration, an act requiring the masculine act of asserting one’s voice, but in doing so “ group members must accept each other” (Murray and Morgan, 1991, p. 79), arguably a feminine act of seeking connection.

This qualitative study examines two online collaborative groups comprised of returning adult learners who performed both masculine and feminine acts in making their collaboration successful. Both of these groups were mixed gender with both genders in both groups performing acts usually associated with the other gender. A number of possibilities may explain why these group members behaved this way, including maturity and experience in collaboration, but other literature suggests that women may feel safer to assert themselves in online spaces, and this study suggests that men in online spaces may feel safer to act in more feminine ways, as well. This latter possibility suggests that the online space can be used for transformative learning that could help people to “unlearn” social encoding.

Review of the Literature

For a group to function in this manner, its members, regardless of their individual sex, must manifest feminine and masculine characteristics. But citing Belenky, Clinchy, Goldberger and Tarule’s (1986) remark “that the collaborative, egalitarian spirit so often shared by women should be more carefully nurtured in the work lives of all men and women,” Ede and Lunsford suggested that men do not have the capacity for the negotiating, yielding, nurturing, and empathizing that groups need for success. While supporting the notion that gender is a social construct and qualifying her argument with “many,” Lay (1991) also cited Belenky, Clinchy, Goldberger and Tarule (1986) and earlier work by Chodorow (1978) and Gilligan (1982), when she posited that these scholars “distinguished the separate knowing of many men from the connected knowing of many women” (1991).

Lay (1991) wondered also whether “men [could] learn these connecting strategies” (p. 155), but Rickly’s (1999) study that supported the notion that gender is a social construct compels us to wonder if men do not have to learn so called “feminine” characteristics so much as they have to “un-learn” socially imposed notions of masculinity just as when women assert their voices they must un-learn notions that women should be quiet and acquiescent. For men and women to engage in this “un-learning” transformative process, they likely must feel safe, and because of the characteristics of online collaboration, cyberspace may offer that safety under some conditions. Indeed, mixed gender online collaborative groups may be a fertile environment for men and women to overcome socially imposed gender roles.

To determine whether such learning can occur, we need to first consider the extant literature that examines the power differential between women and men in online spaces. Some of this literature suggests that women have no more power in online spaces than they do in physical space and are not any more likely to claim power. LeCourt (1999) asserted, for instance, that “’feminine’ modes of expression and ways of knowing [are] discounted as less powerful online, just as they are in the public realm” (p. 156), and studies by a number of other researchers supported that claim (Pagnucci and Mauriello, 1999; Sage, 2000; Selfe and Selfe, 1994; Boese, 1999). Selfe and Selfe (1994) argued, for instance, that the online environment reflects “a larger cultural system of differential power that has resulted in the systematic domination and marginalization of certain groups of students” including women (p. 481). In studies examined by Dene Grigar (1999), moreover, a number of other researchers have pointed to not just negative masculine behavior such as rudeness but also attempts to intimidate, silence and frighten women, behavior that clearly indicates that the ability to empathize and connect has largely been left undeveloped in men. As an empowering place where women can feel free to engage in feminine practices as well as those that society denies them, such as the power to express opinions and have them valued, cyberspace does not seem to offer a great deal.

Yet earlier work (Calhoon, 1996; Harasim, 1990; Hawisher and Moran, 1993; Hiltz, Johnson, & Murray Turoff, 1986) contended that cyberspace at least has the potential to make online classrooms more democratic, even if “not necessarily more egalitarian”, as Cooper and Selfe (1990) claimed. This important distinction suggests that the online classroom enables marginalized students to overcome the “self-surveillance” that Foucalt (1977) contended they practice in order to appease those in power. Thus in this earlier vision of cyberspace, a woman did not have the same preferred status that a male had, but she may have felt more empowered to make her voice heard.

Though tempered with experience, this early vision of cyberspace may still hold promise, and if the online environment provides a place where a woman may exhibit a previously repressed trait, then men, likewise, may feel freer to embrace their own repressed “feminine” characteristics. However, the extant research does not suggest the possibility because this research shows young men acting in virtual reality as they would in physical reality. This research may not yield such evidence because it looks at young men acting together in large groups or in large groups of mixed men and women. In these large groups, the young men act aggressively and competitively, likely because society trained them to do so.

A small online group of mixed genders may be a more fertile environment for women and men to embrace parts of themselves that they cannot in a physical space. In order for a collaborative group to succeed, all group members must take initiative and assert their voices, characteristics attributable to the masculine, while simultaneously negotiating and yielding on issues and nurturing and supporting their group mates, characteristics attributable to the feminine. If all group members must behave in these ways, then both men and women must claim characteristics that society assigns to the other gender. Because cyberspace provides distance and anonymity, participants in an online group may feel safer to adopt behavior that illustrates a blending of feminine and masculine characteristics.

Claiming dormant characteristics that society says belongs to the other gender might be easier for women to do because, socially, it could be less risky for a woman to express masculine traits such as asserting a voice because of the advances women have made. Highly motivated male students may be able to claim feminine traits in order to make a collaborative group successful, and the online environment may make it easier for them to do so because social inhibitions are more easily transgressed online; however, other factors such as emotional maturity and experience collaborating on the job may be equally important factors. Supporting the notion that men in online collaborative groups may feel more empowered to embrace “feminine” characteristics, Gillespie; Julier and Yancey (1999) reported that the men in an online group were “willing to enter into very emotional dialogues with the three women” and that they even “initiate[d] that sort of discourse (p. 306). This study painted a much different picture of male behavior online than some other studies and offered hope that the online environment may be a fruitful place for the personal growth of both men and women.

Mixed gender groups seem to be the ones most likely to be fertile territory for personal growth in men and women as people and as collaborators. In a study cited by Monroe (1999), Faigley (1992) reported that in online classes comprised predominantly of males, behavior tended to reflect the hostility, disagreement, and aggression associated with the masculine (p. 254), but in Monroe’s (1999) observations of her own class, one comprised of eleven females and three males, discussions became “almost oppressively consensual, disagreements routinely dissipating into acknowledgement and apology” (p. 69). These conclusions suggest that in groups comprised of their own gender, people may feel more compelled to behave in ways reflective of social expectations. Later in Monroe’s (1999) article, she reported noticing that her female students who acted with what she described as a Midwestern reserve in the physical classroom “were animated with uncharacteristic liveliness (p. 71) online, suggesting that the online environment made them feel free to fully express themselves, a freedom provided by cyberspace that others, such as Sherry Turkle (1995), have also noted.

As Ritke-Jones (2008) pointed out, the safety that the online environment affords, created by the anonymity and distance of participants, may make an online space perfect for transformative learning events. Using the space as a sort of cushion, learners in online groups, both male and female, may feel safer to assert their voices during a conflict and at the same time they both may find that the online space allows them to be more yielding, allowing them to develop greater empathy for one another. Finally, because asynchronous discussions can be archived, learners can reflect on their discussions under the mentorship of a skilled facilitator to see where they are being too aggressive or passive, masculine and feminine, respectively, to use socially ascribed tags.

Methodology

This study was a naturalistic inquiry into two mixed gender online groups. Based on a thematic analysis, the design and methods of analysis emerged as one of the researchers began looking closely at data to determine what made these groups successful and becoming intrigued by the possibility that the participants were transcending gender stereotypes. From that point, the two researchers designed the study so that they could draw conclusions about the interpersonal relationships in the groups and about the behavior of the participants in those groups.

Data Collection

Miles and Huberman (1994) provided the principles for data collection. The data consisted of coded transcripts of most communications within each group and interviews of all but one member of the two groups. Because every major communication was recorded, defined as anything over a word or two, the data provided a rich description of what happened in the groups during the time that they were in class together. Two sets of notes per researcher were made of the two groups from which themes emerged that provided the kernel for codes, satisfying Gertz’s (1973) criteria for “thick description.” Semi-structured interviews provided triangulation for the conclusions drawn from coded communications. These interviews were conducted using online chat technology such as AIM or MSN Messenger and lasted approximately one hour.

Settings and Participants

Both groups were taking first year composition courses entirely online, using asynchronous threaded discussion and online chat tools. None of the people in either group had met face-to-face before the classes were held, nor did they ever meet during the class. Both groups were comprised of non-traditional students over 30 years old. Group one consisted of two women and one man, and group two consisted of two men and one woman.

Researchers

Originally, one researcher collected transcript and interview data and began memoing data while the groups collaborated. Miles and Huberman (1994) defined memoing as “primarily conceptual in nature” and that memos “tie together different pieces of data into a recognizable cluster” so that in effect the researcher is writing notes to herself that will enable her to develop an overall picture of what is happening (p. 72). After the initial round, more intense memoing was accomplished by going through the data several times. From the memos, themes emerged. The data collector wanted the perspective of a differently gendered person, however, recognizing that her or his cultural encoding, background, and past experiences may have been creating a murky lens through which to see emerging themes. While the invited researcher did see things somewhat differently, fundamentally the researchers saw things similarly and reached consensus on their conclusions. This arrangement assures a high degree of validity and internal reliability.

Data Analysis

After initial memos, each of the researchers made two more sets of memos for each group. From these memos, collaboratively created codes based on themes related to the social construction of gender in the two groups were created. As shown in Table 1, these codes were divided into two categories, masculine and feminine, and were based on accepted notions of how society constructs gender.

The codes were created in an oppositional binary, for instance product/process, hierarchal/democratic, etc., and illustrates the either/or nature of our language and our society. As such, the codes themselves are social constructs. However, because the communications of each woman and man in each group were sometimes coded masculine and sometimes feminine, using these codes actually provided proof that gender is a social construct.

Most communications were coded, the event on the transcript being marked with such configurations as FN for feminine nurturing, or MA for masculine authoritarian. In many instances, a single communication was marked with several different codes to note that a number of things seemed to be happening during that communication. For instance, MPD + FN + MA could mean that in a single communicative event, a group member was being product oriented, nurturing, and active in presenting ideas.

After all communications were coded, they were entered into a matrix that allowed a holistic picture of each group’s communications to form. The matrix listed codes in a far left column and the names of the group members in the uppermost row. The rest of the matrix was filled with the number of times that each group member made an utt erance that was marked with a particular code. This configuration allowed a record of how many times a group member made an utterance that could be coded in a particular way.

Table 1. Masculine and Feminine codes employed.

Masculine	Feminine
Product oriented	Process oriented
hierarchal	democratic
authoritarian	egalitarian
competitive	collaborative
separated	connected
punitive	nurturing
active	passive
structured	fluid

While not coded, each researcher separately analyzed each interview before discussing how the interview verified or contradicted the way the transcripts were coded, thus providing triangulation. The interviews proved to be invaluable as triangulating data. Having a male and a female to analyze data also provided vital triangulation and assured the reliability and validity of this study.

Limitations

Data were only collected from two groups, so while conclusions have been drawn about these two groups, these conclusions cannot be generalized to other groups. Qualitative methods do not seek to draw general conclusions, however. Rather, the strength of qualitative data is their “strong potential for revealing complexity” from which a hypothesis can be formed that can be tested in the future with quantitative methods, more qualitative methods or mixed methods (Miles and Huberman, 1994, p. 10).

Group One

Group One consisted of one woman, Susan, and two men, Peter and Robert (pseudonyms) all over 30 years old. This well functioning group exhibited a high level of socio-emotional bonding as exhibited by various characteristics such as their genuine concern for the well-being of one another. This level of social cohesion fostered the evolution of a group consciousness, and this group consciousness developed a group “system” that allowed for the successful completion of group tasks, “successful” being defined here as a task that was not only completed but one in which the voices of all the group members blended into one after going through a collaborative process.

This group developed a well-refined system, a masculine structure, suggesting a hierarchy, yet the group was highly democratic and egalitarian. During a five-week course, the group made sixty posts. Susan made twenty-four of the posts (40%), Peter made twenty-three (39%), and Robert made thirteen (22%). The mean word count of Susan’s posts was 37, Peter’s was 28 and Robert’s was 24. The almost even split between the number of posts Susan made and those that Peter made suggests an egalitarian relationship. Susan’s higher average word count reflects her role as facilitator but also as caretaker. Robert’s apparent lesser participation does not truly reflect his commitment to the group and his participation in it because many of his communications occurred in chat, and chats were not recorded. The way that we coded the communicative events made by this group is shown in Table 2

Table 2. The coding of communicative events for Group One.

Masculine	Peter	Susan	Robert
Product	15	11	2
Hierarchal	3	1	0
Authoritarian	3	1	0
Competitive	1	0	0
Structured	12	13	4
Separated	0	0	0
Punitive	1	1	0
Active	0	0	0
Feminine	Peter	Susan	Robert
Process	2	3	2
Democratic	1	3	0
Egalitarian	1	2	0
Collaborative	8	11	3
Fluid	0	0	0
Connected	7	6	2
Nurturing	4	8	1
Passive	1	1	0

Susan

Of the forty-four times that Susan’s communications with her teammates received a code, twenty-three were coded masculine and twenty-one were coded feminine. Both her masculine and feminine codes, however, were in some way related to her being the primary cohesive force in the group as well as being its facilitator and coordinator. For instance, we coded her communications eleven times as being product-oriented, two as being hierarchal, four as being authoritarian, and six times as being structured. All of these communications, especially those coded “structured,” indicate that she took on the facilitator’s role, and she excelled at it. She kept her group mates on task and fostered the evolution of a system by which the group could perform its writing tasks. She also played the key nurturing, connecting role in the group, repeatedly communicating in ways that kept the group structured and moving toward its goal of completing an assignment but that also showed deep caring for her group mates. One such communication is the following:

Glad to have you back. We were very concerned. As it is not like you to be gone this long. Glad that everything is ok at your end. We have all but the last section completed. I will post my two tonight when I get home from work. Yes we both got your feedback. I did forward the grant/lee summary that I started to all of you. Please make any adjustments that you see necessary. . . .I have two of my three guys back online…this is a relief…lol

This communication, coded MH, MS, FN, and FCN, demonstrates Susan’s roles as group facilitator and primary nurturer and caretaker. Acting almost like a parent, she admonishes her group mate for not letting the group know that he was going to be absent for a while, but at the same time she displays genuine caring.

The Two Men

Of the two men, Peter’s communications received fifty-five codes, thirty-eight of which were coded masculine (70%) and twenty of which (30%) were coded feminine. Mostly, Peter’s communications illustrated a task oriented approach, a “let’s get it done” attitude. Almost one-third of his communications, however, were coded feminine-connected/collaborative (FC) and feminine-process (FP), indicating that his commitment to his group mates and to the collaborative process. In one such communication, he illustrated his connection to the group:

“Not sure what’s up with [Frank]. He’s MIA bigtime. I should have something on the Grant assignment shortly. I’ll be home the rest of the weekend and Monday. Sorry to be behind.”

In this posting, we see Peter’s concern for a group mate who fell ill early and had to drop the course (and thus was not part of the study because he made no communications) and his commitment to the group by letting them know that he was ready to work. Simultaneously, however, he focuses on finishing the product.

Robert made the fewest number of postings; thus his communications received only nine codes, six masculine and three feminine. Communications on the discussion board do not tell Robert’s entire story, however, because his main function in the group was to serve as a cohesive force and with Susan as the group’s creative force. In their interviews, Susan described Robert as “the whimsical one” and Robert described himself and Susan as “more on the creative side.”

We see this creativity in the assignments that Robert posts, but these assignments were not coded. Rather, we see Robert’s feminine characteristics such as nurturing in his postings about his group mates, especially about the peer who fell ill before the class started, but Robert used other technologies such as text messaging and chat to communicate, especially with Susan and Frank, so these postings are rare. Robert said during an interview that “we used IM like a traditional student commons, even better, we were in the Commons and the class at the same time. I would say that that is where the emotional connection was made.” Also in his interview, he reveals how important he was to the group’s socio-emotional connection when he talks about the loss of Frank to the group: “When [Frank] dropped out it was more of an emotional loss than anything else.”

Group Two

Group Two consisted of one man, Don, and two women, Beth and Sue, all over 30 years old. This group lacked the emotional bond of Group One, and as a probable consequence, it exhibited more hierarchal structuring and less nurturing and disclosing of personal subjects; nevertheless, the group committed to collaborating and nurturing their connection by supporting each other and applauding each other’s efforts and the team’s accomplishments.

Over the course of the semester, this group made 264 posts. Beth made eighty-three of the posts (31%), Don made eighty-one of the posts (rounded to 31%), and Sue made seventy-six of the posts (28%). The mean word count of Sue’s posts was sixty-four, Don’s was forty-three, and Beth’s was twenty-one. These numbers reflect the egalitarian relationship between Sue, Don and Beth, with Beth being the facilitator as well as the caretaker. The way that we coded the communicative events made by this group are shown in Table 3.

Don

Don definitely displayed characteristics defined socially as masculine, and even exhibited some sexism, referring to Sue and Beth regularly as “girls.” Moreover, Don demonstrated a more hierarchal, product oriented approach to the assignments. At the beginning of the semester, for instance, Don immediately wanted to form the group’s hierarchy, asking the three women how they should choose a coordinator and note taker for the group. Don’s quick request to the group to establish a hierarchy suggests that he may have wanted to be the group’s coordinator, but he did not push the issue.

Rather, on a number of occasions, Don simply started discussions, but instead of using questions to begin dialogue as Beth did, Don directed, asking for opinions after he had offered his contributions, as in this communication:

“I have made some additions to our team question. They are in bold print for ease of reading, also so they are easy to remove if we choose not to use it. I felt I needed to participate and here it is. Feel free to give your opinion, positive or negative.”

Coded MS, MP, MA, FC and FD, this communication started discussion on the assignment, fostered a structure for completion and also invited dialogue about it, but in a much more masculine manner than Beth’s way of inviting dialogue.

Table 3. The coding of communicative events for Group Two.

Masculine	Don	Beth	Sue
Product	22	23	12
Hierarchal	14	2	0
Authoritarian	16	11	4
Competitive	1	7	0
Structured	16	16	19
Separated	4	2	0
Punitive	2	5	1
Active	11	5	5
Feminine	Don	Beth	Sue
Process	4	6	3
Democratic	3	2	4
Egalitarian	1	1	1
Collaborative	15	24	14
Fluid	0	0	0
Connected	9	12	6
Nurturing	21	14	13
Passive	13	9	9

Despite his masculine approach, on many occasions during the semester, Don displayed a nurturing, democratic, collaborative attitude. To illustrate, the following communication Don expressed his conservative opinion about Martin Luther King’s “Letter from Birmingham Jail,” about which the group had to write a response, but he negotiated with Sue’s more liberal view to make the assignment successful:

“This was the worst. As I stated before, I really don’t like politics…..really! But I really like [Sue’s] approach. . . Let’s first discuss this one after looking at all of ours and try to come up with a collaborative conclusion.”

Coded, MS, MH, MAU, MA, FC, and FD, this communication distinctly shows Don’s efforts to move past his own views to incorporate those of his group mates, a feminine thing way of proceeding.

Besides his willingness to collaborate, Don also nurtured and supported his group mates, albeit in a masculine, rather sexist, manner, exclaiming in one instance: “HAVE A GREAT WEEKEND, GIRLS. TRY TO GET OUT AND PLAY. WE ALL DESERVE IT AFTER THIS WEEK J ” With obvious affection, Don orders his female group mates to do something self-nurturing, much as a father might order his daughters to take care of themselves. Nevertheless, Don’s nurturing belies the cultural stereotype of the cold, distant, and aloof male.

The Two Women

As we can see from the coding matrix, Beth’s communications were coded almost evenly masculine and feminine, 51% and 49%, respectively. This coding illustrates Beth’s balance of masculine and feminine characteristics. For instance, from the beginning, Beth took charge of the group, coordinating its efforts and making decisions on the group’s behalf, with the group’s approval. Being product focused, Beth made the first suggestion on how to perform the group’s task and with Don, initiated a discussion on how to structure the group itself. Throughout the semester, moreover, Beth kept the group moving towards task completion but by inviting and encouraging collaboration, as in the following post:

I received the emails from [Don]. I was wondering how we should go about choosing a team coordinator/note taker as well. Question one comes directly out of chapter 2, starting on page 20. Should we all give our opinion on the reading and go from there?

This post demonstrated Beth’s focus on completing the assignment (masculine) while at the same time wanting collaboration (feminine).

Sue’s behavior stayed aligned with cultural definitions of the feminine in that she easily adopted a collaborative, almost passive, attitude, yet on one occasion, she asserted her voice during a debate with Don. In this conflict about “Letter from Birmingham Jail” she firmly stated her opinion while simultaneously inviting Don’s different opinion:

As you can see from my response, I chose to answer it by talking about what I don’t like about President Bush (go figure J . I don’t really know how you guys want to handle this. If you want to work on what I have or if we should just choose our own and work from there. Either way is okay with me. I know we might not agree. Just let me know.

Conceivably, Sue may not have felt empowered enough to state her opinion in a physical space, but in cyberspace, she unapologetically claimed her voice while still maintaining a feminine posture towards collaboration.

Discussion and Conclusion

The people in these groups demonstrated that in the context of working with the particular people in their groups during the period that the groups worked together in cyberspace, they were able to transcend (or transgress) characteristics that culture assigns to each gender. We cannot know how these people would have worked together in physical space nor can we know how well that they may have worked together if they had not been strongly motivated adult students. Perhaps the most vital part of the equation, we also cannot know if these people would have transcended socially assigned gender roles if they had worked in all male or all female groups or if the members of the group had been different people.

To answer whether acting in cyberspace allowed these people to transcend gender roles in order to successfully form a collective mind, we can speculate by placing these groups within the frames provided by other researchers and scholars. It appears that the scholars mentioned in the literature review who asserted or supported the notion that gender performance in online spaces mirrors what happens in physical space were correct because clearly the people in these groups performed traditional roles. However, those scholars who asserted that cyberspace promotes an egalitarian and democratic environment where people feel safer to express repressed characteristics seem to have also been correct, judging by the way the people in these groups did not strictly adhere to traditional gender roles.

On the other hand, cyberspace may have had nothing to do with the success of the two groups. Recalling that all of the people in these groups who finished the course were adult learners, the intrinsic motivation of the typical goal-oriented adult learner may have made these people successful in any environment. Indeed, they may have transcended gender roles in any space in order to make their collaboration work. Moreover, from likely having been in more relationships and from having collaborated on more projects, these students had a greater well of collaboration experience from which to draw than the traditionally aged student.

Cyberspace may, indeed, be a place where people feel freer to express repressed characteristics, but ultimately, cyberspace’s power to help people claim themselves may be more in its ability to provide a fertile environment for transformative learning. Besides having a more “flattened” versus “hierarchal” environment that allows for more egalitarian communications, archived communications on which learners can later reflect can help learners to develop the critical self-reflection skills necessary for personal growth and transformation. To illustrate, Ritke-Jones (2008) explained how people involved in collaborative projects can be presented with their archived communications and asked to reflect on them. In order to encourage people to claim repressed characteristics, students could be directed to ask themselves questions such as “were you assertive” or “were you supportive,” and prompted to explain in detail in discussion with the rest of the group. Of course, more research needs to be conducted on how cyberspace can be used in this manner, especially as the web evolves and more technologies surface, but as educators we need to be ever mindful of our obligation to use our tools to their greatest potential.