A taxonomy of learning through asynchronous discussion.
As Internet-based teaching and learning have proliferated, researchers, theoreticians, and pedagogues have recognized that an educationally-viable environment requires students to interact with content, and with each other. This realization has lead to the widespread use of asynchronous discussion conducted through e-mail listservs and web-based bulletin boards. Anecdotal evidence and empirical research offer insights about both the benefits and the problems associated with asynchronous discussion (cf., Brown, 1997; Merryfield, 2001; Pena-Shaff, Martin, & Gay, 2001). Also, instructional theory can help instructors at all levels aim for the benefits and avoid some of the problems (cf., Goodyear, Salmon, Spector, Steeples, & Tickner, 2001; Knowlton, Knowlton, & Davis, 2000; Miller & Miller, 1999; Moller, 1998; Palloff & Pratt, 1999; Prestera & Moller, 2001). But, this body of literature does not fully consider the educational possibilities of asynchronous discussion. Missing from the literature is a description of learning that may come from asynchronous discussion.
This article puts forth a taxonomy that describes possible types of learning through asynchronous discussion. Both social cognitivism and constructivism are theoretical frameworks that might support asynchronous discussion; but because the online classroom has been heralded as an environment conducive to promoting a type of student-centered learning (Knowlton, 2000) that allows knowledge construction among students (Jonassen, Davidson, Collins, Campbell, & Haag, 1995), this taxonomy is based in a constructivist view of asynchronous discussion. This constructivist frame is appropriate because, in spite of the fact that most theoretical literature about asynchronous discussion focuses on prescriptions of how instructors should structure and facilitate an asynchronous discussion, constructivism emphasizes descriptions of how learning occurs. A descriptive analysis is consistent with the purpose of this taxonomy. In fact, a description of learning through asynchronous discussion must precede sound prescriptions of instruction. Perhaps putting such descriptions ahead of prescriptions would provide insights into unclear definitions and vague explanations of designing toward prescriptions. For example, as Bannan-Ritland (2002) pointed out, terms associated with prescription are often not defined carefully. After this taxonomy is explicated, implications for both research and pedagogy are offered.
A LEARNING TAXONOMY FOR ASYNCHRONOUS DISCUSSION
The taxonomy presented here provides a descriptive foundation for different levels of participation within asynchronous discussion and grounds those levels within learning theory. The goal of this taxonomy is not solely to report connections that are already evident through research; rather, the goal is to create the structure for a new framework that might be useful in describing learning within an asynchronous discussion context. Offered within each level of the taxonomy are discussion participants' likely perceptions of three central tenets of constructivism: (a) the educational utility of the environment, (b) collaboration, and (c) knowledge construction. Focusing on these three tenets as the macrostructure of each level begins a bridging of the gap between the asynchronous environment itself and the types of knowledge construction valued by constructivists. The microstructure of each level is grounded in relevant literature. This taxonomy is summarized in Table 1.
Weedman (1999) has noted that some participants take a passive approach toward asynchronous discussion. Passive participants--"lurkers" to use the web-based vernacular--read contributions to the discussion, but they do not participate. Because of their passivity, research tells us very little about the passive participant's behavior within asynchronous discussion. In general, though, passive participants view themselves as having no role in shaping the discussion environment. They also may not see the educational value of collaboration. They probably view knowledge as something that is transmitted to them, not as a dynamic entity that they can construct for themselves. Each of these assertions will be discussed in turn.
Views of the environment. Graham and Scarborough (1999) noted that lurkers "are not regarded as part of the environment" (p. 21). More importantly than dismissing them as irrelevant, a learning theory should attempt to explain the reasons for their passivity: One reason for passivity may be a participant's lack of understanding of the asynchronous environment. Participants sometimes, for example, may not know the mechanics of posting a message to a discussion. Hara and Kling (2000) noted that technical problems can cause a high level of distress among asynchronous discussion participants. In several online courses that I have taught, passive participants have confessed that they took the course to improve their own computer competence. This goal of improving one's computer competence suggests that some passive participants often view the use of computer technology as an end in itself; they have not considered the role of the asynchronous discussion environment in meeting content-based course goals.
Passive participation should not, however, automatically be associated with a lack of understanding about the mechanics of asynchronous discussion environments. Instead, perhaps passive participants are engaged in what Lave and Wenger (1991) referred to as "legitimate peripheral participation." Passive participants may remain on the periphery of asynchronous discussion because they are trying to understand confusing guidelines from the course instructor or ambiguous contributions from classmates. Ambiguous discussions are an element of the asynchronous discussion environment identified by Hara and Kling (2000) and Essex and Cagiltay (2001) as a potential source of distress for participants. Participants' passivity may also be purposeful for cultural reasons. Perhaps passive participants remain on the periphery because they are trying to learn the language of the discourse community and find an entry point into the discussion.
A participant's thinking style (or perceived thinking style) also may influence the need for legitimate peripheral participation. In considering distance learners, Liu and Ginther (1999) have noted that some learners may be analytic thinkers, while others may think visually. Visual thinkers might be more hindered in an asynchronous discussion environment because the primary mode of communication is text, which lends itself to more analytic types of thinking. Perhaps the concept of multiple intelligence (Gardner & Hatch, 1989) comes into play here as well. Some of Gardner's intelligences suggest the value of an asynchronous discussion, but other intelligences appear to mitigate against that value.
Views of collaboration. Passive participants in asynchronous discussion either do not value, or do not understand how to engage in, collaborative processes. As noted in the previous section, their lack of understanding of the mechanics of contributing to a discussion might hinder their efforts to collaborate, but other factors may contribute to a lack of collaboration as well. Consider, for example, that participants' past collaborations in educational settings probably involved face-to-face interaction. Seeing and hearing each other served as a reminder that the educational process is social and collaborative (Knowlton, 2000). Even when the instructor dominates a face-to-face course, audible and visual interactions humanize the learning environment. In asynchronous discussion, though, participants do not have these sensory interactions; as a result, participants sometimes feel anonymous and dehumanized. Furthermore, because gestures, facial expressions, and other physical elements can "contribute subtle (and sometimes not so subtle) meanings or attitudes" (Weiss, 2000, p. 48), the lack of visual and audible cues can, in fact, impede communication, and thus hinder the potential for learning. Feeling dehumanized and experiencing the inability to effectively communicate may lead to embarrassment. Embarrassment may cause some passive participants to withdraw from interacting with others. Their social self-efficacy (see Bandura, 1986) is damaged. Therefore, opportunities to collaborate are hindered. Hara and Kling (2000) noted that a lack of community with classmates and a lack of collaboration with instructors can heighten participants' distress.
Low self-efficacy is not the only viable explanation for a passive participant's view of collaboration. In many educational settings, interaction and collaboration are not valued. Students in these settings are expected to collaborate only with the instructor who serves as the "giver of knowledge." Even here, the word "collaboration" is not accurate. Students in these settings are not collaborating with instructors; they are mirroring the instructor's knowledge. On the rare occasion when an instructor in these settings does initiate a stronger interaction, the instructor still usually tightly controls that interaction. Because of the experiences of participants in teacher-centered classrooms, a new environment where participants are responsible for initiating interaction with classmates may be foreign and overwhelming. Participants are accustomed to the teacher-centered view of education; they do not know how to collaborate. This perspective contradicts the findings of Hara and Kling (2000), who noted that social isolation distressed students. On the contrary, this taxonomy supports the theory that passive participants often are overwhelmed by the expectation to interact with peers. What's more, the ways of interacting within asynchronous discussion are divergent and require participants to take control of an intellectual discourse. For a participant who comes from a traditional, teacher-centered environment, these expectations can cause social withdrawal.
Beyond this two-pronged argument explaining passive participants' understanding of collaboration, we must be aware of the lack of knowledge that we really have about the behavior of passive participants. Perhaps, for example, passive participants are trying to accommodate and assimilate group culture and values. Also, perhaps passive participants are engaged in collaboration with classmates, but their collaboration involves only small groups of classmates by way of private e-mail. They are not communicating in a forum where instructors can witness collaboration; nevertheless, they still may be collaborating with individual classmates.
Views of knowledge construction. The passive participants' views of knowledge construction emerge directly from their understanding of the environment and their perspectives regarding collaboration. Some participants may be overwhelmed by the need to be active; and without an active approach toward participation in asynchronous discussion, knowledge construction is unlikely to occur. As Speck (1998) noted, many students have been trained, not educated. They see knowledge as something constructed by others--usually the instructor, whom they view as endowed with formal authority and thus, by extension, with intellectual authority. This statement about students holds true for online discussion participants, as well. Passive participants view their own role as one of absorbing. The large drop rate in web-based courses may be indicative of this sense of being overwhelmed.
But, some participants simply may be more cautious and want to survey the contributions of their classmates before participating more fully. Participants may not recognize that asynchronous discussion is more akin to conversation than to formal writing in the academic sense (Knowlton, 2003a), and thus they feel pressure to develop cogent arguments that have been revised, edited, and finalized. To accomplish this goal of developing cogent arguments, they recognize the importance of understanding their own argument and counterarguments offered within the contributions of their classmates. They take a product view--not a process view--of knowledge construction. They want to contribute products of their own thinking. Their products can only be shared once they have engaged in a process. For the passive participant, the process is not worthy of being shared with classmates.
Developmental participants exhibit more active behaviors than passive participants. But, the contributions of participants operating at this level still do not substantively contribute to collaborative knowledge construction as described by Jonassen et al. (1995); Tam (2000); and other constructivists.
Views of the environment. Participants operating at the developmental level understand that meaningful interaction can occur in an asynchronous discussion environment, but they may be defining meaningful interaction in terms of the enjoyment that they gain from participating in asynchronous discussion, not in terms of the quality of their learning about course content. For example, during one of my online courses the National Basketball Association's (NBA) playoffs were dominating the national sports news. In one of the course's asynchronous "small talk" forums, a few participants were engaged in making predictions on teams still in the tournament and offering lamentations about teams that already had been eliminated from the tournament. This discussion had nothing to do with the course's content--a technology course for preservice teachers. Interestingly, two students who were very active in this discussion about the NBA were conspicuously absent from other asynchronous discussion forums that were more related to course content. This type of participation may be indicative of those who view the environment as analogous to an interactive video game. The developmental participant understands the asynchronous discussion environment not as a place for serious dialogue about course content; rather, they view the environment as an entertainment novelty.
Views of collaboration. To gain an understanding of developmental participants' views of collaboration, the description in this section emphasizes collaboration about course content. In other words, the types of small talk that were described previously do help participants see that they can meaningfully collaborate through online discussion. However, the constructivists' notion of collaboration transcends small-talk interactions. Instead, constructivists are concerned with the ways collaboration contributes to socially-negotiated knowledge construction (Jonassen, 1991; 2002).
Developmental participants usually do not contribute anything substantive to the content-based discussion, though they are not completely silent, either. That is, a two-word post--"I agree"--or a short sentence--"I thought there were some nice ideas in your contribution to the discussion"--acknowledges the role of interaction among participants. But acknowledgement is not responsiveness. The developmental participant is not adding anything new to the discussion, only tangentially reacting to contributions of others.
For the participant operating on the developmental level, though, these contributions may seem substantive. Developmental participants may view collaboration as a process of morale building and creating a sense of belonging within a community. McMillan (1996) identified "spirit" as one aspect of community. The developmental participants' comments develop a sense of spiritual community among participants by simply acknowledging that their comments "belong" to the discussion. Graham and Scarborough (1999) stated that a sense of community in computer-mediated communication "can be both personally and educationally enriching" (p. 21). Developmental participants may understand the personal aspects of collaboration--to, for example, develop community through a sense of positive encouragement (Palloff & Pratt, 1999) and social reinforcement (Prestera & Moller, 2001)--but they have yet to discover the educational enrichment that can come from online discussion. An allied interpretation of the developmental participants' collaborative efforts may be described in terms of reciprocity. Participants begin to see that they have responsibilities to each other. Therefore, they feel an obligation to contribute to the discussion as payment for what they are learning by reading the contributions of others.
This developing level of collaboration may be indicative of a deficit in the participant's past. Namely, developmental participants never may have been taught how to collaborate in educational settings. Even if the developmental participant has been taught to collaborate in formal educational settings, these lessons in collaboration may not translate into the asynchronous environment. Asynchronously, for example, notions of active listening--a central tenet of collaboration--take on a new dimension and form. Perhaps the developmental participant is struggling to collaborate meaningfully because of these new dimensions and forms.
Views of knowledge construction. As has been described, developmental participants have a myopic view of both the asynchronous environment and the power of collaboration. These views contribute to the developmental participants' understanding of knowledge construction. That is, if developmental participants view the environment as a mode of entertainment and collaboration merely as a source of social motivation, then they are likely to have a limited view of the ways that knowledge can be constructed through asynchronous discussion. Developmental participants do not view online discussion as a place for the creation of new knowledge, only as a place where knowledge can be exemplified by the sharing of basic facts and anecdotal experiences. That is, developmental participants do not contribute in the hopes of constructing, only in the hopes of being told that they correctly understand the facts and having their experiences validated as "normal." They want encouragement, not discussion; they want reinforcement of what they already know, not opportunities to construct that which is not yet known.
Why do developmental participants hold such a narrow view of knowledge construction? Participants often assume that a good contribution is related mostly--if not solely--to mechanics of writing. From a practical perspective, though, participants realize the difficulties of revising and editing discussion contributions. For example, if students take the time to revise and carefully edit their ideas, their contribution will be late coming to the discussion. That is, other participants might have posted the idea already, or perhaps the discussion will have taken a turn in a new direction by the time a carefully constructed contribution is offered. Also, as Jonassen (2002) has noted, many standard online discussion tools--like WebCT or Blackboard--are not designed to support certain types of problem solving. The lack of support that these tools provide for revising discussion contributions is one example. Because of these difficulties, participants may feel hampered by the environment. They assume that the more they write the greater a chance that they will make a mechanical error in their writing. Because of this misunderstanding of what constitutes a strong contribution to asynchronous discussion, participants limit themselves to short, and sometimes single-sentence, contributions.
Other explanations of the developmental participants' limited engagement in knowledge construction can explain their contributions, as well. The developmental participant is trying to gain an understanding of the relationship between classmates' ideas, which are often abstract and theoretical, and their own real-world experiences. Conversely, developmental participants may understand (and agree with) the concrete and practical aspects of a classmate's contribution to the asynchronous discussion, but they are struggling to extrapolate on concrete experiences to connect them with theory. Kolb and Fry's (1975) theory of prior learning can explain this phenomenon. Kolb and Fry argued that for prior experience to result in learning, a student must be able to identify abstract generalizations about that experience. The developmental participant may not understand how to make this transition from practice to theory in their discussion contributions. Land (2000) added further credence to this theoretical connection between prior learning and the behavior of developmental participants. Specifically, she noted that learners must be able to integrate new knowledge with past experiences and use past experiences to metaphorically represent new knowledge.
Though the attempts of developmental participants to translate contributions among theory and practice may not be fully successful, we should acknowledge the attempt as one aspect of a valid educational experience. That is, even short, reactive posts may demonstrate that students have begun a type of "intrapersonal interaction" (Berge, 1999) or "internal dialogue" (Morrison & Guenther, 2000). Perhaps, for example, they are actively engaged in making connections between contributions to the asynchronous discussion and their own interpretations of course readings. Perhaps they are tacitly aware of theory/practice connections, but they have not translated those connections into language. Said differently, the developmental participant may well be thinking about course content, but developmental participants bypass the opportunity to share the process of their thinking; they only share the final products of their thinking. Often, these products are a simple statement of agreement or dissention.
Generative participation is characterized as a participant's attempt to offer commentary about course content. A participant at the generative level views the asynchronous discussion environment as one conducive to articulating an idea. Yet, the nature of a generative contribution does not blatantly show a valuing of collaboration. Participants operating at the generative level view knowledge construction as a private and solitary act. Regardless of the perspectives that participants have about the environment and collaboration, some theoretical evidence exists showing that generative participation can lead to knowledge construction.
Views of the environment. Often, the generative participant's view of the environment is the same as the developmental participant's view. To the extent that generative participants recognize the interactive possibilities of the environment, they view interaction as a novelty, not as an educational necessity. What is added in the generative participant's view of the environment, though, is recognition of the need to use the environment to communicate with the instructor. Participants view the environment as teacher-centered. They struggle to see asynchronous discussion as an authentic place for academic discourse. They use the environment not so much to create healthy opportunities for learning; but to earn a grade. In this respect, generative participants simply view the environment as a place to report to the instructor what they know. From this, generative participants have begun to understand the environment as a mind tool, but they disregard the environment as a tool for distributing and sharing their ideas, focusing instead on the power of the environment to support one-way communication from student to the instructor. The findings of Nicholson and Bond (2003) seem to support this view of the environment, as participants' use of the environment appeared to display a strong "preoccupation with themselves." Even participants' use of the "subject line" evidenced this preoccupation (e.g., "Melonie's Reading Buddy," or "Calvin's message," p. 268).
Views of collaboration. Because generative participants view the asynchronous environment as teacher-centered, they do not view themselves as a part of a dialogue with classmates. Instead, they view themselves giving a monologue to the instructor. The purpose of this monologue is to earn a grade. In this respect, generative participants have a narrow understanding of collaboration. They are, after all, engaged in a type of collaboration with the instructor.
Other explanations for the generative participants' emphasis on a monologue might exist, as well. In a traditional academic model, for example, students have limited time to voice their perspectives. So, collaboration often takes a back seat to the isolated and divergent voices of individual students. Furthermore, the traditional academic model is often based on competing perspectives. Prowess is measured by students' dominance in a conversation, not their ability to find collaborative synthesis. Along this line, students believe that instructors will determine a grade by comparing students to each other to determine that grades fall into a bell curve. The point is that generative participants are influenced by this traditional academic model and thus view collaboration in connection to earning a grade. The fact that others may read their contributions to the discussion is secondary to what they are saying. When taken to the extreme, we can say that the emphasis that generative participants place on monologue might be an act of arrogance--a belief that their perspectives are superior to those of other participants and thus should be given more attention. Less cynically, the generative participant may at least have developed a sense of communal trust. Rovai (2001) declared trust as one aspect of community. Generative participants trust that classmates will treat their contributions with respect.
Views of knowledge construction. Initially, attempts with offering generative commentary may serve the purpose of earning a high grade. For example, when asynchronous discussion is required, participants may be asked to generate a contribution to the discussion by organizing content through summaries and paraphrases, integrating content by applying it to (or grounding it in) new contexts, and elaborating on it in an effort to create something new. In general, participants are not depending on interaction to spur the construction of knowledge. As participants write a contribution to a discussion, they may be aware that an audience exists in the form of their classmates, but the audience is incidental. They are not writing to an audience, they are writing in spite of the audience. Two theories can help explain knowledge construction through generative participation. The first is Wittrock's (1989) theory of generative learning. The second is the theory of "writing to learn."
First, generative participants' contributions should be recognized as a response to a generative learning assignment. Though the purpose of this article is to examine descriptions of learning, as opposed to prescriptions for teaching, we cannot separate the two completely. Generative participants offer contributions to discussions in response to a prescriptive request from an instructor. Wittrock's (1989; Wittrock & Alesandrini, 1990) generative strategies may lead participants toward organizing course content in a contribution to a discussion by writing a summary or paraphrase. Perhaps participants may synthesize course readings with their own experiences. Importantly, Wittrock's work has been connected to online discussion. Morrison and Guenther (2000) suggested that organization, integration, and elaboration generative learning strategies can be valuable in the confines of an asynchronous discussion assignment.
Second, as participants generate content through the act of writing, they can construct knowledge. Indeed, writing is a learning activity (Adams & Hamm, 1990; Young & Fulwiler, 1986). As Fulwiler (1986) stated, "It's not important that writers know exactly where they are going when they start; it's important [that] they trust the process of composing to take them somewhere" (p. 22). Consider, too, a letter written to newspaper advice columnist Ann Landers. The author of the letter began by asking Landers a question: "[S]hould I marry the guy or not." But after describing her fiancee in the letter, she concluded by saying, "Do not bother to answer this. You have helped me more than you will ever know" (quoted in Lindemann, 1995, p. 6). The act of writing allowed this letter-writer to better understand, analyze, and solve her own problem. This notion of writing as a learning activity is transferable to online discussion. For generative asynchronous discussion participants, knowledge is constructed through organizing and elaborating on their own ideas.
Both generative learning and writing-to-learn can be connected to an internalization model of cognition. As Wegerif (2002) explains, learners initially view content as external to them. But through responding to specific instructional strategies, learners internalize the content with which they are interacting (e.g., Knowlton, Eschmann, Fish, Heffren, & Voss, 2004). For the generative participant, knowledge construction is the internalization of content as a result of the act of contributing to the online discussion.
Dialogical participation involves a substantive interaction among asynchronous discussion participants. The notion of "substance" in the interactions suggests that participants understand the legitimacy of asynchronous discussion. Dialogical participants become situated as a part of a community of learners. Learners recognize themselves as situated because they accept "the mutual relation of content and context, of individual and environment, and of knowing and doing" (Barab, MaKinster, Moore, Cunningham, & ILF Design Team, 2001, p. 73). To some extent, the three angles from which this taxonomy has examined learning from a constructivist environment become artificial at this point: Dialogical participants recognize that the environment itself allows knowledge to be constructed, and they view collaboration as a source for knowledge construction. In spite of the fact that these three characteristics of constructivism merge, they are treated here separately for ease of explication and consistency.
Views of the environment. Dialogical participants begin to understand that the environment can increase clarity of communication in ways not possible through face-to-face discussion. They view the environment as a medium for increasing comprehension and focusing on task completion. Walther (1996) supports this perspective by noting that computer-mediated communication may foster an emphasis on task completion more than a face-to-face environment. This perspective has also been confirmed empirically. Jonassen and Kwon (2001) found that students in asynchronous discussion were more task-oriented in problem-solving activities than face-to-face groups of students who were attempting to solve problems.
Further, dialogical participants begin to understand that they can structure the environment to facilitate stronger collaboration, and thus more durable knowledge construction. For example, while many asynchronous discussion participants view basic "rules" supporting strong online discussion as random and purposeless, dialogical participants recognize that following some of these "rules"--such as using the "reply" button to create threaded discussion, double spacing between paragraphs, and using emoticons to clarify the tone of statements that might be misinterpreted (Weiss, 2000)--are not random. Rather, dialogical participants recognize that the savvy use of paragraphing and other conventions of computer-mediated texts can increase clarity of their own contributions and assist them as they interpret the ideas of others.
One indicator of a participant who has a dialogical understanding of the asynchronous environment may be a strong use of asynchronicity. The asynchronous environment allows participants to reflect on previous contributions to a discussion before they respond to those contributions (Lewis, Treves, & Shaindlin, 1997). Participants operating at the dialogical level are likely to take advantage of such an opportunity. As a result of the opportunity to reflect before responding, participants are able to engage in more meaningful exchanges (Prestera & Moller, 2001). Also, dialogical participants likely will recognize the potential for integrating a past contribution to a topic as a key benefit of the environment. By reviewing archived portions of a discussion, for example, participants may be able to quote their classmates or even their own past contributions to begin to synthesize portions of the discussion. The physical synthesis of archived messages may be evidence of cognitive synthesis by the participant.
Views of collaboration. Dialogical participants begin to understand the value of not only consuming classmates' ideas, but also interacting with classmates about those ideas. Dialogical participants are not just concerned with generating an individual perspective. Instead, they become engaged with seeing how others relate to their perspectives. Earlier in this article as developmental participants were described, Graham and Scarborough's (1999) idea of community as being both personally and educationally enriching was noted. For the dialogical participant, collaboration develops a sense of community that is educationally enriching.
Numerous relevant connections can be made between this new awareness of collaboration and theoretical literature about teaching and learning. Speck (1998), for example, notes that instructors must be aware of the pluralism among any group of students. For dialogical asynchronous discussion participants, this concept of pluralism takes on a new significance. They become aware of the importance of the "reciprocal relations of individuals" (Barab et al., 2001, p. 76).
This new awareness is in stark contrast to the generative participation described in the previous section of this article. While generative participants become aware of the ways their own contributions are socially situated within the contexts in which they find themselves, dialogical participants become aware of the situated cognition of other discussion participants. That is, dialogical participants begin to recognize that other participants' contributions, interpretations, and constructions are situated within specific and unique contexts that may be different from their own. Such recognition allows cognition to be distributed. Collaboration becomes viewed as collective thought. (The notion of distributed cognition will be more fully explicated in the next section of this article.)
In some respects, delineating the contrast between the generative and dialogical view is to represent the hierarchical development of asynchronous discussion participants. This notion of participants developing into the dialogical perspective of collaboration is not without support. Nicholson and Bond (2003) observed "development over time from postings that merely served to report first one [participant's] experience and then another's with no replies, to invitations and requests for feedback, and ultimately, to veritable give-and-take conversations among [the asynchronous discussion participants] that extended over time" (p. 267). This observed development fits nicely with the taxonomy of this article in that the contributions to the discussion that reported experiences were generative, but the invitations, requests, and give-and-take were dialogical. The notion of this shift occurring "over time" is indicative of the participants moving "up" the taxonomy.
Views of knowledge construction. To begin an examination of dialogical knowledge construction, we should acknowledge some difficulties in recognizing knowledge construction through dialogue. Distinguishing between knowledge that is constructed through dialogical participation and knowledge that emerges through generative participation can be difficult. For example, as previously noted, threaded discussion might be an indicator that dialogical construction is occurring. But, we would be remiss to assume that participants' physical threading of a response is the result of a true dialogue. For example, Pena-Shaff, Martin, and Gay (2001) determined that most participants "read and constructed their comments based on other participants' messages. However, [they] were unable to find explicit collaboration between participants in the [bulletin board] discussions" (p. 54). Similarly, it is equally tempting to assume that acknowledging another's ideas demonstrates that participants are constructing knowledge through dialogue. In reality, a participant's acknowledgement of a classmate's ideas might simply be a combination of developmental participation and generative participation in the same contribution. That is, a participant may acknowledge a classmate's contribution (developmental participation), but then the participant may develop additional ideas without further consideration of the contribution that is being replied to (generative participation).
Though difficulties exist in recognizing dialogical knowledge construction, we should not dismiss such construction as impossible. Indeed, strong feelings of community among participants contribute to an increased flow of information (Rovai, 2001), and thus efficiency of knowledge construction. As participants are introduced to classmates' opinions and perspectives through online discussion, they are likely to encounter new ideas about course materials (Knowlton, 2003a). For some participants, these new ideas will conflict with their preconceived personal views and perspectives. In an effort to create continuity between their preconceived views and the new ideas that they have encountered, participants are likely to consciously and proactively revise their own ideas in an effort to eliminate cognitive dissonance between what they believe and what others believe.
Once the ideas are revised, dialogical participants will use the processes of debate and discussion to test their ideas through further asynchronous discussion with other participants. The processes of experiencing dissonance, revising, and testing revised ideas creates more cognitively complex and intensely personal learning (Tam, 2000). As Palloff and Pratt (1999) note, for this process to be truly interactive, the participants must be fully inter-dependent. In this respect, knowledge construction has occurred as a result of distributed cognition. The very essence of thinking (a precursor to knowledge construction) is no longer a solitary act. This view is consistent with DeHaan (2002), who argues that "the unit of analysis [for understanding knowledge construction] is no longer the individual's cognitive system, but the interrelatedness of a diversity of intellectual systems" (p. 33). Wegerif (2002) describes the distributed cognition model as a contrast to the internalization model of cognition, which was used earlier in this article to describe the knowledge construction of generative participants.
Though the construction of knowledge may be distributed among a group of participants, a final interpretation of these constructions can only be understood in relationship to the self. This view is consistent with a view about teaching and learning that is articulated elsewhere in the literature: Learning about the self and one's own learning processes is a key purpose of education (Knowlton, 2003b). This view is further consistent with Berge (2002, p. 181) who distinguishes between "interactive learning" (dialogical participation) and "reflective learning" (metacognitive participation). Because participants who view their own contributions to computer-mediated communication as strong are likely to learn more than those who do not have positive regard for the role that they play in asynchronous discussion (Leinonen, Jarvela, & Lipponen, 2003), it is appropriate that the highest level of participation is on the metacognitive level. Metacognition has several components, but central to an understanding of metacognitive participation is the idea that metacognition is connected to "internal mental representations" of the learning process (Hacker, 1998, p. 3).
Views of the environment. The metacognitively-savvy participant views the environment as being well-suited for examining personal transformation. The asynchronous environment allows discussion participants to carefully monitor their own comprehension of the discussion content. Through this type of monitoring, participants come to understand the environment as a place not only conducive to learning about the content of the discussion, but also to learning about the self. This connection between a learning environment and metacognition is consistent with the ideas of Lin (2001), who argues that some environments lend themselves to metacognitive processes more than others.
As a result of this new realization that the environment can allow for metacognitive learning, some participants access the environment to learn about the self. The archives of a discussion, for example, can be used to learn, but metacognitive participants are not solely focused on a reconsideration of course content; instead, they are looking for evidence of personal change. They are considering the relationship between their current and previous perspectives.
Views of collaboration. Part of metacognition involves developing strategies for generating knowledge and monitoring one's own knowledge. Participants can contribute to each other's development of strategies. This can happen overtly. The metacognitive participant, for example, might ask other participants about strategies that they used to understand a text or interpret an assignment. But such collaborative development of monitoring strategies also might be less overt. Sitko (1998) examined such covert strategies in light of coauthoring. Coauthoring provides students with alternative patterns of articulating their ideas because students must integrate their ideas and writing style with the ideas of their peer coauthor. These alternative patterns can cause students to expand their flexibility in terms of the way they express ideas. A similar process can occur in asynchronous discussion. Through reading the contributions of others, participants learn about their classmates' views of course content, but participants at the metacognitive level may also learn about alternative approaches to expressing their own ideas in the confines of an asynchronous discussion (Knowlton, 2003a).
Metacognitive participation transcends engaging in a dialogue. Metacognitive participants not only consider their own role in the asynchronous dialogue, but also they consider how other participants view their role--they see themselves in the second person. This second-person "seeing" may manifest itself in the form of rhetorical questions about their own role in collaborative relationships. Occasionally, metacognitive participants might elicit comments from other participants that help them answer these rhetorical questions--"What contributions have I made to the discussion that you found particularly helpful?" But, more often than not, metacognitive participants' consideration of collaboration is, ironically, internal. They are constantly "sizing up" themselves and the role that they are playing in the social structure of the discussion, but they do not share their introspections with others. Regardless of the exact form that metacognitive collaboration takes, metacognitive participants view other participants as potential mirrors for better seeing themselves.
Views of knowledge construction. Participants operating at the metacognitive level understand that knowledge construction transcends the boundaries of asynchronous discussion. Metacognitive knowledge construction can emerge in two different ways. First, participants might acknowledge the ways that participation in the discussion has shaped their construction of self. How has the conversation changed them? In what ways can they describe their own cognitive restructuring? These questions are consistent with Palloff and Pratt's (1999) view that a true community of participants developed in cyberspace should lead to transformative learning.
Second, participants may be able to reflect on the discussion itself and explain how knowledge construction occurred for them throughout the life of the discussion and perhaps beyond the discussion itself. For example, perhaps a participant has constructed some knowledge that is relevant to another aspect of a course, such as writing a research paper or preparing for a presentation to classmates.
The taxonomy presented in the previous section is obviously not the result of an empirical endeavor. Rather, the taxonomy offers theoretical connections to previously-existing literature. Even the use of the word "theory" may be problematic; therefore, this taxonomy might better be framed using some of the "working words" of Kelly (2004): "explanation," "framework," or "argumentative structure" (p. 123). Regardless of how it is labeled, the taxonomy described in this article can support future work in design and pedagogy, theory, and empiricism.
Course Design and Pedagogy
This article has, to a large extent, ignored prescription in light of the need to articulate a description of learning through asynchronous discussion. A central argument underlying this article is that description is needed prior to the development of more carefully-considered instructional prescriptions for asynchronous discussion. One theoretical connection that now needs to be made is between description and prescription.
Salmon (2000), for example, implies that developmental participation--such as participating in a small talk discussion forum--can help former lurkers move into a more active type of participation. Salmon's view is consistent with Prestera and Moller (2001) who argue that facilitators must change the mindset of participants and help them "break out of their stereotypical roles of information receivers"--lower levels of this taxonomy--and take on the roles of "seekers, explorers, and users" (p. 100)--indicative of the upper levels of this taxonomy. Morrison and Guenther (2000) suggest that instructors should facilitate participants' interactions and not play the expected role of stimulus provider. But, how can online course designers and instructors execute the ideas of Salmon, Prestera and Moller, and Morrison and Guenther? What types of design structures and facilitation can help participants move from lower to higher levels of the taxonomy? Both online course designers and instructors should pursue development opportunities that are designed to answer such questions.
Numerous theories were considered in constructing the taxonomy presented in this article, but this article has not bridged the entire gap between instructional theory (prescription) and learning theory (description) in the context of online discussions. The explanations offered within this article, however, have made a step toward a learning theory, but much work needs to be done. How, for example, do other learning theories influence the taxonomy offered in this article. Where, for example, does the notion of critical thinking fit in? Kuhn (1999) describes a developmental model of critical thinking, and stronger connections need to be made between critical thinking and asynchronous discussion. Jonassen and Kwon (2001) have explored problem-solving in computer-mediated communication. Like critical thinking, problem solving is an assumed "act" in the knowledge construction process. Furthermore, Nicholson and Bond (2003) notes the place of reflection within asynchronous discussion, and reflection assumes both critical thinking and certain types of problem solving.
Critical thinking, problem solving, and reflection are related, yet nebulously so, and theoretical work should explore the relationships and, more importantly, how these relationships manifest themselves in asynchronous discussion: How is problem solving related to critical thinking in asynchronous discussion? Are not they both related to metacognition? Yet, they are not synonyms. Since a learner must situate the self within problems solving, how are critical thinking and metacognition different from reflection, and how do these differences manifest themselves within the type of problem solving that is necessary to be successful within asynchronous discussion? Jonassen (2001; 2002) has begun some of these distinctions through a careful delineation of problem-solving taxonomies and prescription. But, more theoretical work is needed in these areas.
Addressing these questions becomes complicated when asynchronous discussion is considered from a domain-specificity perspective. The framework presented in this article is generic with respect to specific fields of study. But, surely, asynchronous discussion about biology is significantly different from asynchronous discussion about history (Knowlton, 2003a). In other words, the framework that is articulated in this article assumes that hard sciences, social sciences, and humanities all produce the same types of learning. This assumption may not be sound, at least in some levels of this taxonomy. For example, Dunlosky (1998) has noted that we must consider metacognitive processes in light of domain-specific tasks; similarly, Knowlton (2003b) has argued that problem solving is domain specific. To what degree does a specific content or specific body of knowledge constrain or enhance the contributions that asynchronous discussion can make to learning? Learning theorists across disciplines should address this question.
Some of the theoretical implications of this article are related to empirical implications. Design research, for example, has a central purpose of developing theory in the form of "ontological innovations--the invention of new scientific categories" that support further theory development (diSessa & Cobb, 2004, p. 78). Through an empirical approach, the framework in this article could be extended and clarified. Perhaps within a design research framework, the beginnings of this taxonomy as a conceptual frame could help researchers overcome the view that design research tends to be under conceptualized (Dede, 2004). More intriguing, Dede noted that ontologically driven research should evolve from taxonomies to causal relationships or from taxonomies to theories to design heuristics. The taxonomy presented in this article can serve as a starting point for such design research.
Beyond the ideas for theoretical developments that could evolve from empirical research, other areas need to be investigated empirically. Some research exists that examines students' experiences in asynchronous discussion, but such research is bound by discipline. For example, de Vries, Lund, and Baker (2002) examined the role of computer-mediated dialogue in helping learners understand concepts in physics. Other domain-specific investigations are needed, as well. Furthermore, empiricism should begin to focus on specific types of participants through descriptive inquiry. Most obviously, little is known about the behavior of passive participants. Empiricism should explore their activities (and lack thereof). Even as discussed in the dialogical level of participation, participants may be intentionally passive in an effort to actively listen--another type of "legitimate peripheral participation" (Lave & Wenger, 1991). Empiricists should design studies that examine the reasons for peripheral participation and the level of participation that various forms of peripheral participation indicate.
How exactly do students move in and among the various levels described in this taxonomy? This article suggests a linear move from the lowest level to the highest, but empiricism could probably demonstrate that such movement is not linear and quite complex. Some empiricism already suggests this lack of linearity, as the types of contributions that are described as generative and contributions that are dialogical become "mixed" throughout the life-span of an online discussion (Nicholson & Bond, 2003, p. 273). Additional research needs to be conducted in this area. For example, Lin (2001) reports that one approach to metacognition is consistent with the taxonomy presented in this article because it assumes that metacognitive awareness can be achieved in an environment that is conducive to metacognition. On the other hand, Lin pointed out that teaching participants to use metacognitive strategies can also promote metacognition. Research needs to be done on the degree to which learning changes based on the amount of metacognitive training a learner has.
Table 1 Summary of the Taxonomy and Its Constructivist Characteristics Passive Developmental Views of environment * Lack knowledge of * View discussion as a environment logistics novelty, not a * Uncomfortable with the serious educational text-based nature of tool discussion Views of collaboration * Do not understand * Focus on morale collaboration, group building and culture and/or values creation of * Collaborate with small community groups of classmates * View collaboration in terms of the need for reciprocity Views of Knowledge * Do not view themselves * Exemplify Construction as constructors of information to knowledge determine their own * Adopt a product-- comprehension not process--view * Struggle with of participation articulating theory/practice connections Generative Dialogical Views of environment * Understand the * Understand the environment as tool environment as a tool for communicating for increasing clarity with the instructor and maximizing interaction with other participants Views of collaboration * Emphasize their own * Emphasize community ideas through among participants monologue * Acknowledge the * Begin developing situated cognition of communal trust other participants Views of Knowledge * Respond to * Attempt to alleviate Construction prescriptions from cognitive dissonance instructors created by discussion * Construct knowledge * Understand knowledge as through the act of being solely writing internalized Metacognitive Views of environment * View the environments potential for promoting personal transformation Views of collaboration * Search for extensions of strategies for expressing ideas * View others as a mirror for "seeing" themselves Views of Knowledge * Explain how they Construction constructed knowledge through participation in discussion * Understand knowledge as distributed among participants
Adams, D.M., & Hamm, M.E. (1990). Cooperative learning: Critical thinking and collaboration across the curriculum, Springfield, IL: Charles C. Thomas Press.
Bandura, A. (1986). The explanatory and predictive scope of self-efficacy theory. Journal of Social & Clinical Psychology, 4(3), 359-373.
Bannan-Ritland, B. (2002). Computer-mediated communication, elearning, and interactivity: A review of the research. Quarterly Review of Distance Education, 3(2), 161-179.
Barab, S.A., MaKinster, J.G., Moore, J.A., Cunningham, D.J., & the ILF Design Team (2001). Designing and building an online community: The struggle to support sociability in the inquiry learning forum. Educational Technology Research and Development, 49(4), 71-96.
Berge, Z.L. (1999). Interaction in post-secondary web-based learning. Educational Technology, 39(1), 5-11.
Berge, Z.L. (2002). Active, interactive, and reflective elearning. Quarterly Review of Distance Education, 3(2), 181-190.
Brown, A. (1997). Designing for learning: What are the essential features of an effective online course? Australian Journal of Educational Technology, 13(2), 115-126.
Dede, C. (2004). If design-based research is the answer, what is the question? A commentary on Collins, Joseph, and Bielaczyc; diSessa and Cobb; and Fishman, Marx, Blumenthal, Krajcik, and Soloway in the JLS special issue on design-based research. The Journal of the Learning Sciences, 13(1), 105-114.
DeHaan, M. (2002). Distributed cognition and the shared knowledge model of the Mazahua: A cultural approach. The Journal of Interactive Learning Research, 13(1/2), 31-50.
deVries, E., Lund, K., & Baker, M. (2002). Computer-mediated epistemic dialogue: Explanation and argumentation as vehicles for understanding scientific notions. The Journal of the Learning Sciences, 11(1), 63-103.
diSessa, A.A., & Cobb, P. (2004). Ontological innovation and the role of theory in design experiments. The Journal of the Learning Sciences, 13(1), 77-103.
Dunlosky, J. (1998). Epilogue: Linking metacognitive theory to education. In D.J. Hacker, J. Dunlosky, & A.C. Graesser (Eds.), Metacognition in educational theory and practice (pp. 367-381). Mahwah, NJ: Lawrence Erlbaum.
Essex, C., & Cagiltay, K. (2001). Evaluating an online course: Feedback from "distressed" students. Quarterly Review of Distance Education, 2(3), 233-240.
Fulwiler, T. (1986). The argument for writing across the curriculum. In A. Young & T. Fulwiler (Eds.). Writing across disciplines: Research into practice (pp. 21-32). Portsmouth, UK: Boynton/Cook.
Gardner, H., & Hatch, T. (1989). Multiple intelligences go to school: Educational implications of the theory of multiple intelligences. Educational Researcher, 15(8), 4-9.
Goodyear, P., Salmon, G., Spector, J. M., Steeples, C., & Tickner, S. (2001). Competencies for online teaching: A special report. Educational Technology Research and Development, 49(1), 65-72.
Graham, M. & Scarborough, H. (1999). Computer mediated communication and collaborative learning in an undergraduate distance education environment. Australian Journal of Educational Technology, 15(1), 20-46.
Hacker, D.J. (1998). Definitions and empirical foundations. In D.J. Hacker, J. Dunlosky, & A.C. Graesser (Eds.), Metacognition in educational theory and practice (pp. 1-24). Mahwah, NJ: Lawrence Erlbaum.
Hara, N., & Kling, R. (2001). Students' distress with a web-based distance education course: An ethnographic study of participants' experiences. Information, Communication & Society 3(4), 557-579. Retrieved January 12, 2002, from http://www.slis.indiana.edu/CSI/Wp/wp0001B.html
Jonassen, D.H. (1991). Objectivism versus constructivism: Do we need a new philosophica paradigm? Journal of Educational Research, 39(3), 5-14.
Jonassen, D.H. (2001). Toward a design theory of problem solving. Educational Technology Research & Development, 48(4), 63-86.
Jonassen, D.H. (2002). Engaging and supporting problem solving in online learning. Quarterly Review of Distance Education, 3(1), 1-13.
Jonassen, D.H., & Kwon, H.I. (2001). Communication patterns in computer mediated versus face-to-face problem solving. Educational Technology Research and Development, 49(1), 35-52.
Jonassen, D.H., Davidson, M., Collins, M., Campbell, J., & Haag, B. (1995). Constructivism and computer-mediated communication in distance education. The American Journal of Distance Education, 9(2), 7-26.
Kelly, A.E. (2004). Design research in education: Yes, but is it methodological? Journal of the Learning Sciences, 13(1), 115-128.
Knowlton, D.S. (2000). A theoretical framework for the online classroom: A defense and delineation of a student-centered pedagogy. In R.E. Weiss, D.S. Knowlton, & B.W. Speck (Eds.), Principles of effective teaching in the online classroom, (pp. 5-14). San Francisco: Jossey-Bass.
Knowlton, D.S. (2003a). Evaluating college students' efforts in asynchronous discussion: A systematic process. Quarterly Review of Distance Education, 4(1), 31-41.
Knowlton, D.S. (2003b). Preparing students for educated living: Virtues of problem-based learning across the higher education curriculum. In D. S. Knowlton & D.C. Sharp (Eds.). Problem-based learning in the information age, (pp. 5-12). San Francisco: Jossey-Bass.
Knowlton, D.S., Eschmann, A., Fish, N., Heffren, B., & Voss, H. (2004). Processes and impact of journal writing in a graduate-level theory course: Students' experiences and reflections. Teaching & Learning: The Journal of Natural Inquiry and Reflective Practice, 18(2), 40-53. Retrieved May 26, 2004, from http://www.und.nodak.edu/dept/ehd/journal/
Knowlton, D.S., Knowlton, H.M., & Davis, C. (2000). The whys and hows of online discussions. Syllabus: New Directions in Educational Technology, 13(10), 54-58.
Kolb, D.A., & Fry, R. (1975). Toward an applied theory of prior learning. In G. Cooper (Ed.). Theories of group processes, (pp. 33-57). New York: John Wiley & Sons.
Kuhn, D. (1999). A developmental model of critical thinking. Educational Researcher, 28(2), 16-25.
Land, S.M. (2000). Cognitive requirements for learning with open-ended learning environments. Educational Technology Research and Development, 48(3), 61-78.
Lave, J., & Wenger, E. (1991). Situated learning: Legitimate peripheral participation. Cambridge, MA: Cambridge University Press.
Leinon, P., Jarvela, S., & Lipponen, L. (2003). Individual students' interpretations of their contribution to the computer-mediated discussions. Journal of Interactive Learning Research, 14(1), 99-122.
Lewis, D.C., Treves, J.A., & Shaindlin, A.B. (1997). Making sense of academic cyberspace: Case study of an electronic classroom. College Teaching, 45(3), 96-100.
Lin, X. (2001). Designing metacognitive activities. Educational Technology Research and Development, 49(2), 23-40.
Liu, Y., & Ginther, D. (1999). Cognitive styles and distance education. The Journal of Distance Learning Administration [On-line], 2(3). Retrieved October 10, 2001, from http://www.westga.edu/~distance/liu23.html.
Lindemann, E. (1995). A rhetoric for writing teachers (3rd ed). New York: Oxford University Press.
McMillan, D.W. (1996). Sense of community. Journal of Community Psychology, 24(4), 315-325.
Merryfield, M.M. (2001). The paradoxes of teaching a multicultural education course online. Journal of Teacher Education, 52(4), 283-299.
Miller, S.M., & Miller, K.L. (1999). Using instructional theory to facilitate communication in web-based courses. Educational Technology and Society, 2(3), 106-114.
Moller, L. (1998). Designing communities of participants for asynchronous distance education. Educational Technology Research & Development, 46(4), 115-122.
Morrison, G.M., & Guenther, P.F. (2000). Designing instruction for learning in electronic classrooms. In R. E. Weiss, D. S. Knowlton, & B. W. Speck (Eds.), Principles of effective teaching in the online classroom (pp. 15-22). San Francisco: Jossey-Bass.
Nicholson, S.A., & Bond, N. (2003). Collaborative reflection and professional community building: An analysis of preservice teachers' use of an electronic discussion board. Journal of Technology in Teacher Education, 11(2), 259-279.
Palloff, R.M., & Pratt, K. (1999). Building learning communities in cyberspace: Effective strategies for the online classroom. San Francisco: Jossey-Bass.
Pena-Shaff, J., Martin, W., & Gay, G. (2001). An epistemological framework for analyzing student interactions in computer-mediated communication environments. Journal of Interactive Learning Research, 12(1), 41-68.
Prestera, G.E., & Moller, L.A. (2001). Exploiting opportunities for knowledge-building in asynchronous distance learning environments. Quarterly Review of Distance Education, 2(2), 93-104.
Rovai, A.P. (2001). Building classroom community at a distance: A case study. Educational Technology Research and Development, 49(4), 33-48.
Salmon, G. (2000). E-moderating: The key to teaching and learning online. London: Kogan Page.
Sitko, B.M. (1998). Knowing how to write: Metacognition and writing instruction. In D.J. Hacker, J. Dunlosky, & A.C. Graesser (Eds.), Metacognition in educational theory and practice, (pp. 93-116). Mahwah, NJ: Lawrence Erlbaum.
Speck, B.W. (1998). The teacher's role in the pluralistic classroom. Perspectives, 28(1), 19-43.
Tam, M. (2000). Constructivism, instructional design, and technology: Implications for transforming distance learning. Educational Technology & Society, 3(2), 50-60.
Walther, J.B. (1996). Computer-mediated communication: Impersonal, interpersonal, and hyperpersonal interaction. Communication Research, 23(1), 3-41.
Weedman, J. (1999). Conversation and community: The potential of electronic conferences for creating intellectual proximity in distributed learning environments. Journal of the American Society for Information Science, 50(10), 907-928.
Wegerif, R. (2002). Walking or thinking? Images of thinking and learning to think in the classroom. The Journal of Interactive Learning Research, 13(1/2), 51-70.
Weiss, R. E. (2000). Humanizing the online classroom. In R.E. Weiss, D.S. Knowlton, & B.W. Speck (Eds.), Principles of effective teaching in the online classroom, (pp. 47-52). San Francisco: Jossey-Bass.
Wittrock, M.C. (1989). Generative processes of comprehension. Educational Psychologist, 24, 345-376.
Wittrock, M.C., & Alesandrini, K. (1990). Generation of summaries and analogies and analytic and holistic abilities. American Educational Research Journal, 27(3), 489-502.
Young, A., & Fulwiler, T. (Eds.). (1986). Writing across disciplines: Research into practice. Portsmouth, UK: Boynton/Cook.
DAVE S. KNOWLTON
Southern Illinois University Edwardsville, IL USA