Skip to content

Composition Forum 49, Summer 2022
http://compositionforum.com/issue/49/

“Essential Allies in the Construction of Knowledge”: A Conversation with Lee Odell, Dixie Goswami, and Anne Herrington

Neil Baird and Bradley Dilger

As we shared in our introduction to this special issue, we’ve been working together since 2010 to study writing transfer as students transition into their majors. The discourse-based interview (DBI) has featured prominently in our research methods, and due to its ability not only to surface tacit writing knowledge but to encourage deep reflection in our participants, we developed an empirical study to learn about the impact and evolution of the DBI in writing studies (IRB #1509418-3).

This study began with a citation analysis and inventory of scholarship using the DBI, which we completed in collaboration with graduate students Brian Urias and Ran Meyer, and led to interviews with scholars in writing studies and other fields who have employed the DBI in their own research, including Lee Odell, Dixie Goswami, and Anne Herrington. With these interviews, we learned much about how the DBI developed, what all of our participants noticed about the impact and innovations of the DBI since 1983, how mentoring new researchers was vital for the use of the DBI, and what they understood as the future of the DBI in writing studies.

In Fall 2020, we were able to interview Odell, Goswami, and Herrington individually, sharing some of the findings from other participants and our citation analysis. We were then able to facilitate a group interview with them over Zoom in March 2022. The core of this interview piece is the group interview. Where appropriate, we’ve drawn from these individual interviews to support their discussion.

Many of the questions in our conversation guide emerged from conversations with other scholars. These scholars often shared how they came to the DBI as a method, and this made us curious how Odell, Goswami, and Herrington worked together to develop and theorize the DBI. As scholars shared how they adapted the DBI for their specific research questions, we became curious about the kinds of adaptations Odell, Goswami, and Herrington also made across the years. To help support our discussion, we shared a brief document outlining findings from our citation analysis and interviews. Our citation analysis also highlighted the importance of mentoring. We found that those taking up the DBI often had a chair or reader on their dissertation committee who employed the method in their own research. We were thus eager to learn how Odell, Goswami, and Herrington mentored newcomers to the method.

After transcribing and reflecting on their individual interviews, we noticed subtle differences in the narratives they shared. For example, one would downplay their involvement in developing the method, while the others would talk through that person’s importance to the process. We thus reached out to facilitate a group interview, eager to see how Odell, Goswami, and Herrington would respond to each other, and were able to hold this group interview over Zoom in March 2022. Our conversation guide was similar to their individual interviews. At the time, our draft introduction was circulating among our contributors, and we shared it with Odell, Goswami, and Herrington before the interview to facilitate conversation.

The core of this Composition Forum interview piece is this group interview. However, we also incorporate passages from their individual interviews. To do so, we created a reverse outline for the group interview to give us a birds-eye view of the themes running throughout. For example, in responding to our first question about the importance of studying tacit writing knowledge today, Odell referenced the changing nature of discourse and genuine curiosity, Goswami highlighted COVID-19 and other crises and the importance of personal narrative, and Herrington focused on lack of respect for learners’ tacit knowledge. We then turned to their individual interviews to see how these themes played out, often moving passages over to clarify and even magnify the narratives they shared in their group interviews.

Once we had a draft of the complete interview, we shared it with Odell, Goswami, and Herrington as a member check, inviting them to make clarifications or to request redactions. We honored most of their requests. However, in a few cases, we asked for reconsideration, offering rationale for why we thought certain passages should be retained, or be modified to better represent their wishes, mutually deciding how to move forward. Consequently, while what we share here is not a transcript of our group interview, it accurately represents both the interactions we had in our conversations, and the complexity and depth of our individual interviews as well.

Screenshot of video conference interview with Baird, Goswami, Odell, Dilger, and Herrington.

Figure 1. Clockwise from upper left: Neil Baird, Dixie Goswami, Lee Odell, Bradley Dilger, and Anne Herrington.

The Central Role of Tacit Writing Knowledge for the Discourse-Based Interview

Neil Baird: Anne, Lee, and Bradley, Dixie and I were talking before you folks joined, and she mentioned wanting to talk about the importance of this Composition Forum special issue and the urgency of using the DBI to study tacit knowledge today, nearly 40 years after the publication of “The Discourse-Based Interview.” So, what do you folks believe about the importance of researching and studying tacit knowledge today? Why is that still so important?

Lee Odell: I’ve been spending a lot of time in recent years learning about visuals, and it has occurred to me, I have no way to get at the visual choices people make. I know that artists are reluctant to talk about visuals. What I’m missing right now is all of the tacit knowledge that guides what these people are doing. And I feel like if we could get at that, we could help beginning visual people work better.

Dixie Goswami: Yes. That’s central. That’s letting us think about reconceptualizing discourse and what it means for 2022. And it is complicated and fascinating. And we know almost nothing, as Lee said, about tacit knowledge and understanding of people in the role of art.

Lee: A while back, I read a piece in The New York Times about an avalanche. They integrated still images, video clips, pictures, and an illustration of how an avalanche formed to show movement across planes, and put that together with the visual stuff. That was stunning. You know, after reading that, I understand avalanches now. (Laughs.) I’d love to get inside those people’s minds to learn how they did that.

Neil: Dixie, you had mentioned that there’s a certain urgency to making visible tacit knowledge today, based on your work with young people and their adult mentors. So do you mind sharing what you meant by that?

Dixie: Absolutely. Well, I’ll refer to an article in The Atlantic by George Packer. He says we owe our beleaguered children, the victims of our inadequacy, the chance to be better than we are. He talks about the situation they’re in because of COVID and a whole constellation of crises that they have experienced that nobody else has. Not to mention what their families and teachers have experienced!

We are at a crossroads. We are always at a crossroads. But the one in public education involves big analytics. So billions of dollars are being invested right now to organize curricula and increasingly hybrid education for the poor. The poorer the public school students are, the more likely they’re going to get commercialized instruction. And he writes about what that means: an undemocratic approach.

The entire premise of pedagogy in public education being a democratizing process is at risk these days. So it’s urgent for us to develop methods that enhance discourse-based, participatory tacit knowledge gathering.

Anne Herrington: I think the value of trying to elicit the tacit knowledge of learners flips that top-down kind of approach. As I think Dixie is saying, learners’ knowledge is not respected. And these big top-down kind of approaches don’t value that. Discourse-based interviewing begins with a premise of trusting the knowledge of speakers and writers. Trusting their knowledge should drive education, and I think that discourse-based interviewing brings that trusting relationship to the fore.

Lee: When I started researching writing, I thought, “Students don’t have anything to say. I’ll go talk to adults.” Now, when I talk to students in our graphics programs at Rensselaer, I realize, “Holy mackerel, they can do stuff!” But I don’t know how they do it. So they can do stuff now that is certainly beyond me. And I need to understand.

Anne: Well, and that also begins to ask us to take apart what we mean when we say “tacit knowledge.” I will admit that I’m not up to date with the current theorizing around that. But what I think Lee is alluding to is tacit knowledge around discourse that’s so internalized that we can’t articulate it. I can’t articulate the ways of speaking I grew up with. We can notice the differences, but we can’t necessarily articulate them.

But there’s also some more recently learned knowledge that we can begin to articulate. I mean, that’s what I found when I was interviewing undergraduate students. There’s some things they had learned, maybe in classes a few years ago, that they were doing without thinking too much about it. So that’s tacit, and they could explain it, like, “Well, I learned that in my comp class.” They could articulate that tacit knowledge, and that’s still valuable to learn. So there’s at least two dimensions that I know of. The value is getting at both kinds of knowledge as they get articulated in immediate situations.

Dixie: I need to rethink any version of discourse-based interviews in light of participatory, contextualized interviews. I’ll refer to Jackie Royster who would say, unless everybody, the interviewer and the interviewee, is given a framework and a chance to present their own narratives in any way they wish to, as part of the research, the interview is going to be decontextualized. Following that line of thought, all discourse-based interviews are incomplete unless contextualized by narrative.

Lee: One of my favorite participants in an interview was one of the supervisors. I asked a question, and it unfolded all this stuff about his context. And when he finished he said, “Damn. I didn’t think I knew all that.” That’s what we want to find out.

Dixie: But, we have to insist it’s participatory. If we’re going to do this, we have to lay it on the table.

Anne: Well, and that means opening up interviews. So, you may start with a certain set of questions you want to ask, but you’ve got to open that up. You’ve got to open that up to the choices of texts that the participant wants to bring in, and you’ve got to open it up to what you’re going to talk about.

Lee: Yes. One of my favorite kinds of prompts was, “Well, tell me more about that.” You know, just a genuine curiosity about what participants were thinking and why. And that means you may go someplace you never thought you were going to go. But you may learn something.

Anne: It could be an extensive discussion of the chemical engineering process. (Laughs.)

Lee: That’s key for me. Being genuinely curious and appreciative of what they say. Listening carefully. Focusing on general terms and value terms, not to argue but to find out what they think and know. Sometimes people would surprise me and themselves as well.

Dixie: In analysis too. Lee is responsible for saying, “We won’t go in with anybody’s notion of how participants do this analytical work. Every category will arise from the transcripts, from the patterns.”

I’d like to add the urgency of what we can learn from young people, teachers, everybody who has gone through the shift during COVID and other crises. . I don’t think you can do individuals anymore! I think you’re going to have to do networks, groups within settings.

Let’s say that we’ve got a small network, one with about 20 different people in it, and a diverse group, geographically, ethnically, and racially. Could DBIs identify that point at which they become a discourse community of writers? That’s why you say you’ve built productive relationships that will let you construct knowledge across difference.

So, we need to look at that network and ask DBI participants, “You tell me where you changed, where the nature of your responses changed, and what other changes clustered around that.” A tool we’ve got now is everything’s recorded. So we have that data.

Lee: Yes. One of my painful memories is the arrogance with which we approached things some years ago. We knew the questions to ask. We knew the things we wanted. And that was a serious limitation.

Lee: What was frustrating as I tried to work with teachers on visual communication was that their conception was desktop publishing. They had no conception of what was possible. Even students who are so-called “disadvantaged” probably have a phone. And they probably do things on that phone that we don’t know about. They have access to technology.

Dixie: They’re doing everything. They are creating YouTube videos, they’re doing art, they’re doing extraordinary things. Youth activists promote and achieve change that matters. That’s not going away. These young people are essential allies in the production of knowledge. They have formed communities that are so powerful that they reach millions. Discourse-based interviews could develop the evidence, the data and the narratives to demonstrate that children and young people are essential participants in DBIs with extensive and complex tacit knowledge.

Anne: Yes, that’s the shift in assumptions we need. Everyone potentially is a knowledge source. It’s not that someone is going to give us some information that’s lesser than what someone else knows.

How Changes in Understanding Tacit Knowledge are Changing the Discourse-Based Interview

Neil: Anne mentioned earlier that she hasn’t been keeping up on the theoretical movements about tacit knowledge. But within this conversation, you all are sharing how tacit knowledge has changed. We’re hearing that it’s now created by networks of people, for example. Over the years, how have your understandings of tacit knowledge changed as you’ve learned from writers?

Dixie: I was powerfully influenced by Lee and Anne and by Shirley Brice Heath, Jackie Royster, James Britton and Michael Armstrong. I learned a lot from Michael, seeing how he sat down with a series of drawings and stories seven year-olds had written to get at the tacit knowledge that they brought to what they were producing. It was never one-shot. It was a collaborative process. Being influenced by Armstrong’s methodology about the collaborative nature of exchanging and sharing tacit knowledge was a big influence on me. And Royster’s rhetorical approach-narrative contextualization-was foundational.

Britton said there are three kinds of writing. One is perfunctory. One is engaged. And one is impelled. A discourse-based interview will let you know what kind. If a student or an adult or an executive says, “Oh, yeah. You can change it. I don’t care what you put there,” it’s perfunctory writing in the first place. With engaged, they’ll argue with you a little bit. But he says, if you’re interviewing someone and that person argues, “Never in a million years would I use this choice rather than that one,” then you’ve got impelled writing. And if it’s impelled, you had hit paydirt. Because they understood this rhetorical choice was the only one they would accept in a situation that was really important to them. And that’s what I’ve always been interested in: what matters to writers.

Lee: What’s changed for me are the things about which we might be asking. That is the networks, the visuals, etc. The other thing is what we might be curious about. Years ago, all I could think about was audience, purpose, and voice. But now, we can ask, “How does a choice reflect your narrative about your personal experience?”

One thing I utterly failed to follow up on was a woman who sometimes signed her letters differently: Margaret Smith, Meg Smith, M. Smith, Meg, and MS. And I said, “Would you be willing to say Margaret Smith here rather than MS?” And she said, “No, because, I’ll be honest, some in the agency don’t know I’m female. If they know I’m female, I won’t get the serious attention.” So there’s a whole realm of “Who am I in this agency?” that I didn’t follow up on. That’s the kind of thing we need to learn.

Bradley Dilger: I think we’re seeing some of that in the special issue. There’s a lot of folks that are really advancing the discourse-based interview to engage identity a lot more. When you combine that with a participatory approach, it’s really powerful.

Lee: Oh, yes. Again, I was interviewing a history professor and was asking about choices. And he said, “Oh, no, I want to display my learning.” And I didn’t follow up and say, “What do you mean?” Why did he want to display his learning? What did displaying his knowledge entail? Here was a whole conception of himself and his discipline. I just didn’t tap into it. Now I wish I had.

Dixie: Coming to the point where we can construct knowledge together, learn to communicate, and value collective knowledge rather than more hierarchical or individual knowledge, that’s the way changing understandings of tacit knowledge are changing DBIs.

Lee: Chronologically, you can look at changes, say, as a novice in an organization moving through various stages, or you can look into changes to a draft as somebody is doing a screencast, looking at changes and saying, “Okay, you did this, but what about that?” That will probably tell us a good bit about how they see themselves, how they’re changing, how their conception of the audience or organization is changing. So, a chronology of changes would be really interesting.

Anne: Yes. And to consider the backgrounds of the various participants as they come into that conversation. Some people will come in with much more familiarity with working on Zoom, say something like that. They will probably be participating differently from someone who initially comes in, versus someone just trying to figure out this tool. So obviously, considering the tool is something very important here. And what knowledge or experiences from participating in other kinds of tools do you bring to this? What may or may not work very well. That would be important to putting together the process as that group starts to come together and negotiate together.

We talked earlier about our understandings of tacit knowledge evolving. It seems to me that what’s most valuable is that construct as a single construct is being opened up into various dimensions. For example, as soon as you move into the classroom, say, when I was researching a lab and then a process design course. In the process design course, students are really learning a new genre just as they’re learning, literally, a new role for themselves. So it’s not tacit knowledge that’s being unlocked; it’s learning-on-the-go knowledge, and it’s their working-it-out kind of understanding. A very much in-process kind of understanding. It’s that in-process, figuring out how you’re going to do a given genre, as well as the kind of personal understandings. So there are multiple payoffs that you have to think about to even notice.

Defining Characteristics of the Discourse-Based Interview

Neil: So what, for you, are the characteristics of a discourse-based interview?. There are moments where Bradley and I are reviewing our citation research, and we see scholarship described as discourse-based interviews that raises questions for us based on the methods as written.

Lee: Yeah. I have strong feelings about this.

Neil: Well, it’s okay if there are differences between you folks. That’s what the special issue is all about. Are there characteristics that you feel need to be a part of a DBI in order to get at tacit knowledge?

Lee: For me, central: the writer’s choices. The writer’s choices. Sometimes they do X; sometimes Y. What’s the difference? Also, if we didn’t focus on writers’ choices, then we would be just playing English teachers again. So, focusing on choices, and following up and listening to what the writer says and drawing them out.

Anne: I would say, similarly, that it’s respect for the knowledge of the participants. It’s interviewing, using texts that the participants have created. Now, I would say that could include revisions or drafts. That’s still within the repertoire of that writer, and I found that elicits a lot, not only about tacit or prior learning, but also about that immediate situation that you can really unpack. As Lee said, that’s what you want to get to: the unpacking.

I think one of the things that was more important initially, as Lee and Dixie developed it, particularly around businesses, was looking at routine texts. That’s not necessarily required for a discourse-based interview. The artifact doesn’t have to be a routine text. And, indeed, you can learn a lot, certainly in a school environment, when someone is trying to write a text in a new environment.

I remember a lot of the interviewing I was doing was with students learning new texts. And in some instances, they were applying what they had learned from prior texts, what they knew from that immediate situation. And you want to get all that out. To go to Dixie’s point, if you can put that with their own kind of narratives... I think by the time Marcia Curtis and I were doing Persons In Process, we were still interviewing students around some revisions of texts they had made. That was all part of a much larger contextualized study, but all that work shared Lee’s respect for the knowledge of the participants and working with texts they have created.

Lee: Picking up on your notion of echoing a teacher’s voice, I remember once I asked as student, “Could you do X?” And he said, “No, that was one of the two sacred sins.” So there was just knowledge of what you could and could not do, and I failed to find out about it.

Dixie: Can we approach variations and uses of the discourse-based as participatory. Meaning that you have to contextualize it in participation with all the people who are going to be part of it, and part of that contextualization would be everyone bringing their narrative to the interview, the story that they can explicitly tell about themselves. Originally, we were focused on tacit rhetorical knowledge, the rhetorical choices that people made.

A way forward is to examine and learn from participatory DBIs that are now being reported in your collection, and learn how how they help us find questions to the most urgent issues to surviving as a democracy. And that does mean bringing in everybody from the beginning to the final interpretation. We need to bring in the people who produce the data, who gave us the data—we want their take on it. And that would give us another layer that we need. Because we don’t know the markers to look for until DBI participants tell us what we missed. That would be moving to a participatory model.

Designing Studies That Include Discourse-Based Interviews

Neil: We wanted to end with a question that targets an audience we see for our special issue: scholars new to writing research. If you were mentoring a novice researcher taking up the DBI for the first time, what is it that you would want them to know? Anything you would want to share with a person trying out the DBI for the first time or thinking about using it?

Bradley: Yes, and I’ll add that “novice researcher” doesn’t necessarily mean “graduate student.” I came to empirical research having been promoted to associate professor and tenured. I didn’t have any training in it. My graduate school just didn’t offer that at all. So, we’re hoping that the special issue can help fill that gap a little bit, for people interested in research but with limited opportunities for training.

Lee: My first thought would be, “Remember, you’re not the teacher.” That’s the first thing. “What you are is the learner.” You listen and give them a chance to talk.

Anne: Yes!

Lee: Find out what they have to teach YOU. Follow the student. You learn what the choices are, and then follow where they lead you. And not just accepting any blather, either. Ask follow-ups: “Okay, how do you mean that? Well, how does that fit? I don’t understand. How do you mean that?” Genuine curiosity about the reasoning behind the choices. That’s what I would share.

I think the way researchers present themselves, from the very first moment, is absolutely crucial. Because I was always conscious, you know... I walk in there. I can look sort of serious, and I’m tall. So I have to work very hard, as I said earlier, to say “I’m not the English teacher here.” And once people understand you’re genuinely interested in what they do, and what they know, that will really open the gates pretty well. So it’s absolutely essential to say something like, “I’m an English teacher, and I realize that people outside of my profession have powerful insights into ways of doing things. And so what I’m interested in doing is learning some of the ways, some of the knowledge that you all have developed, to guide the various choices you make.”

Dixie: I would say the first thing you’ve got to do is experience it. So you better find a partner. You can’t do this on your own. You will have that person do this with you. And you’ll do this with that person. And you’ll get together and see what you’ve learned. But please don’t start on a research design and go in and do it without having experienced it yourself.

Anne: I think Dixie and Lee have spoken really well to what’s most important. I guess I would say on top of that, you need to be systematic in deciding, “What features am I going to ask about?” For any interview, preparation is important, but in DBIs you have to know your rationale for asking about a particular choice. Unless you’re wanting students to just try to remember and say whatever comes to their mind as they move through it, in which case they’re doing stimulated recall. There’s intention behind choices!

Also, as you go through that process, be humble in your interpretation, so that you’re not imposing your own frame, as much as you can possibly avoid doing that. I mean, being self-aware of what you try to bring to something. I think that’s part of respecting those who are speaking and, as Lee’s saying, I think being the learner.

Dixie: Anne, I’d add that you need a kind of preliminary study, so that at least you have some kind of working knowledge of the nature of the discourse community that the individual you’re working with is functioning in. And that’s what’s complicating your work now, Neil and Bradley, because what it means to be in a discourse community is different. It was pretty easy for us. Now, when you look at the discourse community as an online. . . We used to think of it just at work. And our ideas of audience are incredibly complicated now, but still, they’re there.

Lee: I particularly like the point you just made, Dixie, about a partner. If you have two perspectives on something, you will see something your partner doesn’t see, or they will see something you don’t see. The whole will be greater than the sum of the parts. You’ll be better off. I never did that. That was a mistake. I didn’t know how to do it.

Anne: There’s something to not only having a partner but also having an observer. I found it a lot in my research seminars. I would even ask people to model some kinds of interviewing, and to have a pair go through that with others observing it. They could comment and see how the interview was going and whether the interviewer was coming in a little too overpowering, or dropping things, or not opening up. I mean, there are a lot of ways that kind of observation and comment can add in that learning process.

Lee: So if you have an observer, you make the assumption that the interviewer has tacit knowledge, too, and they’re acting on that tacit knowledge and probably not aware of it. So, it’d be very useful. You could ask, “Did you really mean to do that?”

Bradley: That’s a really important way to think about it. The tacit knowledge of the interviewer, too. How would an observer bring that into the picture in a way that’s participatory? Would it be the same way that you work with a participant?

Dixie: I think it’s important. Again, just thinking about Royster. The interviewer needs to tell you about herself and the experiences related to the focus of the interview. That would have to be part of it. Royster models that in her own work as a rhetorician, where she includes her own story.

Neil: We try to do this in our introduction. It’s not easy.

Anne: Since this is a retrospective, to think contextually, when the work was being developed was a time when composing process research was really coming into importance. Very cognitive-based. So here’s the discourse-based interview, which is looking initially from that social-rhetorical perspective, and taking what in composing process models was stored representations, and opening them up to some really valuable social-rhetorical knowledge.

I think as discourse based interviewing opens up, we see that it’s bringing important personal kinds of information, too, that are part of that narrative. What Dixie was getting into. And so when you start to ask someone about a choice... I’m thinking about a student we studied. In a psychology lab report, he was asked to do things that were against his basic value system. That’s one of the things you can learn that expands what we can and should be getting out of discourse-based interviewing. I think what contributors are saying in this special issue is we need to open that up even more.

Neil: Thank you so very much for being willing to join us and have this kind of reunion. We're so glad to get you folks together and to be able to talk about the DBI in this way.

Lee: I’ve been looking forward to this!

Anne: Thank you both, Neil and Bradley. We really appreciate you doing this and also bringing us together.

Works Cited

Branch, John, et. al. Snow Fall. The New York Times. 20 December 2012. https://www.nytimes.com/projects/2012/snow-fall/.

Herrington, Anne, and Marcia Curtis. Persons in Process: Four Stories of Writing and Personal Development in College. NCTE, 2000.

Odell, Lee, Dixie Goswami, and Anne Herrington. The Discourse-Based Interview: A Procedure for Exploring the Tacit Knowledge of Writers in Nonacademic Settings. Research on Writing: Principles and Methods, edited by Peter Mosenthal, Lynne Tamor, and Sean A. Walmsley, Longman, 1983, pp. 220-235.

Packer, George. Dispatches: Underlying Conditions. The Atlantic, vol. 325, no. 5, pp. 9-12.

Royster, Jacqueline. Traces of a Stream: Literacy and Social Change Among African American Women. U of Pittsburgh P, 2000.

Return to Composition Forum 49 table of contents.