The Language Teacher

Interview with Rod Ellis (Revised for Publication in TLT) by David Kluge

Return to The Language Teacher Online

TLT: Thank you for agreeing to do this interview with the The Language Teacher on the topic of teacher-researchers and action research. In the area of EFL, what has historically been the relationship between the researcher and the teacher?

Ellis: I think the traditional relationship between researcher and teacher in EFL has been much the same as the traditional relationship in any area involving education, which is essentially kind of a positivist relationship where the researcher does research which demonstrates what is the most effective way to teach and this information is then passed on to teachers who adapt their teaching accordingly.

TLT: By positivist you mean following the scientific method?

Ellis: Positivist in the sense that there are definite answers to problems. These answers can be found by using the appropriate research methodology, and when they have been found, the information is conveyed to practioners who then act on them.

TLT: What do you think the relationship is now?

Ellis: I think that the situation now is somewhat more complicated with rather differing attitudes, explicit and implicit, about the relationship between research and teaching, and researchers and teachers. I think that there is still something of the positivist relationship that I've just described. This is often implicit in the kinds of implications for teaching or applications that are written at the end of a research article, and I think it's also implicit in researchers' views about how research should be conducted and what its value to teaching is.

However, I think there are very different research paradigms that operate now with very different views about the relationship between research and teaching. Naturalistic research which seeks not so much to demonstrate what is the most effective solution to a problem, but to try to identify precisely what the problem is, to understand the problem better. That is, research that aims at understanding rather than explaining, seems to offer a rather different relationship between research and teaching. It seems to suggest to teachers that the more that they can understand what teaching a second language involves, the better they will be equipped to actually engage in teaching. Certainly, naturalistic research is not directed at telling teachers how they should teach. And of course action research which seeks to actually make teachers researchers themselves in order to investigate their own classrooms, offers a very different relationship because teacher and researcher become the same person.

TLT: So, what do you think the relationship should be?

Ellis: Well, my first feeling, which I think is probably not shared by everybody, is that research will probably never produce definitive answers to teachers' problems. I think research should be seen as a way of exploring what teaching consists of, illuminating teaching, and perhaps identifying provisional answers to problems, which teachers can then decide whether they wish to actually try out in their own classrooms or not. In other words, I think that the relationship between research and teaching needs to be predicated on the assumption that it is the teacher who have the final answer as to what is done in his or her classroom, and that research can do no more than serve as one of the sources of information that teachers can draw on in order to decide what goes on in their own classroom.

TLT: What are some of those other sources?

Ellis: Other sources they can draw on most obviously are their own actual previous experience of teaching, their previous experience of what works and what doesn't work. Other sources are seminars, conferences, chats with other teachers, e-mail correspondences, etc. All of this information potentially should inform teaching, not just research.

TLT: So it's just one of the tools to improve the language teaching field.

Ellis: Yeah. I think one of the problems is that research tends to have higher status than other sources of information. Clark wrote an article in TESOL Quarterly a few years ago in which he talked about the dysfunction between research and teaching and between researchers and teachers and advised teachers to keep their own counsel. I suppose I wouldn't want to go so far; I think that would be too blinkered. I think teachers should keep their own counsel, but this doesn't preclude them familiarizing themselves with what researchers have found out and asking whether what they have found out has any particular relevance to their own work.

TLT: It should be informed counsel then?

Ellis: Yeah. I certainly think research and researchers have the potential to help teachers decide what they want to do. They provide a source of information and a type of information that perhaps is not available for teachers in the context of their day-to-day work.

TLT: What about the idea of the teacher as researcher?

Ellis: The fact that there will be researchers and there will be teachers does not preclude teachers doing research. In fact, the possible ways of teaching suggested by research need to be investigated by teachers in the context of their own classrooms. Now the investigation can take place informally through normal teaching, or it can take place more systematically. If it takes place systematically, then it calls for some kind of teacher research, so action research, or teacher research, is one way of establishing a meaningful link between the work of researchers and the work of teachers, between research and teaching.

TLT: What do you think are the issues that need to be addressed in the researcher-teacher discussion?

Ellis: Well, obviously one key issue has to do with the status of research findings and what use teachers should make of them. This is what we've already been talking about, and I'd argue that research of any kind, be it experimental or naturalistic, can afford nothing more than what Stenhouse (1975) calls "provisional specifications," which teachers then need to investigate and try out in their own classroom.

TLT: So you're saying that the results of research are not things that have been proven, but provisional statements that need to be continually checked, and re-examined.

Ellis: That's it, and re-examined in particular contexts. Because even research such as experimental research that seeks to achieve generalizability cannot really be generalized to all teaching contexts.

TLT: So in effect then, the concepts of reliability and validity are not as clear-cut as many of the statisticians would like us to believe.

Ellis: No, because there's no such thing as reliability for all populations. A test that might be reliable for one population is not necessarily reliable for another because the second population may have different characteristics from the first one. Therefore any test or any teaching proposal cannot be applicable to all populations. So, teachers need to investigate whether a particular proposal, a particular test is applicable to their particular population, their own classroom.

TLT: What other issues do you think are involved in this whole researcher-teacher discussion?

Ellis: Well, there's obviously the issue regarding what type of research might be most informative to teachers. This is an ongoing argument, and I think perhaps one of the developments in our own field that I am most dubious of has been the increased reliance on experimental research, as opposed to descriptive or naturalistic research. The problem with a lot of experimental research is that it purports to describe cause-effect relationships which are generalizable, and therefore it can more easily be misused than the kind of descriptive or naturalistic research that was done somewhat earlier, a few decades ago. Of course, we still do get descriptive, naturalistic research. I don't want to suggest that it is all experimental; it's not. But if you look at the journals, there is clearly a tendency for experimental research to dominate. And experimental research is problematic.

For example, if one wants to measure learning as a result of some treatment, then how do you measure the learning? Frequently, the studies that are published don't really discuss what they mean by learning, and don't discuss to what extent the particular ways they've chosen to measure learning are valid. The ways that they've chosen, in any case, may be valid for their particular research purpose, but may bear no relationship to teachers' understanding of what learning is. A good example of this would be, for example, grammaticality judgments, which researchers use to measure learning that results from some treatment. But most teachers would not think that what they are in the business of doing is developing whatever is required to enable students to decide whether a sentence is grammatical or not. Their concern is with developing communicative ability to actually use language.

TLT: So you are saying that perhaps many teachers could look at some of the research that is done and say, "That doesn't have anything to do with what I'm really interested in, which is long-term educational gains?"

Ellis: Yeah. I think that there is sometimes a mismatch between how researchers operationalize constructs like learning, and the kind of construct that teachers work with when they talk about learning in their own work. And this is obviously going to make substantial problems in trying to apply the results of that research to teaching. The danger is that researchers will ignore this problem and put forward suggestions about how to teach and that teachers who fail to notice the mismatch will be taken in by them.

TLT: Dick Allwright, in a TESOL Quarterly article (1997) wrote about the issues of quality of research (reliability and validity) and sustainability, that is the often usual "abandonment of research after completing an arduous, troublesome or unsatisfying research project." Allwright, in what he calls "exploratory practice," recommends giving priority to sustainability. Nunan, also in TESOL Quarterly (1997), indicates that standards should be given top priority in teacher research. Which of these two areas do you think should be given priority in research that teachers are doing in the field now?

Ellis: I think that teacher research really has two purposes. One of its purposes it shares with all research, which is to seek solutions to problems or to test out ideas. For that particular purpose, I think Nunan is right; that if we want to test out ideas or if we want to seek solutions to problems, then we want to conduct research that will give us clear and reliable answers to the questions that we ask. So it is necessary to meet standards of reliability, validity, etc. But there is another purpose to research that teachers do to reflect and understand their own classroom, their own teaching, and their own students better. Here one might suggest that it is less important that the conventional criteria of reliability and validity in research are met, and what is more important is that teachers conduct research that is illuminative; illuminative not to others, but to themselves. And in this case, it seems to me that Allwright's arguments about sustainability are very strong. Because, if we require teachers to be technically sound in the research they do, many of them may in fact be frightened off; they may find that it requires more effort than they are able to provide, more time than they are able to give. And in such cases, I think that it's much more important that teachers should be encouraged to persist in their research efforts rather than to give up because they can't do good research.

TLT: If your purpose is to improve your teaching and to reflect on your teaching methods by doing something like action research, yet it doesn't have strong reliability or validity, do you think that it is publishable?

Ellis: Probably not. I mean, Stenhouse argued that there are two characteristics of research, of any research, including teacher research. One was that it was systematic, and the other was that it was published. And probably the sort of research that I think Allwright is talking about, is not going to be publishable. But this doesn't mean that the research cannot be communicated to others, for example through conversations between teachers or, perhaps, in a JALT chapter meeting. I don't think one has to do good research in order to share that research with teachers for and for other teachers to find that research of interest and of value to them.

TLT: By good research you mean...

Ellis: I mean research that is going to meet the standard criteria of validity and reliability. I think it's quite possible that teachers can do research that is not going to meet those criteria but can address issues that can lead to debate and discussion and as such can contribute to the process by which teaching develops.

TLT: Do you also consider that to be good research?

Ellis: I consider it to be useful research.

TLT: Are there any special problems or special benefits in Japan for the teacher-researcher?

Ellis: Probably not. Probably the reasons for the benefits of, and the problems of doing research are going to be similar from one situation to another situation. The problems of doing research have to do with the technical skills of designing a research proposal and then carrying out the research and reporting their results. Those skills are going to be common to teacher-researchers whether they are working in Africa, in Asia, in America, or in Europe. It doesn't really make much difference. The benefits, I think with regard to Japan are going to be very similar. I mean one of the reasons why people do research, certainly why they try to do formal research is because they feel it is going to contribute to their vita and therefore is going to make them more marketable, and I think this is a legitimate motivation. It's probably one of the reasons why I got involved in research in the first place. I don't really see that the problems or the benefits are going to be substantially different in Japan.

TLT: You said that one of the reasons why you first got into doing research was to add to your your resume...why do you do research now?

Ellis: That's an interesting question. I do research now for a variety of reasons; one is sheer intellectual curiosity. I got intellectually involved, for example, in the ongoing debate about what is the best way to teach grammar. And one way that I can contribute to those debates is by actually doing research to investigate different ways of teaching grammar. A second reason is undoubtedly still professional in the sense that doing research is going to maintain my position as a researcher, as an applied linguist, is going to give me material that I can present at conferences for which I am rewarded by the institution for which I work, so, you know, there is that professional reason for doing research. And, I think I do have, and always have had the third reason, and that is that I think that research is one way that one can explore, think about, illuminate, and perhaps improve pedagogy.

TLT: For teachers who have been doing teaching only, and are getting interested in research, what would you recommend them to do?

Ellis: I think in order to do research one has to actually find out how to do research. There's not much sense in saying, "I'm going to research my own classroom" if you don't really know what it means to do some research, so you do need some sort of technical know-how. That's not to say that you need to become an expert researcher, but you need some technical know-how. You need to know how to formulate a research question that is researchable. Then you need to know how you can operationalize a research question, how you can operationalize the elements of your research question, and how you can collect data that will enable you to address that research question. You are going to have to familiarize yourself to some extent with how research is done. So I think probably the first thing that I would recommend to teachers is that either they attend a seminar that introduces them to how to do research,or alternatively you have to get an introductory book to research, such as J. D. Brown's (1988) or Nunan's (1992) book about how to do research--although both of those books are more geared to students doing masters level courses. In fact, I don't think there's a really good introductory book for the language teacher about how they might set about doing research in their own classrooms that would be sustainable in Allwright's sense.There is a need for such a book, but don't ask me to write it because...

TLT: (laughing) I was just going to ask you to do that!

Ellis: I think that those are the two principal ways. The only other thing that I would say is that I don't think teachers should sort of sit around and wait until they feel they know what research is before they try and do a bit, because, you know, the main way in which one becomes a reasonable researcher is by actually doing research. You know, research is a skill, like learning to swim, in a sense, or learning to ride a bicycle. You can read a lot of manuals about how to ride a bicycle or how to swim but it won't actually enable you to swim or ride a bicycle. You have to actually engage in those activities. So, I think teachers, at the same time as reading about research or attending seminars about research, should try to engage in research in their own classrooms.

TLT: They should jump into the water and start splashing.

Ellis: They should jump in--but one word of advice--they should try to jump in by identifying a very small, little issue that they want to investigate. The problem with researchers who are just beginning is that they want to solve the problems of the world, and they're not going to. So they need to identify very, very small issues, for example, a simple descriptive question, like "How do I deal with my students' errors?" because one of the things that research shows you is that often what you actually do in the classroom, like dealing with students' errors, is not what you necessarily think you do. There's often a mismatch. Teachers need to start with a very focused issue, a very focused question.

TLT: What areas do you think would be good topics to explore here in Japan?

Ellis: There are potentially a huge number of issues. I mean, you talk about classrooms in Japan, but there are a lot of very different types of classrooms. There are college classrooms, there are high school classrooms, there are even elementary school classrooms these days, and I think the issues are somewhat different. Let me answer it from the point of view of, say, high schools in Japan. It seems to me that one question that researchers might look at is the actual uses of language, the Japanese language and the English language in English lessons in classrooms in Japan. The kinds of questions, for example, that come to mind are: "How much is English used?"; "How much is Japanese used?"; "For what functions is English used?"; "For what functions is Japanese used in the classroom?" Earlier studies have shown, for example, that Japanese teachers of English in high schools in Japan tend to use Japanese for most of the lesson. In fact, one study showed that some teachers use it for over 90% of the talking time in the whole lesson. So, I mean, I think that is one issue that could be usefully looked at. Another issue is to what extent Japanese teachers are able to introduce, to innovate a communicative methodology in their classroom. Teaching communicatively requires very considerable innovation in a Japanese high school classroom, and I think an interesting issue is what teachers are able to manage that kind of innovation successfully, and how do they successfully manage it, and which teachers are unable to do so, and why are they unable to do so, etc. These are certainly issues that I would be interested in.

TLT: What resources would you recommend teachers interested in doing classroom research to read?

Ellis: Well, I mentioned two, J. D. Brown's book, and Nunan's recent book, a slightly more recent book, again on how to do research. I would also recommend Allwright and Bailey's book which is a sort of review of the different types of research that have been done. It is also useful because teachers need to have some idea of what's been done before in order to decide, perhaps, what they would like to do.

TLT: Would you like to end the interview with any kind of thoughts or comments on this area?

Ellis: Well, to go back to two issues that I sort of discussed earlier, I do believe that teachers can gain a lot professionally, in their own teaching as teachers and as individuals by actually doing research in their own classrooms. I think that it can involve them in teaching. For teachers who have been teaching for a long time, it's difficult to remain at a high level of motivation and involvement. One way, perhaps that you can keep involved is by undertaking research in your own classroom. So I think that I would like to conclude by suggesting that teachers should not be frightened about research--they shouldn't be frightened about doing bad research. They should be prepared to try to become researchers. They should see research not as something that is separate from teaching, but as something that potentially is part of teaching.

TLT: Thank you for your insights, Dr. Ellis.

References

Allwright, D. (1997). Quality and sustainability in teacher-research. TESOL Quarterly, 37(2), 368-370.

Allwright, D., & Bailey, K. M. (1991). Focus on the language classroom: an introduction to classroom research for language teachers. Cambridge: Cambridge University Press.

Brown, J.D. (1988). Understanding research in second language: a teacher's guide to statistics and research design. Cambridge: Cambridge University Press.

Nunan, D. (1992). Research and methods in language learning. Cambridge: Cambridge University Press.

Nunan, D. (1997). Developing standards for teacher-research in TESOL. TESOL Quarterly, 37(2), 365-367.

Stenhouse, L. (1975). An introduction to curriculum research and developmennt. London: Heinemann.