Last changed 19 Dec 1998 ............... Length about 1,000 words (11,000 bytes).
This is a WWW document maintained by Steve Draper, installed at You may copy it. How to refer to it.

Web site logical path: [] [~steve] [this page]

Should teachers be expert in the subject matter?

Contents (click to jump to a section)

A Technical Memo
Stephen W. Draper
Department of Psychology
University of Glasgow
Glasgow G12 8QQ U.K.


Triggered by a discussion on the ITFORUM email list around 1 Dec 1998.


Must or should teachers be expert in the subject matter being taught and learned? Many feel indignant if they do not. But an alternative view is that they should (only) be expert in teaching. A still more radical view is that learning is helped by peer interaction, and no special role for the teacher is needed. The correct synthesis however may be that to be a good teacher -- or rather a teacher good for the greatest number of different learners -- requires expertise in BOTH a) teaching; and b) teaching that particular subject matter, including both b1) the subject knowledge itself, and b2) experience in learning and teaching it.

In the rest of this, I expand the arguments in favour of each of these conflicting points.

Subject matter knowledge

In my most recent written feedback from students, one (but only one, I hasten to add) said I didn't seem to know what I was talking about, and clearly meant this as a bad thing. Similarly one of the discussants in the Itforum exchange adopted this viewpoint. Objections to it are mentioned in the next section, but it does seem to strike a chord in most of us. Would you really seek out someone ignorant to learn from? Socrates has had a great press, but I don't recall the slave boy actually asking him for information (or knowledge or wisdom). The Socratic dialogue was forced on the slave, and what is more Socrates argued the slave didn't learn anything new. The desire to seek information and instruction from experts is greater than ever before: we want to use the best consultants if we can get them.

And in fact the image of a consultant is a good one for us to consider for a teacher: they may possibly not know everything themselves but we certainly want them to act as our gateway to all the knowledge we need on their specialist topic. Note however that expertise does not mean knowing everything so much as knowing how to get the information. I made this point some years ago about knowledge of computer systems [Draper (1985)]. Similarly it is true of librarians: they know only a tiny fraction of the books in their library, but they are better than you at finding out about them. It could be true of good teachers: knowing relatively little, but knowing how to get the information, and access the knowledge. That is what "expert" really means.

I believe Piaget somewhere made the point that a learner must depend on a teacher for the knowledge of what there is to learn: you can't decide if it's a good idea to learn about object conservation or Laurillard's theory until AFTER you have learned a lot about the area. That is the kind of knowledge a teacher really must have, and a learner really and necessarily depends on the teacher having. That is the sense of expertise in which a teacher can't be useful without being a subject matter expert.

And perhaps William Perry represents another aspect of this: it may be that no knowledge of an area can be considered mature if it simply encompasses a single received view, but rather must cover alternative views, their relationships and relative strengths and weaknesses; yet conversely, it is expertise to know which views are interesting at all ("significant") in each area.

Teaching expertise

But equally there are convincing arguments against a teacher needing to be an expert in the subject matter. For a start, demanding it seems to be adopting an instructivist view of learning, in which learning is (only) the transmission of information from expert to learner. If learners demand that teachers be knowledgeable, that just shows how ignorant they are about learning as well as everything else, doesn't it?

The argument above about the nature of expertise may resolve that simple objection, but there are others. An opposite view (from demanding that teachers be subject matter experts) is that teachers need (only) be expert at teaching. A vivid example supporting this for me is my experience of teaching elementary computer programming. As learners struggle with their first programs in language X they frequently ask for help, and benefit crucially from support by tutors. But the tutors do not need to know X so well they can just tell the learner the answer every time. It is enough, and almost certainly actually better, that they just have the confidence that they can discover first the root of the problem and then the solution. Learning that every problem in computing is soluble, and in fact soluble by any programmer, is the key lesson, and this can only really be demonstrated by tutors who have not yet learned by heart every detail of the language and examples being used by the class.

Joe Beckman was taking this kind of position in the itforum discussion: that methods of interaction were what a teacher really must know, not the answers. My own version of this would be my experience of doing first year tutorials in psychology, in which I have no formal training nor systematic knowledge. This scared me because I share the atavistic fear of many teachers of being caught without the answers, but actually it was fine. I spent most of the time structuring discussions aimed at linking the topic of the week to the students' own personal experiences. This requires no prior knowledge of the subject matter, yet is important for student learning. Of course this only half supports this point of view: it was good teaching because it served a learner need too often neglected and not addressed elsewhere in the course, but it relied on leaving the exposition (based on subject knowledge) to other sources (textbook and lectures), so it didn't represent all or even most of the teaching provision.

Interaction only; teachers unnecessary; peer interaction can do it

A more extreme version of this is that all that matters is the right activities and interactions for learners. We might call this learning expertise. Given these, then peer interaction will do, and in fact may be best since other learners will more reliably understand the learner's viewpoint. Thus teachers are unnecessary once learners have acquired the necessary general learning skills.


But perhaps both are essential: both subject matter expertise, and teaching expertise. This is the view of Al Schoenfeld, quoted in Vobejda (1987) "You can have a well-intentioned curriculum, developed by talented mathematicians who don't understand psychology, and it can fall flat on its face. ... And if you have a psychologist who develops a curriculum... that emphasizes rules and procedures, rather than underlying mathematics ideas, the result can be a well-designed disaster." The arguments above perhaps really just show the importance of each of these separately, but not that either can in reality flourish alone (even if the practitioner may be unaware of or undervalue their exercise of the other component).

But in fact even this pair of expertises may not be enough. In a talk I attended, Charles Duncan in describing the impressive EUROMET project (which is authoring extensive professional CBT for meteorologists) said he didn't want to hire subject matter experts (which would just lead to the problem indicated by Shoenfeld) -- what he needed was teaching matter experts: those with experience of teaching the topics and who therefore knew what the main problems were that learners had and had some experience in overcoming them. This of course is entirely consistent with the emphasis by phenomenographers, Laurillard and others on researching learners' experiences and misconceptions and only then designing materials that accommodate that data.

In fact I believe (can anyone tell me for sure?) that the French have a special role and word -- "didactique" -- for the activity needed, which is NOT covered by either subject matter knowledge nor general teaching expertise. Lee Shulman (1986?) and/or Pam Grossman (1989) used the term "pedagogical content knowledge".


So I vote for teachers requiring both subject matter and teaching-method expertise; and in fact requiring all of:

So what is a teacher's function?

QED (and what do YOU think?).


Draper,S.W. (1985) "The nature of expertise in UNIX" in B.Shackel (ed.) Human-computer interaction -- INTERACT '84 (North-Holland: Amsterdam) pp.465-471.]

Grossman, P. (Sept- Oct 1989). "A study in contrast: Sources of pedagogical content knowledge for secondary English" Journal of Teacher Education pp.24-31

Perry, W.G. (1968/70) Forms of intellectual and ethical development in the college years (New York: Holt, Rhinehart and Winston)

Shulman, L. S. (1986) "Those who understand: Knowledge growth in teaching" Educational Researcher vol.15 no.2 pp.4-14

Vobejda,B. (1987) Educational researcher vol.16 no.9 pp.9-12 "A mathematician's research on math instruction" [on Alan Schoenfeld]

Web site logical path: [] [~steve] [this page]
[Top of this page]