Are plans for quality assurance heralded in a new report on the right lines? Three readers think not. The structure of my teaching is overseen by a hierarchy of at least three different committees. We have a teaching committee in the department, there is another one in the university, and then there is the teaching quality assessment panel which comes round once every four or five years.

All these committees were set up for good reasons, educational and political. It is widely accepted that we should follow "best practice", and that students rightly expect us to be explicit about what we are aiming to "deliver" and to have a well-understood system for achieving these aims. We also understood that, politically, we had no choice.

Clearly, this is a belt-and-braces strategy. And when we take into account "appraisal", "accreditation", the "mentor" system, and all the arrangements for finding out what the students think of our courses, plus all the mutual checking up on each other involved in doublemarking, we seem to be developing a safety-pin and suspender-belt system as well. But why stop here? Why not have video cameras in classrooms, like the one trained on assistants in our local fish and chip shop?

Has all this improved the quality of our teaching? It has certainly improved the quality and quantity of the paperwork. It has taken a lot of time and trees. But I have little faith in their ability to pick up an improvement or deterioration in my teaching. Mostly, I know when I am teaching well, or adequately, or badly, and I suspect most of my colleagues do too. When I succeed, it is not because the quality control mechanisms have helped me or scared me; when I fail, these mechanisms will not pick it up, unless the failure is of epic proportions.

³ÉÈËVRÊÓƵ

We all know that it would be political suicide to be seen to reject "accountability". But surely we should not have to accept an inefficient system of accountability? This system has not been planned overall; the quality control measures have been introduced piecemeal and vary from place to place.

The range of measures would do credit to an aircraft manufacturer after several recent plane crashes. But there is nothing remotely equivalent to such disasters in university teaching to provoke such a response. Was there a comprehensive damning report about the quality of our work?

³ÉÈËVRÊÓƵ

Should we not be embarking on a cost-benefit analysis of all these "quality control" measures, taking into account the benefits of alternative uses for the time involved?

I write with feeling, having just spent several hours double-marking work that a colleague has already marked meticulously. Let us stop now, and work out a least-cost system of quality control, based on trust and common sense.

Andrew Fleming Reader in archaeology at the University of Wales, Lampeter.