I FELT painfully sympathetic when I read Valentine Cunningham's account of the ceaseless round of reference writing (THES, January 30). Not because I have written many but because I have asked for so many. In three years of trying to obtain an academic job I have had to plague the same referees ceaselessly.

Each time you ask for yet another reference from your incredibly busy referees you feel a pang of guilt. By the very act of asking you are confessing that all your previous attempts were failures. You hope desperately that they still believe in you enough to dispatch a recommendation, yet you fear that they are really fed up and wish that you would give up.

References are a nightmare whether you are asking for them, writing them or (I am assured) reading them. So can we not largely dispense with them? As a method of sifting talent from mediocrity they have obvious drawbacks. They easily merge into old-style patronage: it becomes a question not of how good the candidate is but of who one wishes to please, whom one is fearful of offending.

Moreover, some referees simply have the art of writing in a more persuasive way than others. One referee's "brilliant" will be another referee's "above average". The reader of a pile of references is trying to glimpse an uncertain entity (the merit of a candidate) through an inexact medium, but probably will not have the time for fine analysis, for "reading between the lines".

ÖйúAƬ

Millions of pointless references must have been sent out, at the cost of God knows how many working hours and stamps, that were hardly looked at, perhaps not even opened. One glance at a CV tells an employer whether a candidate is suitable or not. And this is how it should be. Tangible achievements are objectively verifiable in a way that no amount of gush about "great promise" or "marvellous potential" can be.

What particularly galls me is the demand for references, not at the final stage of an application process, when they can act as an endorsement of character, but as an aid to shortlisting for a job.

ÖйúAƬ

The many research fellowships advertised each year are an excellent example of referencing gone mad. For each one there are 300-400 applicants, each of whom is required to contact two or three referees in advance. That means 600-1,200 references are written and posted before any shortlisting takes place. What useful end is attained by this?

It would be far better to concentrate on the information contained in CVs to draw up a shortlist. Referees' labours would be reduced vastly. The academic world is not particularly large and Oxbridge, where the majority of research fellowships are, is smaller still. The chances that the reader of a reference knows the writer of it are pretty high. This may help or hinder the candidate (particularly in fields such as mine, English, in which methodology is aggressively contested), but only by introducing factors into the equation that have nothing to do with the candidate's suitability for the post in question. In every way, up-front references are likely to complicate an otherwise straightforward relationship between candidate and assessor. Can we not let pretty much everyone off the reference hook?

David Chandler, DPhil in English in 1997.