The reference that said nothing

A reference letter that says a candidate was punctual, pleasant and reliable is usually a refusal to recommend them, and everybody involved knows it.

The convention exists because of liability. In many jurisdictions a negative reference can be challenged, so employers have converged on a practice of confirming dates and job titles and nothing else. What filled the gap was a code. A letter that praises only ordinary virtues, or that is short, or that notably fails to mention the central skill of the job, is read as a warning. The code works, after a fashion, and it has two bad properties: it is invisible to anyone who has not been taught it, and it cannot be challenged, because nothing negative was said.

The people most disadvantaged are those whose referees are outside the professional culture doing the hiring. A teacher writing a first reference for a student, or a small employer in another country, may produce a warm, detailed, enthusiastic letter in the wrong register and have it read as faint praise, or produce a bare factual one and have it read as damning.

Some organisations have responded by abandoning free-text references entirely in favour of structured ones: four questions with defined scales and a space for a single example. The letters are duller and the information is comparable, which is the point. A structured reference cannot be read for tone, because there is no tone.

For anyone asking for a reference, the practical move is to give the referee the questions. A referee who is told what the reader needs to know will answer it, and most unhelpful references are not hostile. They are written by someone guessing.