Leading Article: Doubtful evidence of the handwritten word

IT IS SAD to hear psychologists pooh-poohing so authoritatively the idea that you can tell something about people from their handwriting. Amateur graphology, like armchair detective work, is an amusing way to pass the time.

Unfortunately, the psychologists have a point. Handwriting, like fingerprints, may be hard to hide: those who try to make their writing look like someone else's rarely succeed in doing so, which is why blackmailers so often resort to cutting headlines from newspapers. Yet identifying people from their writing is not the same as exposing their innermost secrets.

All too often, writing can be influenced by factors far removed from the hand that holds the pen. Children at school in continental Europe learn a more uniform style than do those in Britain, and American handwriting is more consistent still. English written by a Japanese, precise and regular in both weight and line, may at first seem to reveal a lot. In fact, it is merely a symptom of the fact that oriental characters require a great deal more manual dexterity than do Roman characters, so someone who has learnt the first is unlikely to have trouble with the second.

Writing is also influenced by fads. Young people whose hands are not yet fully formed experiment with different styles - the current fashion is for fatter Os and As, and stunted ascenders and descenders - just as designers of newspapers experiment with wider space between columns or thinner rules.

But the British Psychological Society cannot yet declare victory over the International Graphoanalysis Society. It may not be scientific; there may be little correlation between the specific conclusions drawn by professional graphologists and those drawn by recruitment specialists. But handwriting certainly reveals something about people - just as do their clothes, their facial mannerisms, their way of sitting, their choice of words and accent, and the amount of polish on their shoes.

None of these, however, is a satisfactory basis on which to decide whom to employ. For that purpose, the two-day 'assessment centre' used by the Civil Service, dull as it may sound, gives a better idea of how people will perform at work. But the best recruitment test must be the simplest: have all shortlisted candidates in the office for a week or two, and see how they get on.