Ivory Towers: Sinister myth is exploded

HANDS up anyone named Kerr or Carr. Now, did you raise your left or right hand? The answer could be significant, for a recent piece of research has suggested that people named Kerr or Carr may contain much the same proportion of left-handers as the rest of the population.

In 'The handedness of Kerrs and Carrs' (British Journal of Psychology, vol 84, December 1993), D Shaw and I C McManus challenge an opinion held at least since 1971 and possibly since the 14th century. The historical position - that Kerrs and Carrs are more left-handed than they ought to be - is supported by three pieces of evidence. Firstly, there is a traditional, anonymous, Scottish poem:

But the Kerrs were aye the deadliest foes

That e'er to Englishmen were known.

For they were all bred left-handed men

And 'fence against them there was none.

Secondly, there is the etymology of the name, Kerr (anglicised as Carr) comes from the Gaelic caerr, meaning 'awkward', and leading to the expression 'Ker-handed' to mean sinistromanual. Thirdly, there are Scottish castles built by Kerrs which have left-handed spiral staircases.

This came to the attention of the Royal College of General Practitioners, whose research unit in 1974 published a report which claimed Kerrs and Carrs were almost three times as likely as anyone to be left- handed. The new research casts doubts on their methodology. The Royal College had asked interested GPs to choose Kerrs or Carrs from their patients, to identify their handedness, and compare it with families selected at random. The research unearthed 200 Kerr/Carrs of whom 29.5 per cent were left-handed. The control group produced 11 per cent.

The trouble came with the media coverage of the research, which was so widespread that many Kerrs and Carrs offered data without going through their GPs. As Shaw and McManus say: 'Essentially the problem was that persons reporting data, be they GPs or members of the public, were more likely to report families that were consistent with the previously stated and publicised hypothesis.' So more left-handed Kerr/Carrs stood up to be counted, leading to bias in the findings.

The recent research proceeded more carefully. They picked, at random, from the London telephone directory, 25 Kerrs and 25 Carrs. They also selected a control group of 50 people with other English-sounding surnames. They then asked, without telling anyone about the surname connection, about the handedness of family members.

Results were obtained for 50 Kerr/Carr and 48 non-Kerr/Carr families ('data from two control subjects could not be used since they were inadvertantly mislaid.') comprising 1401 family members of whom 10.9 per cent were left-handed. Males (15.2 per cent) were significantly more likely to be left-handed than females (6.4 per cent).

Of 706 members of Kerr/Carr families, only 65 (9.2 per cent) were left-handed, compared with 12.7 per cent of the control group. Indeed, in this research the Kerr/Carrs were significantly less left-handed than the rest.

'We must therefore conclude that there is no convincing evidence, despite a wealth of intriguing folkloric and etymological evidence, that modern-day Kerrs and Carrs have an increased likeliness of sinistrality.'