HERE'S a word that's lost the nice sharp edge it had once. Chris Smith, our Culture Secretary, used it last week after his valedictory chat with the outgoing lottery regulator. "I'm very grateful," said Mr Smith, "for the significant personal contribution he has made to making the UK lottery the success it is." You can see why he chose "significant" to describe the regulator's contribution. "Useful" might have sounded too patronising, "big" might have overstated the case. "Significant" was just right: it suggested that Peter Davis's presence had made a difference, without putting too precise a value on it.

So it didn't mean much, which is sad when you consider that the first meaning of "significant", from the Latin, was "full of meaning". A speech might be called significant if it avoided platitudes and conveyed a message. Perhaps it was inevitable that after a time "significant" should have come to be used instead of "important", which was already happening in the 19th century. But I think it was when the Bloomsbury group took up the word that it really began to be a lost cause. "Significant form", said the Bloomsberries, was that hard-to-define quality in a picture or sculpture that stirred the viewer's emotions irrespective of the object represented. This was sensible enough, but it encouraged much talk of "significance" among pseudy critics who thought it would hide their ignorance. The word got blunter and blunter.

Only among mathematicians does it still have an exact meaning. A significant difference between two sets of figures - the difference, say, between the number of illiterate pupils in one sort of school and in another - is one that can't be dismissed as due to possible mistakes in sampling. Some journalists who write of "a significant minority" may be talking the language of statistics without knowing it. Not that dictionaries always help. The Shorter Oxford defines statistically significant as "having a low probability of occurrence if the null hypothesis is true," which to many signifies nothing.

Comments