Indeed, The Times misinterpreted his views so badly that it actually wrote an editorial which supported what he said (or what we all misquoted him saying) about “the new system of assessment for youngsters in their first few weeks in school”.

In his angry letter to me (accompanied by other angry letters to sundry people), he set out the statement which so many journalists failed to understand. It concerned the relationship between testing and the abilities of five-year-olds and setting, or streaming, them thereafter.

In the hope of avoiding further protracted correspondence with the recently appointed junior minister, I reproduce (complete and verbatim) the extract from his reply to press conference questions which he quoted. He seemed to think it clarified his position.

“I think that there are plenty of arguments - and I indicate my own prejudices before (sic) - in favour of setting in particular subjects at particular times but that is not what we are talking about this morning. The assessments are not, in my opinion, the vehicle to achieve more setting and streaming.”

Let us examine that single statement with reference to the context. First, it is not possible to tell journalists what “we are talking about this morning”. If they think that “early assessment” is related to primary school streaming and setting, they will (quite rightly) ask questions about it.

Second, ministers are not allowed to have their own prejudices. They are paid to enunciate public policy. We, the people, are entitled to assume that their personal opinions will soon be turned into official positions - otherwise ministers would not be so stupid as to express them.

However the most important question concerns the relationship between testing and setting new entrants into primary schools. On that subject, The Times leader was convincing at least in one particular. “Examining ability to read, write and count will help teachers to identify their most able charges and those in most need of more assistance. Setting is the logical consequence. ”

It is the logical conclusion that The Times comes to which Mr Clarke takes greatest exception. He complains that the press misinterpreted his views as “he did not see” baseline assessments as the preliminary to more primary school setting and streaming. His myopia is a matter of legitimate public concern. For, whether he notices it or not, setting and streaming is (as The Times properly concludes) the natural and inevitable result of testing.

In his complaint to me about the way I have interpreted his view, Mr Clarke took particular exception to the suggestion that he is out of touch with the way in which primary schools work. He offers as evidence of his up-to-date information a wife who was the chair of school governors and children who were educated in Hackney.

All I can offer, in response to this compelling, intellectual argument, is simple logic. Mr Clarke - defending testing five-year-olds - asks middle-class parents not to coach their children in preparation for assessment. I say that they will do it, whether he likes it or not.

Two basic truths emerge from this strange argument. The first is that - either by ignorance or malice - the Government is encouraging more and more selection. The Department for Education and Employment chooses to call it by another name, but increasingly sheep are being separated from goats according to ability or aptitude. The second truth concerns the machinery of government rather than the philosophy of education.

Mr Clarke has come, rather late in life, to junior office. He is caught in the hideous dilemma which the contradictions of government education policy create. How does he loyally defend increased selection by a government which protests its opposition to selection? Not, I must advise him, by blustering to every editor in London.