I am looking at a letter from Sir Alan Langlands, the head of HEFCE, sent to just about every institution in Higher Education, and many more outside it. It is an upbeat letter, telling of the 534 responses to the consultation HEFCE conducted on proposals for the REF. Amongst other things, it says that “there was widespread support in principle for including an element for the explicit assessment of impact within the REF from higher education institutions mission groups, the national academies, research user representatives and other funders of research, and from a clear majority of academic subject associations”.

It also says that if you want to learn more about these responses, you can go to the website. I was curious enough to do this, since in my experience you can throw a lot of bricks around academic circles before hitting anyone who supports the proposed impact element in the REF. Admittedly the Vice-Chancellors and the rest who drafted the responses are parts of the administrative aristocracy, and need have no brief for representing working academics. Even so, their alleged enthusiasm seemed surprising.

The website, however, says something rather different: “There was widespread acceptance or support in principle for the inclusion of impact as a distinct factor in the assessment, subject to the development of a robust method for assessment and an appropriate weighting for this element”. Spot the differences, anyone?

Logic is not, of course, a STEM subject, and probably did not feature very largely in Sir Alan’s ordinary degree in biological sciences. But first let us meditatate on the difference between ‘acceptance or support’ and ‘support’. A trivial difference, surely, beneath the notice of a busy administrator? Well, suppose I believe that my gap-year son is in India. I can properly infer that he is either in India or China. This is called disjunction introduction, but only logic students need to remember that. What I cannot do is then infer that he is in China. Similarly suppose someone accepts something, perhaps because there is a gun to their head and they have to do so: I accept my bank’s awful terms for an overdraft; the Germans accepted the Treaty of Versailles. We can infer that I accept or support my bank’s terms, and that the Germans accepted or supported the Treaty of Versailles. But we cannot go on to infer that I support the bank’s terms, nor that the Germans supported the Treaty of Versailles.

And then there is that little qualifer ‘subject to…’, which introduces what logic students learn to call a conditional. Suppose I accept a proposal, subject to a condition. Can we silently drop the condition—once more a quibble or qualification beneath the notice of a busy administrator? It would be nice if we could. Think of a day at the races: the bookie accepts a bet which means he must pay Sir Alan a large sum, provided Denman wins the Gold Cup. ‘But look’ says Sir Alan, after Denman fails, ‘you said you accepted that you must pay me a large sum’. Indeed, he might continue, fumbling with the logic one more time ‘you said you supported the idea of paying me a large sum’!

My guess —but it is only a guess, because we are not told any detail—is that quite a lof of the 534 eminent bodies, unhappily resigned to HEFCE’s evident intent to distribute large sums of public money on the basis of ‘impact’, said with a sigh that they could accept that, subject to ‘robust’ methods of measurement, knowing privately that there was no chance whatsoever of such things being found. About as much chance, in fact as the likelihood of an accurate and verifiable measurement of the proportion of the decline of public life due to government foolishness, the proportion due to its mendacity, and the proportion due to the logical and linguistic deficiencies of its officials, although it would no doubt be an enjoyable if time-wasting exercise to run ‘pilot schemes’ on that fragrant trio. I am sorry to say that logic can help us to spot the upshot, but not the cause.