At 25 Dr Soap naturally preferred to set exercises from the stock of ones he had a crib for. By 30 he had his eye in, and began to think of setting (at least in examinations) exercises which had never been set before.

It wasn't easy finding the right level: devising fairly easy problems was an unfairly hard problem; the one with greater knowledge was having to sweat harder than those with less.

In some of his first attempts his students found too little challenge; in others, too much; but since neither was their fault, in neither case did they deserve less than their usual B-. Most eventually learned how to produce a sufficiently correct answer even when they didn't understand the question -- a state of affairs Dr Soap tried all his life to interpret as gratifying.

By mid-career, with increasing maturity, he had discovered that a good exercise might be one that he couldn't actually do himself until after he had seen the best students' scripts. At the same time he was able to find acceptable an increasing proportion of his unaided answers to his own questions -- a clear sign of increasing skill as a setter, or maturer benevolence as a marker.

Latterly, indeed, it had become quite normal to set an exercise he could do to his own satisfaction, and could have done even without working backward from the answer in the books. Where others would have been content, he raised his sights to a yet higher target, and was struggling to set exercises he could have done at their age when he flunked the final Test. He will be greatly missed by fellow-members of the Committee on Innovation in Teaching.

Objectives

If you must talk about educational objectives, do at least have the decency to lower you voice, as in the presence of a mystery.

`It should be possible to design examinations which foster the kind of learning we want to encourage. To do this, we have to be clear ourselves about our objectives'. This feels like another instance of the south-pointing compass: Objective and Exam are no more than names for the two ends of the same needle: my exam is good enough because most people pass it, thereby attaining the course objectives. This ought to be called De Morgan's law, after his `Cambridge examination' of 1853:

"Q. What is knowledge?

A. A thing to be examined in.

Q. What is the instrument of knowledge?

A. A good grinding tutor.

Q. What is the end of knowledge?

A. A place in the civil service, the army, the navy, &c.

Q. What must those do who would show knowledge?

A. Get up subjects and write them out.

Q. What is getting up a subject?

A. Learning to write it out.

Q. What is writing out a subject?

A. Showing you have got it up."

The circularity is inevitable as long as we do not recognise the ought/is ambivalence of the word Objective -- meaning either `Choose a destination and tell someone else to find the best route there' or `Choose the best destination reachable via available routes'. To judge by most specimens, either the gap between objectives and available resources is quite wide -- and no attempt is made to bridge it; or the gap is all too narrow, and the objectives are visibly derived from our favourite activity.

This is not a chivalrous encounter between Sir White-Objectives and Sir Black-Objectives, but a more interesting brawl between various more or less legitimately vested interests. Each of us has such an interest in two or three favourite kinds of exercise or exam whose invention was not a matter of simple rationality. But we can all work back from intuitive acceptance of these discoveries to the kind of plausible principle that looks good on a shield. We are conscious nevertheless that such rationalisation is never quite satisfactory or complete, and that the contest becomes unreal unless each contender works to keep open the channel, in both directions, between rational principles and intuitive practice.

W. Bull: "To achieve near-native ability [the learner] has to know more about the language than linguists have either discovered or described. As a result, the students in our present language programs are faced with two kinds of deficit of knowledge: the one that develops because they fail to discover what is being taught, and a built-in deficit which linguists cannot reduce. Nor is it possible to specify in any scientific fashion the objectives of a language course. There is just no way to differentiate between what is being taught to the student and what is learned by the student."

Emerson: "We do what we must, and call it by the best names we can, and would fain have the praise of having intended the result which ensues."

Oracles

An oracle gives you a message which, with very little work, translates into what you really wanted all along. This gives you the confidence to act decisively. Later it turns out that you were being given a painful lesson on the wishfulness of translation, and another no less painful on the necessity of a decisiveness which doesnot require to be based on over-confidence.

Ordinary language

Coleridge:

" ... a selection of the REAL language of men... (Wordsworth)

I object, in the very first instance, to an equivocation in the use of the word `real'. Every man's language varies, according to the extent of his knowledge, the activity of his faculties, and the depth or quickness of his feelings. Every man's language has, first, its individualities; secondly, the common properties of the class to which he belongs; and thirdly, words and phrases of universal use. The language of Hooker, Bacon, Bishop Taylor, and Burke differs from the common language of the learned class only by the superior number and novelty of the thoughts and relations which they had to convey. The language of Algernon Sidney differs not at all from that, which every well-educated gentleman would wish to write, and (with due allowance for the undeliberateness, and less connected train, of thinking natural and proper to conversation) such as he would wish to talk. Neither one nor the other differ half so much from the general language of cultivated society, as the language of Mr. Wordsworth's homeliest composition differs from that of a common peasant.

For `real' therefore, we must substitute ordinary, or lingua communis. And this, we have proved, is no more to be found in the phraseology of low and rustic life than in that of any other class. Omit the peculiarities of each, and the result of course must be common to all. And assuredly the omissions and changes to be made in the language of rustics, before it could be transferred to any species of poem, except the drama or other professed imitation, are at least as numerous and weighty, as would be required in adapting to the same purpose the ordinary language of tradesmen and manufacturers. Not to mention, that the language so highly extolled by Mr. Wordsworth varies in every county, nay in every village, according to the accidental character of the clergyman, the existence or non-existence of schools; or even, perhaps, as the exciseman, publican, or barber, happen to be, or not to be, zealous politicians, and readers of the weekly newspaper pro bono publico. Anterior to cultivation, the lingua communisof every country, as Dante has well observed, exists every where in parts, and no where as a whole."

Outwitting the censor

abbé Galiani : "Savez-vous ma définition du sublime oratoire? C'est l'art de tout dire, sans ëtre mis à la Bastille, dans un pays ou il est défendu de rien dire."