The question set is (whether deliberately or unavoidably) too wide or moot to be `covered' in the time. For your own sake you must decide -- i. e. discover as you go along -- which aspect you will focus on, and for the reader's sake you must indicate (without getting involved in) the nearest of the territories you won't be getting into. This sort of skill is far more creditable than any memory-dump, however compendious. It might even be more creditable than the right answer.

It's not an easy line, trying neither to be thought an ignoramus nor to treat your reader as one. Some things are too elementary to be worth saying at all; others, though as elementary, really ought to be said, but only if it can be done lightly, incidentally, allusively.

Exposition and Criticism are just two of the strands that have to be interwoven. If you remind readers, now and again, and unobtrusively, which strand they're on they will be grateful. Even road-traffic needs not only signposts before the turn-off but little reassurers after it.

What direct exposition fails to achieve will not be made good by a double dose of the same; but only by indirectness, the light shed incidentally in the course of (apparently) conversing about something else.

If you have summarized a controversy without servile literalness and with a due sense that the loser's errors are likely to be more creditable to him than your refutation is to you, it matters not at all whether you conclude for the `right' side. Nor is there any requirement to conclude at all costs. A hung jury shouldn't pretend not to be (and in any case the pretence is transparent). On the other hand, since less is at stake than at the Old Bailey, a verdict can be tried for size, in the language of trying-for-size, and with the best supporting case you can make out.

Are there not models, to convey by example what cannot be conveyed by precept or correction? But in all the weary years of essay-setting I have never known a colleague show to the class specimens of a range of answers to the same question. It's unthinkable. But whereof one cannot think, thereof one must practise thinking.


"Synonymy is suspect because no two terms have exactly the same meaning." What is it that hangs on the existence ofexact synonymy which would not be near-enough satisfied with near-enough synonymy ? We all need to talk about înear-enough X╣, and always call it îX╣ for short, calculating that the dangers of doing so, though obvious enough, are outweighed by the benefits. Can there be any point in considering whether two terms have exactly the same fuzzinesses ?


Some exam questions seem to be looking for standard commentaries on standard issues, some for non-standard commentaries on standard issues, and a few for the detection of non-standard issues. This variety would be a little less unbearable if it were freely acknowledged.

Few of the many teachers engaged in examining think half as hard about the problems of examining as they do about their own research. Few have carried such reflections as they have had to the point of writing them down, whether for their own clarification or to invite criticism from colleagues; fewer still, to the point of publication, which invites criticism at large. And every June they reproach themselves for not having done any of these things.

Rubric (Arts): If anything in the following passage intrigues you, talk about it in the way and at the length that seems appropriate. (This is almost the same as a blank rubric, but not quite.)

Some examiners accept that it is the examinee's right to read the rubric as uncooperatively as a computer might. Rather fewer enjoy attempts to jolt the student out of routine sullenness by high-voltage ingenuity.

Aristotle: "Clearly the art of examining does not consist in knowledge of any definite subject."

P. Hilton: "[The exam system] seems designed to select principally for the curious ability to obtain correct answers to imperfectly understood questions."

Augustus de Morgan: "Only yesterday a friend told me that while walking in the street, violent and frightful screams startled her, and on enquiring at the house from whence they came, she was told that a young lady was dangerously ill of brain fever, having just passed a College examination."

R. F. Harrod: "What was this Greats? If in this centre, an arsenal of British thought, there could be such inspissated parochialism and complacency, were we not indeed in a rather parlous position? I frankly view the tendency of the dons to encroach upon the life and time and energy of the undergraduates with suspicion."


Aldrich is critical of the sterilized but idle examples so common in the philosophical literature; but his own are just as far from being case-histories. The trouble with `neutral' examples is that they are not as neutral as they look. Their vocabulary shows an undue proportion of well-worn low-tech items like tables and chairs, set in that most serene of environments, the professor's study. Such examples are like fourth-form lab-experiments: they are guaranteed to demonstrate exactly what the teacher wanted, provided he does the demonstrating himself.

But where the difference between asserting and denying is a real-life controversy the vocabulary is of a more protean kind; the range of meanings in one's own use, to say nothing of one's interlocutor's, cannot be relied on to correspond neatly with the range of words. If anything, it is rather more likely that all the positions worth defending will fall within the range of permissible meanings of a single key word. The controversy amounts to a war to decide what that word is to mean from now on. Rarely is the victory clear-cut (and never as the result of any single battle). Even more rarely is the contest recorded by the lexicographers.

Can a mere proposition-sentence ever count as an example? Yes; in privileged cases, with a serious attempt to reconstruct just the circumstances its utterance was intended to alter, and the motives for doing so, and the reasons for concealing some of that intention and most of those motives, and the tone which puts the best apples on top. In short, when it ceases to be a mere proposition-sentence.

Everything depends on how the example or case-history is introduced. `Once upon a time...' contrives to convey strong overall reassurance while disclaiming all guarantees. This (it seems to say) will be intriguing, and perhaps rewarding, in proportion as you let your imagination off the leash. No need for it to be shackled by anxiety: no-one can be hurt in the world of Let's-pretend, a.k.a. Hypothesis-land.

If the introduction-slot is empty, the default reading seems to be: Nowhere-&-nowhen-in-particular, and a voice trying to sound disembodied. Now why would any voice try to do that? That puzzle alone raises anxieties enough to shackle the imagination.


The Apostle teaches (a text), and the Fathers also teach what the same text means. Can this be the same verb?

There's an embarrassment inherent in the interpreter's position. The commented text is of higher status than the commenting text. Nevertheless the latter implies, what the former omitted, that high-status texts cannot manage without lower-status ones.

Something must be done to stop the second of the above uses of the verb from being mistaken for something very like the first, or as the route to it, or as that which isn't supposed to need still-lower-status interpretations. This last use seems to declare an end to the process it has itself begun -- like those who are broad-minded about all changes in the language up to the time of their own O-levels.

What if it should turn out that the Apostolic text were itself at second hand?

We exegetes serve a narrow bracket: those who can't manage the books without our help, but can manage us without anyone's help. We can't get away with just re-expounding, because it's unbecoming to set up as a better expositor than, say, Hume or Mill; and because it generally turns out that we are not in fact expounding a giant who will survive our explanation of him, but some intermediate interpreter of intermediate eminence; and because to imply that there are n standard interpretations to be memorised is exactly as reprehensible as implying that there is only one. To accept any of a range of answers is fine; but not if in doing so we imply that the questions raised by the text are conveniently few, all uncontroversially known.


There is inevitably something odd about the Exercise. It has to stand in some defensible relation (neither too like nor too unlike) both to the Exposition which preceded it and to the Examination which will follow it. It must be accessible to all but the dimmest, and at the same time stimulating even to the brightest. Its rubric seeks to be pellucid, but problematic. And it should try not to bore the marker.

Most exercises are (and any could become) quite highly stylized; to that extent they are exercises in not putting a foot wrong. They allow radical misapprehensions to survive unguessed-at. Or is that their function? for radical misapprehension casts aspersion on both the misapprehender and the misapprehendee.

The good teacher is half expert, half Svengali. He knows, under those respective hats, (i) his stuff, and (ii) how to make it someone else's. The devices available to him are the Exposition and the Exercise. In practice the proportion of Exposition to Exercise may vary from zero to infinity; and the relation between them, where they both exist, also varies widely. In general, Arts courses do not aim at a close relation; and in the common case where Exposition is all lectures and Exercise is all essays, the absence of any statable relation is counted a merit. Such zero relationship is guaranteed, in many university-level language courses, by having no Exposition for the Exercise to relate to. After all, we can all think of some things that are learnt without Exposition, or without Exercise, or without either.

Overstating the merits of new exercise-types, and the demerits of old ones, keeps the educational world turning. After all, if we innovators understand our own proposals, it can only be in a programmatic, pre-experiential sense; the principles they exemplify may not be quite so unmistakable, so undeniable, so timeless as we implied. Multiplication tables, or prose translation (or any other bugbear) was an innovation once, promising more than it achieved, but achieving something. We had better concede that in some teachers' hands old exercise-formats go on achieving results long after you and I have exposed their radical unfairness, or unclarity, or dreariness; and of course even unfair exercises can be used instructively by students not concerned to get away with the minimum. Would-be stimulating exercises can always be tackled as if they were humdrum, and vice versa.

The exercise fulfils its function when it produces something beyond comparison more interesting than an exam answer: -- core-samples from prospecting in the library. Those who have caught that bug, on however small a scale, need no further convincing that the process is worthwhile. They now have a reason for writing; which is to say, they meet at least the first condition of writing something worth reading.

The purpose of an assignment isn't just to show you remember what teacher said. It is to randomize things so that, in addition to indicating how much has been retained on some point, the student incidentally reveals unsuspected misapprehensions. So all assignments are traps, of a kind where the hunter doesn't know what he is trapping for, and where it is in the prey's interest to trip as many wires as possible. This is the only way for student and lecturer to understand what each has failed to understand. Half the teacher's purposes in setting an assignment are discovered after the event.


A data-set (some personally collected) +

at least two competing analyses of it, of nearly equal plausibility +

a criticism of the fit between each analysis and the set +

a criticism of the shortcomings of the set itself, and its mode of collection.

Can questions on a text be, in principle, simultaneously unambiguous, clear, fair, and challenging? The possibility of having to cross the line from `On the whole Yes' to `On the whole No' is one I no longer find unthinkable.

Exercise, Arts

In any of its eccentrically-named forms (Essay, Prose, Translation) the Arts Exercise combines two major advantages: it is simple to set; and there is no point in expecting benefit from anything less than a long-term treatment -- to be measured in years. The combination gives a foolproof and wearproof solution to the organizational part of the teacher's problem (What the hell to set for next week's work), though leaving untouched the intellectual part of his problem (What the hell to say about last week's work). With such a course one has no difficulty in dashing off the syllabus, or in resisting colleagues' raids on all that lovely time.


Epictetus: "Do I go to my teacher as to the oracle, prepared to obey, or do I too, like a snivelling child, go to university to learn nothing but the history of philosophy and to understand books I didn't understand before, and, if given the chance, to explain them to others ?

Man, you have had a fight with your slave at home, and turned the house upside down and disturbed the neighbours, and turn up at my class with a knowing air and sit there poking holes in my my explanations and complaining about the teacher. You have come in a spirit of envy, depressed because the parcels from home have stopped coming; and while the discussion is going on you sit there worrying about domestic matters and wondering what people are saying about you back home. `They think I'm doing well, and expect me to come back knowing everything there is to know! I suppose I was eager to learn, once, but it's such hard work, and nobody sends me anything, and the baths at Nicopolis are a dead loss, and you should just see my digs, and I'm not getting anything out of the course.'

People say, `No-one gets any good from the lecture-room.' But which of you come with the expectation of being cured ? -- of having their judgment improved? of working towards an understanding of what their real needs are? Why then are you surprised if you take away only the qualities you brought with you? For you do not come prepared to lay your present qualities aside, or of correcting them, or of getting others in exchange. Far from it. Ask yourselves whether you get what you came for. You came hoping to learn how to chatter about principles; that at least you are now fluent at. Doesn't the course give you enough opportunity for displaying your precious principles? Can you not hold forth for hours on the Liar paradox and on contrary-to-fact conditionals? Why then do you complain, when you are getting what you came for ?"

Experience, lessons of

All this supposes I know what worked for me -- itself a far from convincing claim, because so gratifying if true. It's more likely that what worked for me can't be reconstructed: we survive on mythoepic versions. All that varies -- but it's enough for useful disagreement -- is the degree of complacency with which each of us contemplates his own idealization.

So the best exposition would be an edited replay of what seemed to work for me, eventually -- massaging the facts as all histories do, leaving out the dead patches, artistically improving the sequence, hinting at which bits would repay meditation and which are passage-work. But it would be wary of sacrificing, in the name of the quintessentially methodical, any of the benefits of the case-history.

On the subject of what didn't work we are all more articulate, though still not obviously reliable -- and of course not much help constructively.


There are (let us suppose) level-1 experts in some field.

Then at level 2 there are those who claim to be also experts at transmitting their level-1 expertise -- but fewer of them, and the evidence to justify the claim is a good deal less stringent than at level 1.

Then at level 3 there are those who claim to be also experts at the transmission of expertise in general -- but fewer still, and with evidence even less stringent. [This remark evidently claims to be level 3, or worse.]

As we go from level 1 to level 3, the need for expertise rises and its reliability falls; painfully, in both cases. This is inevitable (and therefore not in itself a scandal).

At level 1 some disciplines like to think they are `harder' than others; at level 3 they're all in the same boat. That is, the `harder' they start, the harder they must be prepared to fall. What they have in common is a need for bi- or tri-lingualism -- the recognition that different ways of talking, and in particular a wider range of degrees of asseveration, are needed where what we say has different ceilings of reliability.

This may entail greater tensions for academics who take pride in a high ratio of Understanding of basic concepts to `mere' know-how (at level 1). It isn't easy to switch languages when one moves from a domain where quantifiably-testable theory is supposed to be the sine qua non to others where it is something else -- an ideal, a dream, or a constraint. Yet domains exist where even `Basic Concept' isn't a basic concept -- that is, where such talk can be a way of begging a question. (Come to think of it, how long since we scrutinized the credentials of `Domain' ?)

Use of the word "expertise" leads, I fear inevitably, to a binary division of the world into experts and non-experts, with only the former (in the opinion of both sides) having what it takes to voice the Verdict of Science.

In a variant of this, not so different as may at first appear, there are `scientific' experts who are sure to dis agree, and `generalist' experts who can tell which set of `scientific' experts is the more reliable. Both variants are static and divisive, the first assuming one, the latter two unbridgeable gulfs.

Occasionally one comes across a different kind of expert -- so different as to deserve a different name -- who has a lively awareness of both the responsibilities and the temptations of expertise, and who tries to convey to outsiders the more important `knowns' and the degree of confidence with which each of them is `known'. Talking in these terms would be easier if there were a neutral term between Confidence and Unconfidence, and another between Known and Unknown. For lack of these, even confessions of hesitation must be expressed in the terminology of certainty.

Matching the degree of asseveration to the degree of understanding is a Sisyphean task: on either side, the slope is steep to the exaggerations of immodesty and over-modesty. No-one could be expected to stay on top all the time; pity the teacher who cannot smile wryly as he starts back up.


To help me see what he meant he struck a metaphorical match; not realizing how long it took to recover, both from the white dazzle it brought and the black dazzle it left. Still, I'm grateful; though there wasn't quite time, between the dazzles, to tell if we were looking in the same direction.

Too little explicitness leads to obscurity, which attracts the wrong readers. Too much begins to sound like condescension, which repels the wanted readers. Some is needed for stage-1 learning -- the push in the saddle that starts the rider off. Too much prevents stage-2, self-pedalled learning.

Resolution: to stop presenting as if it answered all the questions an explanation few of whose implications the hearers are yet in a position to absorb; and so either antagonizing them (if they resent this sort of pressure) or seducing them into pretending to understand. An explanation may be presented quasi-autobiographically, as having provided the crucial stimulus (but no more than the stimulus) which ultimately led, after further inputs which there isn't time to enumerate, to the explainer's present fuller understanding; but the temptation must always be strong to understate the difference between immature and mature understanding, and to present the explanation as constituting sufficient grounds for full understanding. I hope I leave you with the feeling that you have in some sense 'understood' the above explanation, inadequate though that sense may be.

Explanation is the autobiography of that subset of one's former perplexities which the hearer is still struggling with. How could one explain what one had never been puzzled by, or to one who sees no puzzle?

An educational system earns its keep if it does nothing more than maintain the right meaning of (what counts officially as) Explanation. It isn't that hard, if you repeat often enough mantras of the form: `This (form of words A) is explained by that (form of words B)' or `A, because B'. It would be worth totting up the frequency of such phrases, compared with that of less simple-minded schemata, in a random manual; but in order not to cause needless offence we should pick one that is now amusingly out-of-date, from say fifteen or twenty years ago. (`Maintain' here has the engineering sense, which extends to a little debugging, provided it is done discreetly.)

Isaac Todhunter: "`Sir,' said Todhunter to Clerk Maxwell, `If a young man will not believe his Tutor, a gentleman and often in Holy Orders, I fail to see what can be gained by a practical demonstration.'"

George Eliot: "But let the wise be warned against too great readiness at explanation: it multiplies the sources of mistake, lengthening the sum for reckoners to go wrong."


From the lectern, my exposition looked like the M1; from the hall, like Spaghetti Junction.

Expounding is heavy work. At every point there is a choice of lines, for every line a choice of words, all of which, though at first reasonable candidates, turn out to have snags. We earn our keep by knowing a snag when we see one, by our cunning at devising a way round it, by our readiness to acknowledge (notwithstanding that investment of cunning) a further or worse snag in the hard-won solution, and in general by our stamina at this unending game.

Between the classical theory and my annual exposition of it there is an embarrassingly obvious chasm. Only one of them is in print, and so open to scrutiny from all comers. The printed text may try to sound authoritative, and the oral exposition of it may try not to; but it is the former which invites criticism on the fairest terms.

Exposition-1 offers itself as an improvement on the original, or doesn't do enough to stop being taken as one. Exposition-2 offers itself as a better-than-nothing substitute for the original, or doesn't do enough to stop itself being taken as one. Exposition-3 says `The only point of this is to be tested against the original, line by line', but disclaims responsibility for ensuring that a realistic amount of time is left free for this (and even for estimating how much time that would need). Exposition-4 is a scattergun mixture of all the others, hitting assorted targets in proportions it considers unknowable and unimprovable.


Kleist: "It is strangely inspiring to speak to someone directly facing us; a glance which proclaims that our half-expressed thought has already been grasped often gives us the expression for the whole other half. I believe that many a great speaker did not yet know at the moment that he opened his mouth what he was going to say. But the conviction that the profusion of ideas that he needed could be drawn from the situation and from the resulting stimulation of his mind made him bold enough to begin, trusting to luck."

Back to Contents