Writers with a reputation for clarity are often complimented in terms that locate the virtue in the wrong place. ‘Delightfully clear sentences’ is usually the theme, and there are many variations on it. What they show is that clarity is taken to be first and foremost a property of the sentences. This is a mistake: clarity is first and foremost a property of structure. Just as the architectural design of a building does not inhere in the bricks, and might remain the same if the whole building were constructed of other materials, so the chief repository of clarity in a verbal communication is a structure that can remain the same when all the words and all the sentences are different, as they are in a translation. Needless to say, it is nevertheless important to have clear sentences, just as it is important to have sound bricks and strong walls.

Not only does a reader need to be clear in his mind about what is under discussion at every given moment, he needs to understand why it is being discussed. How did he get to this point? What steps were gone through to bring him here in such a way that the question now reached seems to have arisen naturally? And what expectations does this suggest about the next step, or choice of steps? Only if a reader is clear about these things too does he really know where he is. And they are matters of order. Structure is of the essence.

Some philosophers, most importantly Kant in his Critique of Pure Reason, lay out a structure like this with the utmost clarity, yet in unclear sentences. In his case it was because he had spent many years thinking his critical philosophy through, but then wrote it down hurriedly because he was afraid of dying before he finished writing the book. The result is clear thinking expressed in unclear sentences.

An essential truth is that every extended piece of speech or writing contains two kinds of structure simultaneously. Taking our terms from Susanne K. LangerFootnote ¹ we can call these ‘discursive structure’ and ‘presentational structure’. Discursive structure has a time dimension, presentational structure does not. We apprehend a discursive structure moment by moment as we listen or read. It comes at us thought by thought, or point by point, in words and phrases each of which follows the one before and leads to the one after. This process can vary a great deal in quality, depending on the degree to which the components are well chosen and assembled, and then skilfully joined. It can be marred by apparent contradictions, interpolations and digressions, non sequiturs and repetitions, all of which may be expressed in clear sentences.

The other kind of structure, presentational structure, is what stands before us when the discursive structure is complete. We see it as a whole, a Gestalt, its parts exhibited simultaneously. The crucial attribute here is proportion, the accurate representation and placing of all the parts in their right position, size and function, relative to one another and to the whole. It has been widely argued (for instance by Wittgenstein in the Tractatus) that presentational as against discursive structure cannot be depicted in language, since any attempt to do so is inescapably discursive, unless it be merely a replication of what it purports to describe. On this view, a presentational structure is one whose properties can be exhibited but not stated: they can be shown but not said. Presentational structures may exhibit formal beauty, not only in expository and argumentative prose but also in logic, in mathematics, and in scientific and other such theories. Perhaps they are at their most beautiful in the arts, above all in music, which possesses both discursive and presentational structure. The story is told of the composer Sibelius playing a record of one of his works to a visitor: when the performance ended, the visitor asked him to explain the work – whereupon Sibelius, without uttering a word, played the record again.

It has always seemed to me, as a matter of direct experience, that both the creation and the apprehension of the two different kinds of structure are non-linguistic. In the case of music this is self-evident. But it is equally true of structures exhibited in language. I recognize immediately what the Nobel Prize-winning scientist Jacques Monod is describing in his book Chance and Necessity when he writes (page 146 of the English edition): ‘I am sure every scientist must have noticed how his mental reflection, at the deeper level, is not verbal: it is an imagined experience, simulated with the aid of forms, of forces, of interactions which together barely compose an “image” in the visual sense of the term . . . In everyday practice the process of simulation is entirely masked by the spoken word which follows it almost immediately . . .’ Even in everyday matters I am aware of the word as coming after the thought, not as accompanying it, still less as being it. Sometimes I have difficulty in finding words to express what I already know I want to say. I am always conscious of the formulation of sentences as being a kind of in-filling of spaces in a structure that I need to get clear in my mind before the sentences can be formulated. When Monod goes on to say (on the same page): ‘But, as we know, numerous observations prove that in man the cognitive functions, even the most complex ones, are not immediately linked with speech (nor with any other means of symbolic expression)’ he is saying only what my most immediate experience has already shown me. I am aware that ‘modern philosophers, some of them . . . believe that we think in words’,Footnote ² but it has always puzzled me how they can. I would have supposed, rather, with Chomsky, that ‘everyone who introspects will know at once that much of his thinking doesn't involve language.’Footnote ³

The mere existence of music is enough in itself to show that discursive structures can be characterized by the highest degree of complexity and sophistication without the use of language, either in their formulation or in their apprehension, and perhaps without the use of symbols at all – there are many cases on record of composers completing pieces of music in their heads before, sometimes even without, writing them down. But most of us have had the experience of suddenly seeing our way through an argument, perhaps quite a long and complicated one; and the very fact that this is sudden, a click (a popular phrase for it in English is ‘the penny drops’), shows that the explanation cannot be that we run through the rest of the argument to ourselves in language. We see it all at once, a Gestalt, without any use of words. What has happened is that we have seen the argument's presentational structure. But we cannot say what it is we have suddenly grasped, except simply to say that we have got hold of the rest of the argument. If we are challenged we can prove that we have by rehearsing the argument discursively, but we cannot reproduce in words what it was we grasped in the instant when we grasped it.

The achievement of maximal clarity in exposition consists in the accomplished creation and communication of both presentational and discursive structures, the former being the more important, because they constitute what we normally mean by ‘content’. When we achieve more than a superficial understanding of an exposition put forward by someone we grasp first (in time) its discursive structure and then, through that, its presentational structure. If creator or recipient operate on the level of words alone this is glibness in the former and the mugging up of a subject in the latter. Real creativity and real understanding consist not in spinning or soaking up words but in the creation and apprehension of presentational structures which the words articulate or exhibit. Language is their medium, not the focal point of their creation or understanding.

If one is to make something fully clear to others one must first make it fully clear to oneself. This means thinking it through to the bottom, to the point where one has a complete grasp of its presentational structure. This process, as I have quoted Monod as saying, is not verbal. From there one needs to embark on what may be a long and laborious task of creating a discursive structure which is articulated in words and which, when the listener or reader has followed it through, will place the presentational structure as clearly before the reader's mind as it had been before the writer's. In this process, taken as a whole, the use of language, both for the writer and for the reader, is intermediate. Its clear use is important but not indispensable. Think of Kant.

The first and most essential stage in achieving clarity, the stage of deep non-verbal reflection, is something about which, in the nature of the case, little can be said. To a few remarkable beings, clear presentational structures seem to come as the leaves to the trees. Mozart sometimes said that when a piece of music came to him it arrived all at once. There may be quite a few of us to whom this happens occasionally. However, most of us, most of the time, need to struggle through prolonged, frustrating, unverbalizable gropings to get the necessary pre-verbal grasp of what it is we want to say. This requires prolonged concentration, and for it we usually want to be left undisturbed for long periods. Hence the familiar self-isolation of the creative worker, his basic need being not solitude or time but absence of distraction. Any stimulus strong enough not to be ignorable is a threat to creativity. Keeping all, or nearly all, such stimuli at bay for the necessary periods of time calls for enough determination to say No to other people, including our nearest and dearest, and to insist on doing our own thing regardless of what they may want or think. This makes demands on our characters as well as on our minds. It would be idle to pretend that accident of circumstance does not also play a part, sometimes a decisive one. Schopenhauer said that he would never have been able to accomplish his life's work if his father had not left him financial independence. I suspect that something close to this is true of a number of great thinkers and artists who have not said it. Obviously, too, many vicissitudes besides the need to earn a living can hobble a creative person.

In communication, what might be called primal unclarity occurs when the first stage in the process has not been properly carried out – when the writer has not thought his subject through to the bottom before beginning to put what he hopes to say into words. In those cases something unfinished and unfocused is passed on to the reader, who is then left to complete a process which the writer has not been through. This happens not infrequently in professional and academic life, partly because the writers are over-busy and therefore hasty, and partly because they feel they can leave their professional colleagues to work things out for themselves if given the gist. At a deeper level, though, there is a constant temptation to hold back from the testing of one's ideas which is involved in the process of thinking them through fully. Ideas that are left just that little bit vague are so much more accommodating.

Stuart Hampshire drove home what is at stake when he praised Bertrand Russell's uncompromising lucidity. ‘It's a question of not obfuscating – of leaving no blurred edges; of the duty to be entirely clear, so that one's mistakes can be seen; of never being pompous or evasive. It's a question of never fudging the results, never using rhetoric to fill a gap, never using a phrase which conveniently straddles, as it were, two or three notes and which leaves it ambiguous which one you're hitting. Russell's prose excludes even the possibility of evasion and of half truth . . . there's always this extraordinary nakedness of clear assertion. His doctrines and arguments stand out in a hard, Greek light which allows no vagueness.’Footnote ⁴

My central contention – that the foundations of this sort of clarity do not lie in the sentences, and are therefore not achieved in a struggle with words but in the creation and perception of appropriate structures – is borne out by Russell's account of his working method in his essay How I write:

Two striking things here are, first, that Russell's prolonged and essential labour went not into anything to do with language, the verbal articulation of his discursive structure – he could, and did, produce that off the top of his head – but into the creation of the presentational structure; and second, that this act of creation proceeded below the level of consciousness, so that his own first apprehension of the presentational structure was of it as a whole.

In the creation of discursive structures the challenge is to use only indispensable elements and to get them in the right order. This needs to be an order that satisfies two sets of criteria simultaneously, one with respect to the subject matter, the other with respect to the reader. As regards the subject matter, one needs to start from the most securely established position that can be found, or to establish it oneself, and then proceed in such a way that each step in the exposition or argument follows logically from the one before. When the subject matter itself is multifarious, or the argument complex, much trial and error may be needed to find a single path that will thread every necessary point into a logical sequence without jumps, backtracking or repetition – often there is only one such sequence; and it may not be self-evident. The familiar experience of writing an opening paragraph and then not knowing how to go on is most often due not to the fact that one has nothing more to say – one's bafflement and frustration give the lie to that – but to the fact that one has started in the wrong place. One has chosen a base from which it is not possible to make the required journey. Finding the right starting point can be the most difficult thing of all, but if one persists in continuing from the wrong one the whole structure will be distorted. One will find oneself constantly having either to repeat oneself, or make not-yet-justified leaps in the argument, or produce ad hoc considerations without warning. The result will be a serious, perhaps debilitating lack of clarity – no matter how lucid the sentences.

All this is easy to say but difficult to do – and sometimes, it has to be faced, impossible to do. If a presentational structure is so complex internally that its every constituent is held in place by three or more others, there may be no way at all of rendering it in the linear form constituted by a single line of exposition, and therefore no way of expounding it without jumps or repetitions. An explosion of frustration at this caused Rousseau to write a heart-cry into The Social Contract when he was less than a quarter of the way through it: ‘All my ideas fit together, but I cannot express them all at once.’ In a case like this, the best that can be done is the best of a bad job. But if some degree of unclarity is unavoidable it becomes all the more important to avoid avoidable unclarity. Too many people give up the struggle and trust that the unavoidability of some unclarity will be taken to excuse whatever unclarity there is.

With the other set of criteria, those that relate directly to the reader or the listener, the starting point needs to be something he understands and finds interesting even though he is as yet in ignorance of what is to come. If he holds a mistaken view about that, or has mistaken expectations, these should be addressed at once, not left to distort his understanding. From the beginning, each step should be such as to commend itself to the recipient in the ever-changing positions he occupies. If I have something I want to share with him, my task is not to start from where I am and carry my treasure towards him step by step, it is to start with him as he is, where he is, and bring him step by step towards me and what I want to share with him. The commonest systematic fault in communication on the part of able people is that they tend to start their attempts from the position which they themselves occupy, and then proceed from there, so that throughout the whole process there is a gap between what is being said and what is being understood.

Quite a lot of university teaching is vitiated by this defect. Teaching calls for close attention to others, an active empathy with them. One needs to begin with some insight into their state of mind regarding what it is one wants to impart. The realism of one's judgment about this governs the effectiveness of one's communication. At its worst, everything one says may be crystal clear to oneself, only for one to discover later that one's audience has understood little.

Another common fault in the articulation of discursive structures is the inclusion of inessential material. This is not just neutrally bad, it actively misleads, for it causes the reader to assume that the explanation, when completed, is going to contain that material, so his anticipation leaps ahead in the wrong direction. Also, it puts irrelevant burdens on his concentration and his memory. The best discursive structures are characterized by an economy that gives them aesthetic appeal. And the fact that every constituent is indispensable makes for bold outlines which render the communication more effective. There is no reason why the result should be dry. A well-made discursive structure, like a well-made presentational structure, has aesthetic properties, though its properties are not timeless. A good exposition or argument hits the right tempo and is shrewdly paced; it may create suspense and arouse excitement; it may perhaps quicken pace, increase tension and build to a climax; the climax may bring some kind of revelation to the audience, which will be followed not by an anti-climax but by a resolution which is both interesting and satisfying in itself. Where such dramatic procedures are exploited to divert attention from poverty of content they smack of charlatanry, but where there is satisfying content and the drama is metabolized into the structure they characterize the work of great thinkers who are also great literary artists, such as Plato and Schopenhauer.

It is not possible to teach people everything they need to know about how to make aesthetically satisfying discursive structures. Each structure needs to be adapted to its content. Nevertheless there are a few procedures that can usefully be taught, and circumstances in which they are appropriate are common. For example there is the Popperian schema incorporated in the formula P¹ → TS → EE → P², where P stands for ‘problem’, TS for ‘trial solution’ and EE for ‘error elimination’. The student is taught to start with a clear exposition of the problem to be confronted, including why it is a problem and why it is worth confronting. He then puts forward his proposed solution. He then considers all the objections to this solution that seem to him to merit attention, especially the most attractive alternative solutions. In the course of subjecting these to critical consideration he learns from them things that enable him to correct or improve his own solution. Finally, having established the preferability of his amended solution to the alternatives, he describes the new situation with particular reference to new problems it uncovers. Admittedly this is a formula, but it is a rich one. If all philosophical and scientific papers were written in accordance with it, most would be clearer than they are.

Defects in discursive structure make for what we could call secondary unclarity, corresponding to the primary unclarity which we located in presentational structure. Only when we raise our heads from this to a tertiary level can questions arise of clarity or the lack of it in sentences. Writers who care passionately about clarity will care passionately about tertiary clarity, and will be involved in a struggle with language aimed at lucidity of verbal expression. But no matter how clear their sentences, if the two underlying structures are unclear the reader will be baffled by their work. This is a common experience for students when they encounter, shall we say, the later philosophy of Wittgenstein. The sentences are lucid, but the discursive structure is bitty, characterized by unexplained jumps, and the presentational structure is only doubtfully existent. The result is that it takes the reader, every reader, a long time to discover why what is being said is being said, what the point of it all is – and therefore, in spite of the lucidity of the sentences, what it means. The inverse of this situation is also to be found – with Kant, as I have said, as the best example. The presentational structure of The Critique of Pure Reason is a model of clarity, and its discursive structure is at least intelligible, yet the sentences are so ill written that the reader has to fight his way through them. Even so, it can be claimed for Kant that his lack of clarity obtains only at the most superficial level, that of language.

There are, however, other philosophers with whom, even when they have something worthwhile to say, unclarity prevails at all three levels. This is the case, for example, with Fichte, Schelling and Hegel. Some of their work borders on the unintelligible. They had been shown by Kant's example that a thinker who is exceedingly difficult to read may, partly for that reason, win a reputation for profundity; and because they coveted this reputation for themselves they deliberately expressed themselves obscurely. In the twentieth century the worst philosopher in this respect who was nevertheless worth while was Heidegger.

With writers such as these one is brought up against moral issues posed by the implications of clarity. ‘La clarté est la bonne foie des philosophes’, wrote Vauvenargnes. (‘Clarity is the good faith of philosophers.’) That one should try to be as clear as possible is a matter not only of intellectual honesty but also of moral obligation to the reader. As Popper said (page 29 of Modern British Philosophy (ed. Bryan Magee)): ‘It's not just a question of clarity, it's a question of professional ethics.’Footnote ⁶ A great thinker, like a great artist, usually feels a high degree of impersonal devotion to his work. And the more highly he values it the more highly he values its communication to others. If he employs unnecessary obscurity what that indicates is that communication of the content is not his primary concern. What such a person wants, as a rule, is to impress the reader – and sometimes himself. Even worse is his attitude to the subject matter, namely a relegation of its importance. A writer who dresses up unclear thought in colourful rhetoric and wide-ranging allusion in order to persuade his readers that the thought is profound and the thinker a genius is using his subject matter for an end unconnected with itself. None of the philosophers I have just named was a charlatan pure and simple, but they all strove to give the impression that there was more to what they had to say than there actually was. This means they must have been aware of its limitations. They had contemporaries who wrote in the same way but had nothing to say, and these are now forgotten, their names uncovered only by historians of ideas. With such mountebanks the presentational structures either do not exist or cannot be made to hang together. The discursive structures do not hang together either. But their cracks and crumblings are pasted over with a wallpaper of words whose decorative qualities impressed some of their contemporaries. But whatever it was that impressed existed solely on the level of language. Beyond that there was nothing.

Not only the desire to impress but any other impurity of motive is likely to cause unclarity, because in so far as it makes itself felt it causes things to be said which would not have been said otherwise and are superfluities or distortions. Common among these are the desire to live up to a certain image of oneself, or to gain promotion in one's profession, or a reputation outside it; the desire to evade other challenges or responsibilities by seeking refuge in writing; the desire to make money. These things and others reveal themselves in writing – sometimes as the motive for writing – and always muddy the waters of communication, if only because they are themselves being communicated, usually unconsciously.

Not all unclarity is culpable. I have instanced already one cause of unavoidable unclearness: a structure so complicated that there is no way of rendering it in an unbroken line of exposition. This insoluble problem causes me to envy the musical composer, who can not only keep several voices or strands of argument going simultaneously but in a way that keeps their interrelationships on perpetual display. One of the worst limitations of verbal utterance is its lack of anything corresponding to polyphony. Another is when what is being expressed is so radically new that the writer has been unable to find a way of articulating it without doing violence to established uses of language. Language now needs to be stretched, paradoxes embraced, metaphors resorted to. Kant on the transcendental ideality of the empirically real, Heidegger on conscious self-awareness, Freud on the unconscious, Einstein on time and space, these were all saying something so innovating that to say it at all they had to use language in ways that were at odds with its established use. This calls for tolerance and imagination on the part of the reader if he wants to understand. In such cases readers who deny the meaningfulness of what is being said are being conceptually conservative to a fault. Linguistic philosophy was conspicuously culpable in this regard. Wherever established usage is set up as the only criterion of meaningfulness, the possibility of conceptual innovation is precluded. While keeping our guard up against oracular charlatans we need to be perpetually open to an unprecedented use of concepts.

The need sometimes to use concepts in unprecedented ways in order to express the new explains how a writer who strives for and achieves clarity can sometimes be underrated. There is a feeling that if what he says is clear it cannot be original. Worse than that, there is confusion of the clear with the self-evident, and hence with the obvious. There are people who feel they have no need to read anything if it is entirely clear, because they sort-of know it already. This is misplaced. Although some of the greatest thinkers have been unclear, others have been wonderful communicators. One thinks of Plato, Descartes, Hume and Schopenhauer among philosophers. With writers like these the writing contains everything necessary for an understanding of itself. Nothing has been left in the inkwell. But clarity in writing, as in mountain lakes, disguises depth. No one, so far as I know, has accused Plato of being superficial, but this charge has been made against all the others. They have sometimes been met with the reaction: ‘I see what you're saying but why are you bothering to say it?’ Yet it does not at all follow from the fact that something is clear to me when it is put before me that I could have thought of it for myself. From the fact that I understand a book it does not follow that I could have written it. The clear is not at all the same as the obvious. But there are readers who feel an extra respect for writing that demands an effort from them, even if that effort goes into work that could have been done by the writer. The fact that they are having to put more into it makes them feel they are getting more out of it.