Frye says that to some extent 'structure' and 'system' are synonyms of 'diagram', so of course I had to consult the OED to get a feel for the etymology. I'm surprised to find that the Greek diagramma (see the LSJ and Middle Liddell entries here) has the uses:

I. 1. that which is marked out by lines, a figure, a plan, or 2. a geometrical figure, diagram; II. a written list, register; III. a decree, edict.

For the second sense, an obsolete one for 'diagram' in English, the OED has 'a list, register, or enumeration; a detailed inscription; also, 'the title of a booke', and its citation from 1631 is about monument inscriptions: 'An Epitaph is..an astrict pithie Diagram, writ..vpon the tombe..declaring..the name, the age..and time of the death of the person therein interred'.

(Then there is diagrammizo (LSJ here), 'divide by lines: hence, play at checkers'.)

I find this surprising because I had been thinking about what one might say about the presence of the diagrammatic in certain forms of prose, where a literal diagram with the pictorial clarity of something like Plato's line doesn't automatically suggest itself. You do get the sense that for philosophical writing with a metaphysical ambition and aiming for a certain degree of transparency in its 'view' of the world (not just words in a book, or words expressing thoughts, but words that are a means for the recognition of what really is, what exists, in the world), diagrams concerning an author's ontology can naturally be read off their writing. And they needn't even be dictated; there is plenty of choice about what one might wish to represent in such a diagram. Taking after a teacher of mine, I often represent Descartes' ontology with a table, mainly so I can represent the dependencies of certain kinds of things (modes, finite substances) on others (finite substances, infinite substances). I suppose if one wanted, a diagram of the meditator's self, or mind, and its contents, could easily be made. Spinoza, same sort of thing.

What gave me pause, in thinking this way, was trying to apply similar thoughts to someone like Kant, someone at least within the ambit of easily diagrammable early moderns, but also someone to whose thought this notion of transparency seems iffy to apply. It's been a long time since anyone tried to teach me any Kant, but I still recall the difficulty they had representing the idea of the self, or at least the 'I', that is supposed to fall out of the transcendental deduction. Likewise their attempts at drawing screens between the knowing subject and the things themselves, or between some sort of formless manifold and the subject, or at somehow representing the field of experience as structured by categories provided by the subject; and woe at any attempt that seemed to suggest we had better hide a thing-in-itself somewhere back behind some other part of the diagram, a fat scribbled dot or an X or a question mark. Kant's own tactic in §16 seems telling: 'It must be possible for the "I think" to accompany all my representations...' sidesteps any pictorially diagrammatic aspirations in favor of the discursively logical, the linguistic. And as far as I know anything about Kant (gossip, hallway talk, summer reading), there's nothing inappropriate about treating him as logic-centric, in a broad sense, and in a way that harmonizes with my feeling that the notion of transparency (of prose as a window to the world) does not suit him.

So far so good. What gave me pause was the obvious thought, in connection with Kant, that his logic-centricity ends up somehow being the complement to the shape his prose takes, the multiply articulated structure of the first Critique, the array of terminology that coordinates somehow with that structure, the legalistic way its sentences unfold, the large sections whose aim is to give arguments about arguments (rather than, say, arguments about things). It seems that an unsurprising effect of all this is that 'diagrams' of Kant's thought tend to look like his writing, like the complicated tables of contents - in fact, like the diagrammatic devices he preferred, tables and parallel columns and such.

Well, that didn't give me pause. But what did was: appreciating that something like this hyper-articulation of the verbal means of philosophical writing seems characteristic of a lot of philosophical writing after Kant, what might one think about the possibilities for extracting diagrams from this writing, when it lacks that complement to the prose, the logic at the heart of the project? There, at least, efforts at diagramming, however complicated or multipart they might get, seem to enjoy the promise of bottoming out, or at least terminating, in something solid, or at least appreciably limited, graspable: lists of categories, forms of judgment. (Imagine the attraction this must have.) Past a certain point, once logic becomes dissociated enough from traditional categories, this rooting down in logic even seems to mean that the diagrams come themselves to populate the philosophical writing, almost to infect it down to the word, letter, symbol. So far so good; one can always suppose that the surface appearance of modern philosophical prose diagrams out in terms of what it has to say about such structures, or what is expressible in such structures. But... what if an author is something more like Wittgenstein, or something more like Cavell? Where should efforts to diagram their thought seize upon the terms to be represented and related in diagrammatic form?

I ask because I've got two thoughts in mind. One is that when Frye includes 'chapters' and the whole rhetorical apparatus of prose-navigation - one of my favorite tools, or starting points, for investigating the poetics of any prose - I have the sense that he's not wrong - even what is thoroughly, conventionally rhetorical about a text, and even what it it borders on or meets up with the paratextual, is somehow diagrammatic - but that these are features of a text's diagrammatic character that should properly be subsidiary to whatever has most to do with its dianoia, its thought (Frye likes to gloss this as 'meaning' in Anatomy, which is fair enough in connection with the interpretation of philosophical writing, too).

The other thought is that, for certain kinds of philosophical writers, what you are likely to come up with when you seek diagrammatic structure in their writing is just the sort of stuff that has to do with the shape of the text more than the shape of the thought. And if that shape of the text does not seem to subtend 'an argument', as for example I think the Investigations does not and as Part Four of The Claim of Reason makes no pretense toward doing, then to what avail would someone writing about such a text (e.g. me) try to capture something of that shape for the purposes of communicating something of the workings or the effect of the text to one's own readers?

The surprising thing about the etymology of 'diagram' is that it invites retrieval of the sense that there's nothing particularly distinguished about the at base Platonic notion of the diagrammatic, as a permanent kind of relation among elements of thought or elements of the world, such as might serve as the typical referent of philosophical writing - nothing particularly distinguished about this as against the notion of a list, of stuff that is together just in that it is all written here in the same place, side by side or one after the other, and otherwise together because of what the list-maker means by making a list, not necessarily because of something about its items (conceived independently of the list-maker or the list-reader) or the relations between them. (Shades of Foucault's Borges.) A sequence is not a list, but there seems to be some affinity between them. And a great many of the forms of prose I would like to say are nondiagrammatic are also fundamentally sequential. Though perhaps they would seem more diagrammatic if I were to represent them using selves and readers, acts and lives.

Northrop Frye on what I would call the diagrammatic aspect of writing:

'The two elements of subconscious association which form the basis for lyrical melos and opsis respectively have never been given names. We may call them, if the terms are thought dignified enough, babble and doodle. In babble, rhyme, assonance, alliteration, and puns develop out of sound-associations. The thing that gives shape to the associating is what we have been calling the rhythmical initiative, though in a free verse poem it would be rather a sense of the oscillations of rhythm within an area which gradually becomes defined as the containing form. We can see from the revisions poets make that the rhythm is usually prior, either in inspiration or in importance or both, to the selection of words to fill it up. This phenomenon is not confined to poetry: in Beethoven's notebooks, too, we often see how he knows that he wants a cadence at a certain bar before he has worked out any melodic sequence to reach it. One can see a similar evolution in children, who start with rhythmical babble and fill in the appropriate words as they go along. The process is also reflected in nursery rhymes, college yells, work songs, and the like, where rhythm is a physical pulsation close to the dance, and is often filled up with nonsense words. An obvious priority of rhythm to sense is a regular feature of popular poetry, and verse, like music, is called "light" whenever it has the rhythmical accentuation of a railway coach with a flat wheel' (pp. 275-76).

'The first rough sketches of verbal design ("doodle") in the creative process are hardly separable from associative babble. Phrases are scribbled in notebooks to be used later; a first stanza may suddenly "come" and then other stanzas of the same shape have to be designed to go with it, and all the ingenuity that Freud has traced in the dream has to be employed in putting words into patterns. The elaborateness of conventional forms - the sonnet and its less versatile congeners the ballade, vilanelle, sestina, and the like, together with all the other conventions that the individual lyric poet invents for himself - indicates how far removed the lyrical initiative really is from whatever a cri de coeur is supposed to be....' (p. 278).

'We are now in a position to find more acceptable words for babble and doodle, the radicals of lyrical melos and opsis respectively. The radical of melos is charm: the hypnotic incantation that, through its pulsing dance rhythm, appeals to involuntary physical response, and is hence not far from the sense of magic, or physically compelling power....' (p. 278).

'We have several times noticed the close relation between the visual and the conceptual in poetry, and the radical of opsis in the lyric is riddle, which is characteristically a fusion of sensation and reflection, the use of an object of sense experience to stimulate a mental activity in connection with it. Riddle was originally the cognate object of read, and the riddle seems intimately involved with the whole process of reducing language to visible form, a process which runs through such by-forms of riddle as hieroglyphic and ideogram....' (p. 280).

'If there is such a thing as conceptual rhetoric, which is likely to increase in proportion as the discursive writer tries to avoid it, it seems as though the direct union of grammar and logic, which we suggested at the beginning of this essay might be the characteristic of the non-literary verbal structure, does not, in the long run, exist. Anything which makes a functional use of words will always be involved in all the technical problems of words, including rhetorical problems. The only road from grammar to logic, then, runs through the intermediate territory of rhetoric....' (p. 331).

'[T]he reader... may at any rate be willing to admit the possibility of links between grammar and rhetoric, and between rhetoric and logic, that have a neglected but crucial importance.

We remember that a good deal of verbal creation begins in associative babble, in which sound and sense are equally involved....' (p. 334).

'The link between rhetoric and logic is "doodle" or associative diagram, the expression of the conceptual by the spatial. A great number of prepositions are spatial metaphors, most of them derived from the orientation of the human body. Every use of "up," "down," "besides," "on the other hand," "under" implies a subconscious diagram in the argument, whatever it is. If a writer says "But on the other hand there is a further consideration to be brought forward in support of the opposing argument," he may be writing normal (if wordy) English, but he is also doing precisely what an armchair strategist does when he scrawls plans of battle on a tablecloth. Very often a "structure" or "system" of thought can be reduced to a diagrammatic pattern - in fact both words are to some extent synonyms of diagram. A philosopher is of great assistance to his reader when he realizes the presence of such a diagram and extracts it, as Plato does in his discussion of the divided line. We cannot go far in any argument without realizing that there is some kind of graphic formula involved. All division and categorization, the use of chapters, the topotropism (if I have constructed this correctly) signalled by "let us now turn to" or "reverting to the point made earlier," the sense of what "fits" the argument, the feeling that one point is "central" and another peripheral, has some kind of geometrical basis....' (pp. 335-36).

The Inward Morning falls into two parts, or it is kept over two stretches, from August through October of 1952, and from July through November 1953. Bugbee's thoughts, and the writing he was capable of, were diverted by his teaching responsibilities. (This is both heartening, and disheartening, to read.) When he starts up again, it is in much the same mood as his first beginning (August 1952: 'I feel that for me a crisis is at hand', July 1953: 'the first day in which I faced the crucial year ahead'). But inflected; now his effort to steady himself, to get his bearings, includes some reference to the professional culture that places little value on, or can offer little understanding of, explorations such as his. It is especially his way of writing that seems conspicuous, problematic, to him here.

The other day while reading I was struck by what seemed like a predominance of questions in Bugbee's prose. For example, of the roughly twenty sentences of the September 26, 1952 entry, half of them are questions; often one right after the other.

That characteristic isn't as pronounced as I thought, though it does seem telling about the kind of writing Bugbee needs for his undertaking. I was probably struck by it because I perceived a similarity to Cavell's prose; there are passages in The Claim of Reason where he will similarly 'answer', or follow, one question with another, with another, with another. I think Mulhall has associated this feature of Cavell's writing with what he sees as a similar one in Heidegger.

Bugbee is far more personal than Cavell, at least in any familiar sense. The first stretch of The Inward Morning is halfway filled with Bugbee's sketches of three memories 'from experience long ago' that press upon him 'as if they bore the image of conclusive meaning which our situation may yield if only our mode of being be true': he presents them as if they were autobiographical sketches, titled 'Swamping', 'Building a Dam', 'Rowing'. His resumption of writing in July 1953 is presented in a similarly frankly personal way: 'Here I am in a situation about which I feel the need to set down a few words; it is the actual situation which I must somehow work through and beyond'. Then, his assessment of his situation as a writer, mentioned above. The writing and the work are inseparable: 'My task has been to learn to write in a vein compatible with what I can honestly say in the act of trying to discover what I must say'.

There are similar moments in Cavell's writing - and often right in front of you, liable to be overlooked - but when he makes a point of personal involvement it often seems meant to foreground his making of a personal claim, his act of asserting his involvement (so that you can see it's a thing to do), rather than anything like the experience on the basis of which he makes the claim. Experiences of this sort, when they make an appearance in his writing, often seem to border on the inconsequential, the contrived. Perhaps, though, he is sensitive to invoking experiences too eloquently, or experiences that call for too much eloquence, sublimity. As if powerful avowals were liable to thwart his aims.

When the personal, when experience, is confined to the proem or a bit more license to stray is only taken near the close of a text, decorum seems to win out over the self.

Dorothy Wordsworth begins the Grasmere Journal out of resolve 'to write a journal of the time till W[illiam] & J[ohn] return'. (They are off to Yorkshire. They return about a month later. Dorothy's expectation of their return extends uncertainly over days.)

Apparently, one reason Dorothy keeps her journal is that 'I shall give Wm Pleasure by it when he comes home again'. The editor notes that 'even Coleridge seems not to be thought of as a reader of the diary'. She speculates (on what basis I don't know):

'We are privileged, by reading this private diary, to know something of the daily life of a poet, though not from his own point of view and not with a central light focused upon him. The diary was for him, and thus is not primarily about him. Dorothy was not a Boswell recording a hero. When she stopped to recollect details, to write an extended and careful description, to re-read and improve her prose, this must partly be because she was offering for Wordsworth's consideration selected items of their common world. He might have forgotten, had not Dorothy's prose taught him to see again, the leech-gatherer or the shore of daffodils. Wordsworth has paid tribute to Dorothy's power to make him see and hear, 'She gave me eyes, she gave me ears.' The gift was perhaps mutual.'

Of course, Thoreau begins his when Emerson asks him if he keeps one.

Rilke's fictional journal-keeper doesn't know why he begins to keep his journal; or at least not completely.

Newly arrived in Paris, Malte writes two entries just detailing his impressions of place before he comes to address why he is writing—

'I am learning to see. I don't know why, everything penetrates me more deeply, and doesn't stop at the place where it always used to end. There is a place in me I knew nothing about. Everything goes there now. I don't know what goes on there.'

—and the thought seems to be that he has to write to find out what goes on there, what goes on inside him, with the new emergence of this depth to him.

He goes on to deny any point to writing to others, i.e., for others. 'Why should I tell someone that I am changing?' So he chooses a form in which one writes for oneself.

What makes a writer begin to keep a journal?

In the first entry of The Inward Morning, Tuesday, August 26, 1952, Henry Bugbee writes, 'I have yet to discover how to say what moves me to the endless search and research, the reflective turning over in my mind of experience'. The day's entry ends:

'I feel for me a crisis is at hand. I look back over my writing and I discern hints of what I can genuinely say, but undeveloped for lack of riding them through as they come to me. I look back over the last four years of work, and I feel dismayed that so little is set forth, and shown for what it has been. What is needed, I have concluded, is a record based on just one principle: Get it down. Get down as far as possible the minute inflections of day to day thought. Get down the key ideas as they occur. Don't worry about whether it will come to something finished. Don't give it up when faced with the evidence of miscarried thought. Write on, not over again. Let it flow. Don't haggle with the naturalists. Don't be stopping to jam the idea down someone's throat. Give it a chance. If there can be concrete philosophy, give it a chance. Let one perception move instantly on another. Where they come from is to be trusted. Unless this is so, after all is said and done, philosophy is arbitrary and idle.'

In the preface, a few years after the journal had ended, Bugbee meets the reader in an apologetic mode, apologetic about the book's being a journal:

'A life's work takes shape slowly. There is a periodicity about it. At intervals of years there comes a real show-down. Then one discovers, within the scope of his powers at the time, what he has been about.

These pages represent such a period for me. Since they rounded out in the fall of 1953, the period which they represent has assumed its place more unmistakably in the rhythm of years. I could not then appreciate as I have come to do that they would have to stand or fall pretty much as they were written, and why this should be so.

In the first place, I have tried to rework the material here presented. I could see that it left much to be desired. After I had culled out all that seemed irrelevant to the basic undertaking, there were still 'the good days' and 'the bad days.' I could imagine many questions which might be justly raised about my meaning, when it remained far from clear in my own mind. And since the themes which occupied me in these pages undergo as much reflective analysis as they do, why not organize at least some of the ideas which are recurrently developed into a more systematic form? Why not cut free from the bad days and supply a better substitute for the continuity of those days in the overall task? Over and over I have tried to act on such considerations as these. And each time the heart went out of the ideas themselves. They lost their actual exploratory cast. I found I was in danger of betraying the very undertaking in which experience yielded them their measure of meaning and support. Finally, I have come closest to establishing continuity with the work of these pages in the interim when I have forgotten about them, when I have worked from where I am—as I did in writing them.

I can therefore say that subsequent trial has confirmed the intimations on which I resolved to act from the outset of this work. As I would put it now, the guidance of meditation, of the themes received in meditation, is the fundamental feature of the work; and the themes of meditation live a life of their own, perhaps wiser than one knows in their advent and departure, in the things they gather to themselves as relevant to their formation, in the memories with which they visit one and establish their own concrete meaning. It was my work to attend upon such themes, in the very rhythm of daily life; to follow them where they might lead; not to put them off when they came to me, not to bid them stay beyond their actual departure; and not to try to make more of them than I presently could.

The present day—that is the dwelling of meditative thought. Consequently this work is in journal form. Not because it is a philosophical notebook or diary; it is neither of these. It is basically a work which required to be done within the day, from the actual human stance which they day might afford, whatever that day might bring.'

What does a writer keep a journal for?

The translator of Musil's Diaries offers more than one interpretation:

'One has the impression that Musil felt that all these authors failed to put in the preparatory studies, the intense observation of human beings, the background training of mind and pen, the deskwork, the back- and mind-breaking effort that was the cross a creative writer ought to bear in order to achieve the standards demanded by Geist—that mental-spiritual continuum in which the greatest writing took shape. Vital to such preparatory studies was the regular work in the Diaries.'

Another follows a description of Musil's slow pace in writing and meticulousness in (extensive, repeated) revision, and a description of how he was 'unusually sensitive to the eyes of others resting on him':

'Musil appears to have felt that writing for publication was worse even than exposing face, clothing, posture to a cameraman. It was exposure of a more intimate kind—of the mind at work. In the Diaries, Musil works not for the public but for himself, his critical threshold is lowered, he writes fluently, spontaneously. He is no longer the buttoned-up, tautly organized author of essays, or reviews, or chapters of the novel, hyperaware that the eyes of the Viennese or Berlin reading public are upon him; this is Musil in relaxed, private mode—he quite often makes mistakes, misspells, gets names wrong; he occasionally lets himself go to the extent that he produces misshapen sentences, his syntax is awkward, even ungrammatical—in short, he is reassuringly human when no one is looking!'

'... in small works there is no point in having a mix.'