Ordinary language is all right.
One could divide humanity into two classes:
those who master a metaphor, and those who hold by a formula.
Those with a bent for both are too few, they do not comprise a class.
Street clothes, that's my favorite kind of clothes.
(I added the exclamation point myself, that's right, I do what I want.)
Genette says that 'the paratext is what enables a text to become a book and to be offered as such to its readers and, more generally, to the public'. So, naturally enough, it would follow that a form mostly lacking paratext has not been made so as to be offered to the public; something about it remains private, perhaps even despite publication. What Genette says about prefatory material seems to apply to most other paratextual elements as well: the material's main functions are to get a book read, and to get it read properly. This suggests that what 'enables a text to become a book', what converts it from something private to something public, somehow casts its contents into readable form, where the term 'readable' covers a whole range of factors from those by which a book attracts readers (title and genre indications which help readers identify reading material of interest, back-cover blurbs that tantalize or quote testimony of reliable readers or renowned authors), to those by which prospective readers are first introduced to the author and his or her purposes (prefaces, forewords, and introductions especially, which Genette sees as occasions for feats of persuasion placing a high value on the text without 'antagonizing the reader by too modestly, or simply too obviously, putting a high value on the text's author'), to those by which the reader's progress through the text is managed (I would count intertitles here, though that way of conceiving of their function in a text, as titles of parts, downplays the greater importance of the structure they help indicate, especially in what Genette calls 'didactic texts', about which he says only a tiny bit) and, of course, effected (the 'text itself', which it seems shouldn't be denied a role in the reader's reading—consider, after all, some of the famous limiting-case texts which are generally regarded as inherently unreadable).
An unpublished journal (one written for private use, with no thought of publication) is not likely to contain any of that. A published journal is likely to contain, at most (excluding non-authorial paratextual material, like Searls' introduction for his abridgement of Thoreau's journal), dates as headings for the entries (perhaps a little regularized, as compared to unpublished journals, as a sensible concession to readers when the original entry dates are incorrect or erratically maintained). But nothing else. (OK, a generic title too: 'Journal'.)
What does all the paratextual apparatus give the reader, help the reader to do, that is simply not done (or not doable? or without a point? or just harder?) when reading books which lack most of that apparatus, like journals? In what way is what the reader does connected to the (paratextually-constituted) book's being public, and the journal's being private?
About the use of descriptive intertitles, delayed for ancient historical texts until the late fifteenth- or sixteenth-century editions of medieval chroniclers, Genette says: 'the titles in these editions are synopses in the indirect style, noun clauses introduced with "How…" or complements introduced with "About…"' (Paratexts 309). The synoptic function is one that aphorisms' titles seem hard-pressed to serve, probably given the degree of compression already at work in their texts. But its other side is the titles' use in something like a table of contents or a plan communicated in-text: with no effective way to refer to each of a text's individual parts synoptically, it seems difficult to coordinate their relation to one another in anything like an overview, or synopsis of the whole.
Distinguishing between thematic, rhematic (i.e., designating their portion of text formally, as it were as an object, e.g. in the case of entire collections of each kind, 'Odes', 'Tales', 'Essays', 'Dialogues', etc.), and mixed intertitles (i.e. titles of parts of a text, like episodes or chapters or poems or whatever), Genette notes that 'thematic intertitles not preceded by rhematic indications of the "Chapter Number Thus-and-Such" type are in reality very rare in any period, perhaps because without such indications the narrative text could easily be taken for a collection of separate novellas' (Paratexts 298n4). That issue of reader uncertainty is connected in complicated ways to the kind of text the whole text is, and to its mode of discourse. Discussing cases of the absence of intertitles, Genette says:
'The intertitle is the title of a section of a book: in unitary texts, these sections may be parts, chapters, or paragraphs; in collections, they may be constituent poems, novellas, or essays. it follows, therefore, that a completely unitary text—that is, an undivided one—can contain no intertitle. To the best of my knowledge this is the case, for example, of most medieval epics, at least in the state in which they have come down to us, but it is also the case of some modern novels, such as Sollers' H and Paradis. It would be tempting to say the same of Ulysses, but as we know, its situation is a bit more subtle. Conversely, some texts are apparently too highly segmented—I mean chopped too fine—for each section to bear its own intertitle. This is the case for collections of fragments, aphorisms, thoughts, and other maxims when the author, like La Rochefoucauld, did not deem it necessary to group them—as La Bruyère would for his Caractères—into thematic subgroups forming chapters and warranting the assignment of an intertitle to each' (295).
Apparently, in narrative discourse which is not (formally) unitary but somehow segmented, a reader would be too likely to try to read a text with thematic intertitles, but lacking rhematic indications before each section, as somehow (narratively) dis-unified. So making its intertitles both thematic and rhematic is a way of helping the reader to read the book properly, avoiding unnecessary confusion (and work).
It's not clear how well Genette's remark about fragmentary (etc.) works being too segmented to bear thematic intertitles holds up, though. We know there are works like Wittgenstein's Investigations that readily invite coherent readings despite superficial fragmentation, presumably because of the care Wittgenstein took in arranging remarks. (In Strawson's early review he notes: 'Wittgenstein did not gloss his thoughts; but he arranged them'.) And even work like La Rochefoucauld's, at a further extreme of compression and clear lack of discursive continuity, with a preponderance of single-sentence remarks, apparently didn't enjoy the effects of extreme segmentation automatically, as a kind of consequence of form: Blackmore, Blackmore, and Giguère note in their OUP edition of the Maxims that he presents his views
'as a series of detached reflections, a "heap of diverse thoughts"… the truth and validity of which are to be weighed individually, without reference to any previously established line of argument or any system of philosophy—whether or not such a system may have originally generated them. And in practice this is how most of La Rochefoucauld's correspondents and other contemporary readers (such as Queen Christina) responded to them, accepting one maxim and rejecting another, without feeling that the book's items were logically interdependent and must stand or fall together. The 1664 'Note to the Reader' gave haste as the main reason for the lack of order, but in subsequent editions this argument was abandoned, and, far from seeking a more orderly presentation, La Rochefoucauld strove to heighten the appearance of discontinuity: maxims were progressively abridged, removing supportive arguments and leaving them more terse, more naked and exposed; sequences of maxims on similar topics were removed or disrupted' (Collected Maxims, pp. xxvii–xxviii).
—which at least suggests the possibility that titles could have been omitted not just out of formal necessity (at a level of segmentation that would not bear intertitles) but avoided so as to achieve particular effects.
What effects, exactly, is hard to say. But we know from Nietzsche's aphoristic works that intertitles (non-rhematic, or more than just rhematic—in all modern cases bare numbering seems to be supplied pretty universally as a matter of convenience, with little worry of falsely implying sequence or system) are possible. In every publication from Human, All Too Human I through The Gay Science—five books, counting the supplements to Human separately—Nietzsche titles every single aphorism (with Human the exception, with a few chapters ending with free continuations of a titled aphorism into subsequent, untitled sections to form miniature essays), with thematic grouping in Human (titled, and perhaps discursively arranged, chapters of aphorisms), and merely rhematic grouping (or not definitively and explicitly thematic, given some obvious patterns) in the others (numbered 'books' of aphorisms in Dawn and The Gay Science, save for the title of Book 4 of the latter, and the title of the later expansion by Book 5). That consistency is broken in later works, in which Nietzsche's intertitling practices get more complex (in interaction with different generic and rhetorical aims), but in those works even where one could have expected the most consistency with previous practice, titles are not given: most (but not all) of the sentence-long 'Epigrams and entr'actes' of Part 4 of Beyond Good and Evil are, like the rest of that book's numbered sections, not given titles although the titles of its higher-level parts seem newly intended to suggest a continuous argument and not just thematic groupings.
Perhaps that reinforces a suggestion of unity or continuity to those aphorisms, contra the formal implication of disunity which was reinforced by the deliberate intertitling of aphorisms in earlier works.
Again, to what effects? Although the uniform use of intertitles can certainly be used for rhetorical effects—Adorno's very Nietzschean Minima Moralia shows a fondness for ironic negation of commonplaces or other sorts of formulaic expressions in the titles of its essayistic aphorisms, which are often even revisited for emphasis in the aphorisms' conclusions—Nietzsche often seems content to identify the subject or broad point of an aphorism in its title, without necessarily using the opportunity to inflect it with his authorial point of view. That's not the whole story, because seemingly neutral titles often precede remarks which could, I guess, modulate the reader's attitude toward the subject under discussion, or directly imply that Nietzsche's attitude toward it is not a conventional one. But the thing about which this attitude is communicated is generally the same one conventionally referred to by the title phrase, and its manner of doing so is usually pretty direct.
Personally, I've never been a very good reader of Nietzsche's titles; I tend to ignore them when reading, and not to try to incorporate them into my readings when I get down to the work of interpretation. (One of my students was once very exercised by my neglect, because he always took care to try to figure out how the aphorism went with its title, and found some that weren't so obvious.) Perhaps that's out of a sense of redundancy, of a lack of functionality: they often just seem to duplicate the identification of an aphorism's subject that is already accomplished by the text of the aphorism, and which has to be grasped anyway if what the aphorism says about the subject is to be understood.
But that much is true of La Rochefoucauld's maxims as well, where the consistently extreme brevity might make it seem more as if titling would be utterly redundant. Imagine trying to title these maxims (V:30–36):
We are more able than willing; often we imagine that things are impossible because we want to excuse ourselves in our own eyes.
If we had no faults, we would not derive so much pleasure from noting those of other people.
Jealousy is sustained by doubt; and it either becomes a frenzy or comes to an end, as soon as we pass from doubt to certainty.
Pride always finds some compensation for everything; even when it relinquishes vanity, it does not lose anything.
If we had no pride, we would not complain of it in other people.
All men are equal in pride; the only difference is in the ways and means by which it is brought to light.
It seems that nature, which has so wisely arranged the organs of our body for our happiness, has also given us pride to spare us the pain of knowing our deficiencies.
Clearly a consistently thematic arrangement (which, remember, La Rouchefoucauld worked to disrupt) would deprive him of the opportunity to make a variety of transitions. But some level of thematic continuity lets him talk about indirectly about pride without saying 'pride', or lets him say unexpectedly different things about pride by deferring references to it (a very French-Classicist move, stylistically, I suppose). But at this level of compression, and without explicit (say, titled) thematic grouping, what choice of titles would there be? Redundant and distracting ones like 'Pride', obviously. Or more contentful ones which would amount to commentary on the maxims they titled, thus possibly (a) associating some view of the matter with the author, the one conventionally assigned ownership or responsibility for this paratextual material, or (b) obtruding a view assignable to the author in between the reader and a maxim carefully contrived to, let's say, corner the reader, per the 1664 'Note to the Reader': 'The reader's best policy is to start with the premiss that none of these maxims is directed specifically at him, and that he is the sole exception to them, even though they seem to be generally applicable. After that, I guarantee that he will be the first to subscribe to them'.
But if La Rochefoucauld's lack of titles effects a sort of transparency in remarks which are directed at the reader in a systematically ambiguous way, so that what the reader must think about is (what he thinks about) the subject, the phenomenon, in actuality (in his life and in the lives of others), perhaps what Nietzsche's relatively neutral titles do is to suggest the existence of these phenomena as objectifiable subjects, as potential matters for study, with some degree of detachment despite their (acknowledged) entanglement in our lives. This wouldn't just be a matter of objectivity about morality, in the face of prejudices declaring its examination as a phenomenon off-limits (a theme Nietzsche hits pretty regularly in this period). It's also a matter of scope, or method: Nietzsche wants to claim that even minor subjects can be philosophically revealing, particularly when studied 'psychologically' in an approach inspired by the French maximists. But that requires winning those subjects standing, getting them to count as phenomena worthy of reflection—say, by putting every single topic of discussion on the same level, through its appearance in the title of an aphorism. Nietzsche's titles become ways of making his subjects independent of what he happens in each case to say about them (so that, say, he can change his mind, or leave room for observations by others). When titles disappear in later works, or rather start to move around in order to serve more plainly argumentative functions, that seems to be because Nietzsche is newly interested in making arguments, in persuading and provoking, and in (risking, and negotiating, and avoiding) being identified with a point of view.