A Brief History of Paragraphs and Punctuation

Punctuation began with the paragraph. “Since the second century BC, the basic unit in a western text has been the paragraph or capitulum,” M. B. Parkes declares. “This identifies a principal topic in a text, or a point of focus in an argument or narrative” (65). At first, the para-graph meant a “mark beside” the text, as its prefix indicates. The practice of writing a continuous block of letters without spaces, known as scriptio continua, dominated in Rome and Europe from the second to the sixth century CE. Because orators memorized the content and rhetorical structure of discourse, writing served primarily as a record. Precious parchment was filled with ink. Servants deciphered script as opaque as computer code and read aloud to learned listeners. A Latin K marked the kaput (“head”) of a section in the margin; it evolved into the pilcrow (¶). The word chapter also derives from the mark for a capitulum (“little head”).

Across the next thousand years punctuation developed as Europe’s republic of letters evolved. The slow deaths of classical languages left ecclesiastical lectors in need of guidance, so systems of glyphs arose to illustrate rhetorical parts and grammar. Signposts for section, entry, or commenter helped seminarians navigate through the Bible or glossaries; encyclopedias would later adopt the same paragraph-based order (Châtillon 28). Meanwhile, gaps appeared between words to aid silent reading. Paul Saenger details how blank space quickens visual perception by clearing the ground on which textual units figure. In learning to read silently, the brain effectively automates recognition of words and sentences as units of information. Punctuation accompanied the shift from reading aloud to reading silently.Footnote ⁴

Indentation took hold with paper and printing. Notable letters that announce new chapters required different graphic techniques than regular letters, so a blank space would be left in the first pass, to be filled in later. But blankness alerts the eyes well enough without needing the fancy letters filled in, and the adoption of paper favored this economical alternative (Houston 16). Paragraph indentations cue the mind to close the preceding chunk of information while cueing the eyes to drop down and open the next part like a typewriter’s ratchet mechanism.

Seventeenth- and eighteenth-century French grammarians established conventions to maximize clarity. Gilles Siouffi explains how the language’s basic unit slowly changed from the period—a rhetorical unit of thought from which the mark takes its name—to a sentence ruled by grammar. At this point, sentence punctuation could be distinguished from blank spaces and other marks that arrange text on the page. Nineteenth-century publishers and the printing profession would enforce these new norms, as Annette Lorenceau recounts.

In the hands of the genius who bucked convention, however, punctuation became a unique element of style. In 1871 George Sand declared, “la ponctuation est le style bien compris” ‘punctuation is style well understood’ (Impressions et souvenirs 91):

Sand links expressive punctuation to freedom of thought and to the speaker’s body. Although Flaubert disagrees with her on a point that I examine in the next section, both profess pure style as the arbiter of punctuation usage, against rote convention.

The paragraph emerged as a tool for creative fragmentation with the modern genres of the novel and the prose poem. The novel’s “broad and elastic form” cultivated experimentation, especially in France (Auerbach 491). Romanticists led by Charles Nodier, Stendhal, Victor Hugo, and Honoré de Balzac availed themselves of irregular paragraph lengths to inscribe “natural” rhythms, as Jacques Dürrenmatt explains (183). Jacques Rancière agrees that the “genreless genre” of the novel cultivated a variety of styles that “merged with life itself” (14, 15). As a flexible container, the novel conglomerates rhetorical structures. In tandem, the prose poem emerged in the 1840s as fragmented structures on a smaller scale. New prose genres thus gave paragraphs license to participate in style.

Paragraphing styles also followed national trends. In his entry “Alinéa” ‘Indentation’ for the 1886 Grande Encyclopédie, Adhémard Leclère criticized fellow French writers for fragmenting their prose with too many small paragraphs; some even indented after semicolons! He cautioned that such short, “rude” phrases made the style too light for the reader to grasp the unity of thought or to appreciate stylistic contours. Sand likewise deprecates excessive commas, exemplified by Jules Michelet’s “très-haché” ‘chopped up’ style (Sand, Impressions et souvenirs 99). At the other end of the spectrum, Leclère chides German, English, and American writers whose paragraphs extend for page after page like a soporific medieval forest. (Henry James would surely have taken umbrage at that remark.)

In the shadow of strong norms that emerged for sentence punctuation, the paragraph found a new freedom, especially in the hands of the creative stylist. I now turn to Flaubert, whose engagement with Sand demonstrates how adherents to the doctrine of pure style nevertheless disagreed on specific usage and on certain general ideas.

Flaubert and Pure Style

Flaubert developed a conception of style that is not personal or conventional but purely aesthetic. The work of writing therefore shifted from the easy élan of genius and the divine to a difficult “travail” of scrupulous revision, driven by a devotion “presque sacrée” ‘almost sacred’ to art for art’s sake (Philippot 8, 97). The young Flaubert hatched the ideas for Madame Bovary and Bouvard and Pécuchet in the 1840s but did not publish the former until 1856 or undertake the latter until the 1870s. His assiduous revisions aim at a lucid yet indeterminate movement that agrees with Kant’s view of the aesthetic, as I discuss below.

Flaubert’s punctuation inherits the idiosyncrasy and virtuosity of Hugo and Chateaubriand. Anne Herschberg-Pierrot describes Flaubert’s unconventional use of semicolons, ampersands, and commas as a “private” system because its logic remains opaque even to specialist scholars (par. 2). With regard to paragraphing, one does not detect any system at all. Peter-Michael Wetherill characterizes the intervals across his oeuvre as “vagues, flexibles” ‘vague, flexible’ (79). For instance, the second, 1874 edition of A Sentimental Education favors longer paragraphs; meanwhile, Flaubert was composing the short paragraphs of Bouvard and Pécuchet. This variability agrees with Sand’s view that style is “élastique” ‘elastic’ and knows no absolute rules (Impressions et souvenirs 91).

However, Flaubert disagrees with her view of punctation as an expression of personal voice. That equation leads Sand to criticize excessive punctuation as the mark of a man who “s’aime trop lui-même” ‘thinks too much of himself’ (Impressions et souvenirs 93; Impressions and Reminiscences 77). Flaubert evidently felt targeted by her remark, as he reveals in a 31 May 1873 letter: “Je ne suis pas complètement de votre avis sur la ponctuation. C’est-à-dire que j’ai là-dessus l’exagération qui vous choque. Et je ne manque, bien entendu, de bonnes raisons pour la défendre” ‘I don’t fully share your views about punctuation: that is, mine exhibits the very excesses you disapprove of. And needless to say I can adduce good reasons for defending them.’ (Correspondance 4: 669; Steegmuller and Bray 315). Flaubert never states his good reasons for immodest punctuation, but his elevation of style over subject provides a clue.

Sand elsewhere criticizes excessive punctuation on the classical ground that a text’s parts should be subordinate to the whole: “Si, même avec beaucoup d’habileté, vous insistez à égale valeur sur tout ce qui peut former une sorte d’angle d’incidence, vous détruisez l’effet principal au profit d’effets secondaires” ‘If, even with much ability, you insist on allowing an equal value to every thing that can form, as it were, an angle of incidence, you destroy the primary effect for the benefit of secondary effects’ (Impressions et souvenirs 95; Impressions and Reminiscences 79). Too many marks fragment the logical unity of the period of thought. Sand disparages the vain man who “pèse chacune de ses paroles” ‘weighs every word’ and thereby fails to communicate clearly (93; 77).

But Flaubert does weigh every word and even every comma, for the good reason that every detail matters. “Each part of a composition is as important as the whole,” observes Stein of Flaubert and Paul Cézanne; with a Kantian turn of phrase, she says each part becomes “an end in itself” (“Transatlantic Interview” 15).Footnote ⁵ In elevating style over subject, Flaubert suspends the classical hierarchy of elements such that color and rhythm may carry as much weight as plot or character.

Consider the 29 January 1853 letter to Louise Colet in which Flaubert discusses paragraph revisions for Madame Bovary:

Note how the cascading flow between paragraphs that Flaubert seeks is reciprocal, and therefore not merely plot’s downhill bent. Perfect structure inhibits a freer, vital movement. Where Sand bemoans dense details that distract from higher thought, Flaubert seeks a unity that arises from fine detail as much as it derives from structure.

Kant’s distinction between determinate judgment and aesthetic judgment can clarify the opposing views of Sand and Flaubert. The subordination of details to higher thought and subject corresponds to determinate judgment; meanwhile the reflective judgment of the aesthetic discovers unity through play. “The judgement of taste is quite independent of the concept of perfection,” Kant explains, because true perfection involves the object’s conformity to a purpose (Kant’s Critique, sec. 15). The disinterested play of imagination prefers the imperfectly “wild beauty” of nature, whose harmony with the understanding happens freely and subjectively (sec. 22). Kant’s description of the sublime as an effervescent and variegated “Bewegung des Gemüts” ‘movement of the mind’ rhymes with the movement of style that Flaubert seeks (Kritik, sec. 24; Kant’s Critique). From this point of view, Flaubert does not aim to disrupt all thought and logical coherence; rather, he favors the subtle harmony of the aesthetic over the conventional order of understanding.

Sand and Flaubert agree that punctuation and style rise above the grammarian’s brute rules. Sand nevertheless reverts to classical conceptions of organization through subject and authorial voice. On the one hand, she recommends that punctuation serve the main subject of discourse without distraction. On the other hand, she locates absolute authority for punctuation in the body of the poet. This latter view seems to agree with Szendy and Serça that aesthetic power originates outside language. Meanwhile, Flaubert develops an alternative vision of language and literature that has a capacity for harmonious play within its excessive forms. I find that Kant’s approach to the aesthetic as play guided by the imagination and unruled by concepts corresponds to Flaubert’s pure style.

Given that the wildness of play is always particular, let me turn to an example. Thereafter I comment on the role that paragraphs have in setting the conditions for the conventional order of representation or, alternatively, for play.

Genre Play

I must refrain from introducing Bouvard and Pécuchet, because it is precisely the ambiguity of genre in the opening passage that interests me:

This description could be compressed seamlessly into one paragraph (as some translations unfortunately do). But the four tableaux invoke the unmistakable sequence of void, waters, darkness, and light that opens Genesis:

Flaubert parallels the ancient text with an image of empty heat, followed by the canal’s black pool, then a blindingly bright sky cut by Paris’s stony skyline, and two men who appear from nowhere. His short paragraphs also imitate the prose-poetic verses of the vernacular Bible, while the holy book’s high register is echoed in “Au-delà” ‘Beyond’ and “grand ciel pur” ‘great pure sky.’

No literary gesture is more grandiose. Hegel agrees with Longinus and Nicolas Boileau that Genesis verse 3, “Let there be light: and there was light,” epitomizes the sublime; Flaubert duly noted Hegel’s explanation when he read Lectures on Aesthetics in 1844 (Flaubert, “Notes” 293). For the Christian philosopher, just as God’s divine will and its infinite awesomeness stand apart from the finite natural world (and from primitive polytheism), the sublime interrupts the immediacy of primitive symbols through reflection that compares the idea and the sensible form that represents it.Footnote ⁶

Flaubert’s pastiche invites reflection, but some might find his invocation of Genesis profane. The opening passage depicts the inertness of a sweltering summer afternoon more than it heralds a divine mover. In lieu of a creator, things act autonomously here: the sky “se découpait en plaques”—as if the sky fragmented itself into metal—while the built surfaces “éblouissaient,” or glowed, as if emitting light themselves. The industrial setting of canal, barges, dirty water, and hot stone is far from pastoral or heroic. The scene culminates in nothing more than two clerks who exchange pleasantries, establish a friendship, and commence a series of pratfalls that have all the “predictability of a Tom and Jerry cartoon,” as Dennis Porter puts it (376). And while Bouvard and Pécuchet’s ignorance, earnestness, and awe befit Eden, the homosocial duo and Flaubert’s mocking exclamation marks do not. Such carnivalesque play recalls François Rabelais, and can even pose theological riddles: Is man’s stupidity also the image of God? Could God create boredom and meaninglessness?

The genre of Flaubert’s novel morphs because frameworks of high and low registers that typically guide interpretation float unmoored. Here diction conveys awe, there characters appear clownish. Flaubert cultivates the reflexivity characteristic of Hegel’s sublime but in an ironic mode.

A second object of pastiche—realist and naturalist literature—further complicates genre in the passage. As setting, wharves fit the program of Émile Zola’s experimental novel: to observe the effects of environment on behavior (especially of the lower classes) in the laboratory of fiction. The precise air temperature and “absolutely” deserted street also indicate conditions for a scientific experiment. However, Georges Kliebenstein points out that a proper experiment would note “degrees Celsius” as the unit (452). Indeed, only the physics of the tongue seem to account for the specification of “thirty-three” degrees and the equally alliterative “Boulevard Bourdon.” Sensual prosody thus undermines the scientific sobriety of naturalism.

Do these details instead produce what Roland Barthes criticized as an “effet de réel” ‘reality effect’ (88)? Barthes argues that superfluous details in Flaubert’s descriptions exemplify the illusory power of representation. Echoing Hegel’s criticism of the primitive symbol, Barthes suggests that realist literature veils the medium of communication in the same way that ideology and power deceptively veil themselves. He praises classical genres such as heroic elegies or ekphrasis that showcase descriptive language for purely aesthetic ends.Footnote ⁷ Barthes would say that Flaubert chose Boulevard Bourdon to convey the patina of reality. Ferdinand Bourdon, a colonel killed at Austerlitz, has never been a household name (and still awaits a Wikipedia page). Balzac would have instead set the action on the opposite side of the canal, on the iconically named Boulevard Bastille. It does seem that details in this passage merely perform the actual stuff of a Sunday afternoon in Paris, to an extent.

Yet Flaubert’s rich detail can be more generously read in the tradition of classical ekphrasis that self-consciously showcases language and writing. The second paragraph reflects on the medium of writing in describing the canal water as “black as ink.” Flaubert’s clever variation of Homer’s wine-dark sea suggests the deathly stillness of print, and a netherworld of total darkness is further evoked by the transitional phrase “plus bas” ‘farther down,’ which tracks the eye’s journey down the page. Timber and barrels hint at pens, paper, and the rotary press cylinders of a culture industry then emerging. While Genesis describes nature as a creation of the divine word, Flaubert suggests that the buildings, canal, and street are a kind of human writing upon the earth as so many letters, lines, and spaces upon the page. Queer comma usage—missing at two places in the first sentence, then excessive in the second—seems to mimic the artificial flow and arrest of water in canal locks, as do the short paragraphs. The next tableau, of the bright sky shining through houses, also mimics the white page between words and letters. Hard consonants seem to chisel stone structures on the tongue as if in reply to the first paragraph’s gratuitous alliteration. The repetitions in thirty-three degrees, Boulevard Bourdon, and a double row of barrels could be metonyms for the endless copying that the whole novel figures.Footnote ⁸ In contrast to the sober aim and polished language of the realist novel that Barthes disparages, Flaubert’s virtuosic description soars in the ether of contemplation.

But even the languor of ekphrastic excess does not endure. A three-word fourth paragraph, “Deux hommes parurent,” brusquely changes rhythm. Bouvard and Pécuchet arrive onstage with an unexpectedly dramatic flourish. Pierre Cogny detects here a parody of popular novels: “un bon roman noir ne serait pas introduit différemment” ‘a good crime novel wouldn’t start any differently’ (41). Flaubert introduces yet another contrast in genre at the moment the novel moves from description to action. The discontinuity is stark and intentional.

In reading the passage I have discovered a handful of competing genres: ancient myth, naturalism, ekphrasis, and pulp fiction. The diversity of meaning that Flaubert’s text yields is not an accident, Steve Murphy observes, but “programmé” ‘planned’ (80). Meticulous drafts make clear that Flaubert aims at exquisitely endless contradiction. Irony and satire arise from the storm of interpretive logics, not from a cynical narrator. By favoring multiple possibilities over one dominant genre, the play of the text and its uncanny harmony exemplify Kant’s view of the beautiful as an indeterminate judgment.

To evaluate the part that paragraphing has in this particular gem and in literary aesthetics generally—my initial question, whether paragraph divisions are accessories or primary elements in the work—I now introduce a formal scheme and then return to analyze the passage.

Formal Analysis of Paragraph Division

Writing handbooks express perfectly the purpose of the paragraph: in argumentation, paragraphs differentiate logical parts; in stories, they structure plot into scenes and into discrete actions within scenes, such as changes in speaker within a conversation.Footnote ⁹ Yet the conjunctions in this seemingly simple description already reveal that paragraph divisions can perform multiple rhetorical functions, despite the absence of any inscription upon their blank matter. Indeed, differences between scenes or parts of an argument do not dissolve if indentations are omitted. That multiplicity and redundancy suggest that paragraph divisions (and punctuation) are material accessories to relations inherent in the text.

In fact, paragraph names multiple objects: the blank indentation and the block of text between the breaks. In turn, the block of text does not simply stand alone. Indentation simultaneously circumscribes one holistic block, creates a difference from its neighbors, and indicates the greater ensemble to which these rhetorical units then belong (scene, act, and chapter, for instance). The negative space of blank indentations cues the reader to organize the text in all these ways at once.

To reiterate, paragraph divisions can be said to have three concurrent functions within reading: to indicate autonomy, contrast, and pattern. These modes are ordinal, not exclusive (either/or) types, meaning that the simple relation can integrate into more complex relations. That ambiguity affords opportunity for play. Here I see a logical triad, similar to the triadic pattern that Peirce derives from Kant and Aristotle, of how entities can relate (Peirce 1: 148). I propose this formal scheme for analyzing the functions that paragraphs have in normal narration and the subtle alternatives that certain cases discover.

Consider again the opening passage from Bouvard and Pécuchet. Pray read it in continuous script to observe the mind organizing the text into parts, unassisted by punctuation:

This exercise gives a general impression of how punctuation and paragraphing guide reading. Blanks and marks distinguish words and paragraphs, through which syntax and the higher orders of narration become manifest. Those structures reinforce perception of parts, even retrospectively, as the telos of story or argument coalesces. The curiosity of this passage lies in the variety of part-whole relationships that emerges.

Description sets the scene before the heroes appear. Paragraph divisions create suspenseful beats that draw out observation, from distant overview to lower and then upper parts. Short paragraphs also establish a rhythm of narration that integrates the text on another level. Proust, in a less complimentary remark, once compared Flaubert’s characteristic “rythme obsesseur” ‘obsessing rhythm’ to the intermittent noise of a distant excavating machine (“À propos” 83). In a normative way, Flaubert’s paragraphs emphasize the patterns of drama and rhythm that organize the units of the story as a whole.

Yet these same paragraph divisions in Bouvard and Pécuchet can emphasize ways of reading that play against the normal order of the story. For example, indentation frames each paragraph as a picturesque tableau—an empty street, an idle canal, and a city skyline—that showcase the writer’s fine expression. The autonomy of paragraphs brings to mind Flaubert’s dream of a book composed solely of sentences, mentioned in his 25 June 1853 letter to Colet: “Je voudrais faire des livres où il n’y eût qu’à écrire des phrases, comme pour vivre il n’y a qu’à respirer de l’air” ‘I would like to produce books which would entail only the writing of sentences (if I may put it that way), just as in order to live it is enough to breathe’ (Correspondance 2: 362; Letters 189). A book composed of sentences would frame each one as its own paragraph and thus render a series of prose poems that showcase prosody and the picturesque in the ekphrastic tradition. Indeed, Michael Fried elucidates the author’s obsession with verbal expression in Flaubert’s “Gueuloir. ” Guy Larroux identifies similarly autonomous paragraphs in other Flaubert novels (476). Here division of the description into paragraphs is not necessary to action; their autonomy invites the reader to admire composition for its own sake.

A play of contrasting verb tenses between the third and fourth paragraphs also merits attention. The first three paragraphs employ the imperfect tense to give the impression of a timeless scene: “il faisait,” “il y avait,” and so on. Then, as if sounding a gong, a preterite verb brings the protagonists into existence: “deux hommes parurent” ‘two men appeared.’ Along with the sudden shift in tense, the rhythm also changes. Proust admired Flaubert’s employment of a brief paragraph followed by white space to create an “extraordinaire changement de vitesse” ‘extraordinary change of speed’: “La chose la plus belle de L’éducation sentimentale n’est pas une phrase, mais un blanc” ‘The most beautiful thing in A Sentimental Education is not a sentence, but a blank’ (“À propos” 84). In this case, the paragraph break amplifies abrupt changes in tense and rhythm and even draws attention to the artifices of the comic tale and the printed text.Footnote ¹⁰

Flaubert’s paragraphing also highlights his play with genre. Short paragraphs mimic ancient verse translated into vernacular prose but also allude to the list structure of encyclopedias. The picturesque tableaux underline the distant observation of the naturalist, as well as delimiting the stage of the classical poet. The liberal indentation of Flaubert brings out these generic patterns.

To review, the paragraph divisions in this passage facilitate uptake of narrative organization to an extent, but its paragraphs also emphasize prosody, the picturesque, and allusions to various genres. To judge the passage beautiful, one would find the harmony among these diverse meanings and their arrangement compelling.

In formal terms, aesthetic play could be called a degenerate game. Peirce describes free play as the degeneration of a higher-order relation (law) to simpler relations of contrast and similarity. “In order to discover something new, we must introduce chaos into order,” explains Johannes Ehrat in his overview of Peirce (105).Footnote ¹¹ Bernard Suits might dub this randomness an “unnecessary obstacle” that constitutes a game (41). As the reflexes of ordinary representation slacken, more possible relations emerge. Play involves seemingly irrational movement that challenges the intellect. But Genette traces a through line from play to reason when he states, “it is the effort we make to overcome a difficulty that makes us conscious of things” (Aesthetic Relation 180). In literature, play can prompt reflection on language, intertexts, or the medium of writing. Communication transforms into a beautiful game of making sense.

Another point I want to make with continuous script is that all the text’s normal and aesthetic relations are already present. This reading exercise confirms and extends to paragraphing the point made by Nina Catach that “la ponctuation, du fait de son caractère de redondance, se prête particulièrement bien aux effets littéraires, à la liberté stylistique” ‘punctuation, because of its redundant character, lends itself particularly well to literary effects, to stylistic freedom’ (17). Punctuation marks and blank spaces ornament the semiotic body of the text by quickening readers’ perception of its manifold parts and, in certain cases such as this one, of their beautifully complex relations. Any text needs a physical support, but the persistence of play in continuous script challenges the idea that literary aesthetics originate from the materiality of literature. (Exceptional works that radically foreground ink and page are perhaps better classified with plastic arts that take reading as their subject than with literature proper, if one is to differentiate the arts at all.)

Molecular Narration

My analysis of the opening passage bracketed the fundamental role of the narrator. Paragraph divisions and punctuation help establish the narrator as the explicit organizer of all the story’s parts by sharpening distinctions among the voices of characters and narrator. Marks that indicate quotation are normally accompanied by indentation before passing to another speaker, in English; in French, indentation led by a dash can also introduce a quotation. Unpunctuated text usually implies the direct voice of the narrator’s persona.

Flaubert’s so-called impersonal narrator has been widely recognized as a literary innovation. His narrator resists stable identification with the author, with a character, or with any personality as such. Yet it does not disappear entirely into the impassive voice of scientific observation or official report. Kliebenstein describes a position “mixte” ‘mixed’ between objective distance and the author’s subjective presence (457). Florence Pellegrini says that Flaubert’s narrator develops a “distance empathique” ‘empathetic distance’ that is at once radically critical and “éminemment commune, partagée” ‘eminently common, shared’ (par. 15). The “Nous” ‘We’ that famously opens Madame Bovary exemplifies the complexity of Flaubert’s narrator.

Jonathan Culler relates the impersonal narrator to aesthetic autonomy in The Uses of Uncertainty. If communication involves a sender and a receiver, Flaubert’s narrator renders that basic exchange uncertain by adopting the language of people from various positions. Shifting perspective creates “a text which stands alone, upheld not by its speaker but by the tensions among its styles” (78). Because communication has practical purpose, Kant would say that Flaubert’s uncertainty suspends that purpose while maintaining purposiveness.

The dialogue passage below illustrates how Flaubert’s paragraphing supports such an impersonal narrator. French punctuation marks have been preserved in order to observe their effect on perception. At the outset of chapter 3, Bouvard and Pécuchet have failed at farming and seen their home distillery go up in smoke. So the pair concludes that they must learn basic science. This scene satirizes a vague chemistry textbook and its doltish readers. Their struggle to understand how one entity can have distinct parts or various manifestations reflects the paragraph’s “molecular” division of the work of literature.

The end of this passage exemplifies the normal aim of distinguishing narrator from other characters: “— ‘Moi, non plus ! »’ disait Bouvard.” Indentation and dash indicate a change of speaker. The conventional appendage “he said” identifies the speaker as Bouvard. The exclamation mark and the French guillemet, with prominence and shoulder space that catch the reader’s eye, insulate character speech from the narrator’s voice. Furthermore, all of these devices reinforce the conventional order of representation in which Flaubert’s narrator mediates between the reader and the world of characters and action.

Interestingly, Flaubert does not conclude the scene with the conventional “il dit” (“he said”) but switches tense. Up until this point, all action has been narrated exclusively in the preterite, simple past tense. Instead of concluding the scene with “dit Bouvard” as one would expect, Flaubert writes “disait Bouvard” ‘Bouvard kept saying or would say.’ (D. F. Hannigan’s translation ignores the subtle difference.) The imperfect “disait” could mean that Bouvard repeats his line, that this scene repeats somehow, or that this type of scene was habitual. What first appeared as a linear sequence of unique moments now begins to blur with subsequent paragraphs that summarize Bouvard and Pécuchet reading other books. In making temporality ambiguous, the text mimics the ambiguity between acid and base that perplexes the characters. The passage could represent one instance or a repeated series, “according to circumstances.”

Vague temporality also implies the narrator’s subjective presence. Proust remarks that Flaubert’s use of the imperfect tense turns every action into an impression (“À propos” 75). If impressionist painting implies a watcher, Flaubert’s imperfect tense implies the memory of the narrator. In retrospect, this scene ceases to transparently observe a fact and becomes representative of a typical occurrence. Raymonde Debray-Genette agrees that Flaubert’s impressions collapse the familiar show/tell distinction, blending dialogue with narrative summary (42–43). By highlighting the contrast in verb tense, the division between paragraphs four and five amplifies the narrator’s subjective blur of action into impression.Footnote ¹²

Other paragraph divisions also play against conventional representation. The blank space between the first and second paragraphs represents the characters’ blank stares, as if so perplexed by Regnault’s ambiguous definition, “les corps simples sont peut-être composés,” that the pair is left speechless. Given that the first and second paragraphs in this passage both cite Regnault, the division is not necessary to distinguish voices; Hannigan’s translation even omits the break. This remarkable indentation represents speechlessness and thus dispels the normal invisibility of the physical page and print.

And given the seriality of indentation, the next paragraph division consequently represents another speechless pause in response to further equivocation by Regnault. The blank space that follows “« rien d’absolu »” ‘nothing absolute’ becomes an icon for nothingness. Flaubert not only mocks the physical sciences’ pretense to irrefutable truth. The dumb beat also gawks at Bouvard and Pécuchet, who only understand absolutes. If one pauses long enough in the blank space, one sees one’s own reflection as a reader blindly confident that I too have interpreted the text correctly and that I know better than these fools. As irony circles round, print’s authority is evacuated and reestablished. Indentation does seem to impose an effect here that is not inherent in the text, comparable to the radical visual play of Laurence Sterne, Stéphane Mallarmé, and Guillaume Apollinaire. Hugo also used the blank page to represent nonresponse, as Michel Sandras points out (109). By representing with blankness rather than with words, Flaubert throws into relief the conventional regime in which language communicates action.

Another aesthetic contribution of paragraph divisions in this passage also deserves mention. Punctuation and indentations normally clarify who speaks, but their excess in one place reveals the general ambiguity of literature. At first glance each speaker’s identity appears obvious: quotations from Regnault’s textbook, a third-person narrator in the third paragraph, Pécuchet reading in the fourth, and Bouvard concluding with his reaction. However, the second paragraph blends the voices of narrator and characters: “On les distingue en métalloïdes et en métaux, – différence qui n’a « rien d’absolu », dit l’auteur. . . .” The guillemets apparently contain Regnault’s text, but it is not clear if “dit l’auteur” is reported by the narrator or by a character reading aloud. The familiar register of On (“one”) indicates that Bouvard is reading aloud to Pécuchet or vice versa, but a lack of guillemets meanwhile indicates the narrator as speaker by default. The dash that separates the two halves of the sentence compounds the discrepancy. Does the dash indicate a change in the character speaking, or a shift from narrator to character? That scenario can be read multiple ways: characters begin the sentence and narrator takes over at the dash; or, conversely, the latter part quotes the characters reading aloud while the sentence begins by copying Regnault’s textbook, in which the indefinite pronoun may represent the pedagogical voice of French instruction manuals that read (as in cookbooks), “first one gathers the ingredients, then one mixes them in a bowl.” One can thus identify the voices of this paragraph’s parts in multiple ways, as if performing ironically the textbook’s subject of nondistinctions between metals and metalloids.

The ambiguity of the second paragraph also drifts back to the first, retrospectively. Does “« les corps simples sont peut-être composés »” quote Regnault verbatim, Bouvard reading aloud, or Pécuchet’s inner voice? Italics could have clarified matters, but Jean-Luc Seylaz notes that Flaubert renounced that typeface (30). The strong emphasis of the guillemets betrays an uncertain object of emphasis. Readers cannot be sure if the scene involves the characters speaking to each other aloud, if the narrator reports the inner voices of the characters who read, or a combination of the two. Hence the prosaic representation of dialogue in this scene reflects on the indeterminate boundaries between reader, character, and writer in the way that lyric poetry plays with the first-person pronoun I to represent writer, reader, and other subjects (epitomized by Walt Whitman in “Song of Myself”). Seylaz remarks that Flaubert made the French pronoun On the axis for this novel because it involves “glissements” ‘slippages’ from the personal to the impersonal (28). These slippages reflect how language complicates the boundaries between self and other through sharing the same tongue. The division between paragraphs one and two makes more slippery the distinctions between persons in communication.

In sum, the scene that opens chapter 3 is laced with features of style that diverge from typical narration: an imperfect tense shift, superfluous paragraph division, missing quotation marks, and an indefinite pronoun, On. These delicate gestures blur representation like the brushstrokes of an impressionist painting, and, in parallel, blur ontological distinctions among things, ideas, the self, and others. Such ambiguities weaken the order of the narrator and mimic the ambiguity of the molecule. This irony reflects on the human limit to know or represent things in themselves, a theme that recurs throughout the novel as the protagonists pursue and fail at an encyclopedic range of endeavors.

I also see an ironic reflection on the beautiful through figures of chemistry. Beauty has traditionally been associated with perfection, and perfection defined by unity in variety. In Metaphysics Aristotle states, “In the case of all things which have several parts and in which the whole is not, as it were, a mere heap, but the totality is something besides the parts, there is a cause of unity” (1045a9–11). The unity of different organs in the body commonly illustrates this perfection in nature. In satirizing the chemistry textbook’s uncertainty, Flaubert suggests that the classical vision of nature’s perfection breaks down at the molecular level. And yet, at the same time, ironic reflection on that imperfection creates a metaphysical unity, like a flower that blossoms from a grave.

The beauty of Bouvard and Pécuchet arises from ambiguity. I agree with Patrick M. Bray that the book is more than the manifesto of a skeptic who rejects all knowledge (18). Rather, it seeks a satirical sublime. For Kant, the sublime can occur only in nature because the experience overwhelms even the concept of an art object; however, he indicates that certain genres of beautiful poetry can present the sublime (Kant’s Critique, sec. 52). In this way Flaubert’s encyclopedic novel represents the sublime crisis of understanding. The certainty of human knowledge is suspended, through which arises an awesome demonstration of the imagination’s capacity for organic unity from no certain ground.

In his drafts for Bouvard and Pécuchet, Flaubert left a note instructing himself to “multiplier les attaches et les suspensions” ‘multiply the attachments and suspensions,’ as Jacques Neefs highlights in his preface to Yvan Leclerc’s study of the novel (3). The two excerpts that I examined reveal rich networks for free association within a conventional story. The opening passage evokes various genres while the reading scene in chapter 3 blurs time and voices into an impression. These examples support my focus on play as a central element of the aesthetic that happens within language, not against it.

Paragraph division has a minor but illuminating role in Flaubert’s style and in literature generally. At bottom, the blank space of indentation is a dumb ornament that emphasizes rhetorical joints within the text that are complex. The ordinal schema of autonomy, contrast, and pattern that I develop from Peirce elucidates how a paragraph can serve at once as a pedestal elevating a particular enunciation or image, as a curtain between scenes, and as a drum marking the crescendo of action or whatever rhythm. Peirce’s understanding of representation as a dynamic process of sign relation also presents a new perspective on play as the degeneration of strict rule into simple association. According to the Kantian view I adopt, beauty happens in the play of cognitive powers in reflective judgment. The text’s material support does not initiate aesthetic movement; rather, paragraphs and punctuation can ornament the matrix of relations that constitute language.

My study intends to highlight the relevance of formalism as a critical method and as an understanding of the aesthetic. Formal schemas can pry open the complexity of representation in language. When Johan Huizinga says that “all play has its rules,” he means that certain rules create conditions of indeterminacy in which play flourishes; there is not a rule for how play itself works (11). Although my excerpts from Flaubert do not exemplify all these dimensions, Michael Silverstein’s social semiotics and Caroline Levine’s expansive conception of forms do indicate how stereotypes, power relations, and other meaning can be analyzed as formal aspects of representation in language. Their work builds on Frances Ferguson’s accusation that New Criticism was not formalistic enough, owing to an innocent reliance on dramatic tone (viii). Indeed, Kant’s oft-criticized exclusion of affect from beauty becomes less defensible as the notion of form broadens (Guyer).

Yet Kantian idealism remains important for literary study. Fernando M. F. Silva and Bernd Dörflinger’s collection of essays Kant on Poetry / Kant über Poesie recovers from unpublished texts and commentary his mature understanding of poetry as the anthropological practice par excellence. Positioned as a mediator between empirical experience and pure reason, poetry and literature lead development of ideas that form the sensus communis. This mediating role aligns with the surprising claim that aesthetic judgments are both subjective and universal, and it testifies to literature’s role in founding the objective on an immanent basis, as Luc Ferry puts it (155). One can also glimpse in this mediation the anti-imperial commitment that Sankar Muthu outlines. Kant’s affirmation of humanity counteracts the skepticism that deconstruction and empiricism share, as Ferguson notes, and tempers the isolation and entrenchment that skepticism exacerbates. Derrida nonetheless reminds us that writing merely ornaments a mysterious play of reason that knows no better image than a blank.