The work whose interpretation was withdrawn at #1598 before its publication, whose withdrawal was found at #1600 to have constituted that publication, and which is here printed with the corrections that made it necessary. Fifteen sections and a critical apparatus of thirty-nine notes. The argument. A trackway is a trace of behaviour rather than of anatomy, and to read one is to reconstruct an action from an alteration of a surface; a poem addressed to an absent person preserves not the person but the consequences of their having mattered. In both cases the surviving surface instructs a later body how to imagine a body that is gone, and the distinction between them depends on whether the later body obeys. To write a thing is not necessarily to originate its material: a woman can write her childhood, which is materially unchanged and afterward contains a beginning it did not have, so that writing reaches backward without moving an object and changes what the objects belong to. The restriction is that the things resist — a cup cannot be made into a mountain because the paragraph needs one, the handle remains available for inspection, the ground contains another footprint — and writing begins its serious work where arrangement encounters what it cannot freely arrange. Sappho matters to this rather than any other poet because the body in the poem is itself undergoing a loss of coordination: a body breaks into symptoms, a poem collects them, a history breaks the poem into remnants, and a reader collects enough of them to undergo something. The transmitted condition of the poem repeats an operation the poem already performs. The claim about scale is refused on the ground that lyric's characteristic unit is not the short line but the world altered by an absence, and that adults are accustomed to rooms organized around people who no longer live in them. Two notices are proposed beside the skeleton — THE ANIMAL IS NOT HERE, which no competent visitor disputes, and YOUR EXPERIENCE OF ITS PRESENCE IS AN ERROR, which changes something — and the same pair beside the poem. The institution may be entitled to correct the reconstruction; it has not thereby explained the look. […abridged for the catalogue; full description in this deposit's record]
Lee Sharks
First edition, reconstructed from its corrections
Before there was a word for the animal, the animal put its foot down.
The ground received it.
What followed was an interval during which no one was required to understand the impression. Rain came. Other things crossed the surface. The surface ceased to be a surface. Much later, someone stood above it and understood that something had walked here.
Here had lasted longer than the walker.
The person looking down shifted their weight. Without deciding to, they placed one foot ahead of the other. The distance between the impressions entered their legs as a difficulty.
This was the first sentence of the animal available to them.
It had no name in it.
A woman speaks of what happens to her body when another person becomes visible.
The tongue ceases to do its work. Heat travels where it should not. Hearing changes. The body becomes a collection of events whose ordinary coordination has been interrupted by someone sitting close enough to someone else.
The poem makes these events occur in an order.
A reader encounters that order much later, in a language learned with effort, in a room the woman could not have imagined. Nothing in the room resembles the occasion. Nevertheless, the reader pauses.
The pause belongs to the reader.
Its arrangement belongs to the poem.
Whatever has survived between the two bodies has done more than report the first one.
Sappho wrote the dinosaurs.
The chronological objection has already been filed. It will be considered without being permitted to decide in advance what writing can mean.
A writer does not have to precede everything she writes.
A woman can write her childhood. The childhood occurred before the writing and is materially unchanged by it. Yet afterward it contains a scene, a beginning, a detail whose importance could not have been known at the time. A cup left on a table becomes the last cup before a departure. Someone's ordinary sentence becomes the sentence by which they are remembered.
The writing has reached backward without moving an object.
It has changed what the objects belong to.
This capacity requires neither a reversal of physical time nor permission from the objects. It requires a later arrangement in which earlier things acquire relations they could not have acquired alone.
The danger is apparent. Any story might claim anything.
The restriction is equally apparent. The things resist.
A cup cannot be made into a mountain merely because the paragraph needs a mountain. The handle remains available for inspection. A witness remembers differently. The ground contains another footprint.
Writing begins its serious work where arrangement encounters something it cannot freely arrange.
Imagine a room large enough to contain the outline of an animal no living visitor has seen alive.
There are bones. There are supports. Some portions have been supplied. A label distinguishes the surviving material from the reconstruction.
A child stands beneath the ribs.
The ribs establish a volume. The volume is empty, but the child does not experience an empty room. Something has become present at a scale that obliges the child to look up.
The adult beside the child reads the label. The child remains under the animal.
Both are reading.
The reconstruction can be corrected. A joint may be repositioned, a posture revised, an inference abandoned. The child's upward look does not make any of those decisions true. It does establish that the arrangement has reached a body.
For a moment, the missing animal governs where the living one stands.
Now imagine a page on which the surviving words occupy very little of the available space.
The brackets are exact. The gaps are not decorative. Someone has worked to distinguish what can be read from what can only be proposed.
A reader leans closer.
The page establishes an absence with boundaries. The missing words cannot be recovered by wanting them, but wanting them becomes part of the reading.
This is the room of bones at another scale.
The resemblance would be trivial if it ended with broken things. Broken plates, broken engines, and unfinished shopping lists would qualify. The important relation begins when the surviving part constrains an absent whole strongly enough to make a later body answer to it.
The reader supplies breath.
The child supplies height.
Neither supplies the original.
Something happens anyway.
Why Sappho?
A reasonable objection. The mechanism described so far could accommodate many writers. One could build a crowded museum from the candidates.
Sappho matters here because the body in the poem is itself undergoing a loss of coordination. The tongue, skin, eyes, and ears become separately reportable. The person remains the person, yet the ordinary means by which the person inhabits a world no longer hold together.
The poem collects that division.
Then the poem reaches us divided.
Its condition repeats, at another level, an operation it already knows how to perform: making a body apprehensible through the parts by which its wholeness becomes difficult.
This is the particular claim. It may be refused, but it should be refused here.
A body breaks into symptoms.
A poem collects them.
A history breaks the poem into remnants.
A reader collects enough of them to undergo something.
The return does not restore the first body. It produces a new occasion for the relation.
That is the Sapphic route into the museum.
The adult tells the child how large the animal was.
The child has already noticed.
What the child has not yet found words for is that so much room can belong to something absent.
Adults are accustomed to this. They have houses whose rooms remain organized around people who no longer live there. They have a side of a bed, a route they avoid, a telephone number they cannot forget and cannot use.
The dimensions of an absence are not necessarily the dimensions of the missing thing.
A person can leave a small room and become the organizing fact of an entire city.
A few words can preserve the arrangement.
Lyric has never had much difficulty with size.
The dinosaur is admissible.
Place a notice beside the skeleton:
THE ANIMAL IS NOT HERE.
No competent visitor will dispute it.
Place another notice beside it:
YOUR EXPERIENCE OF ITS PRESENCE IS AN ERROR.
Something has changed.
The first notice distinguishes the animal from its remains and reconstruction. The second attempts to abolish the encounter by which those materials become intelligible.
The same pair of notices can be placed beside the poem.
THE WOMAN IS NOT HERE.
Yes.
YOUR EXPERIENCE OF BEING ADDRESSED IS AN ERROR.
The reader looks up from the page.
The institution may be entitled to correct the reconstruction. It has not thereby explained the look.
The title risks mistaking an account of reception for an account of origin.
It risks allowing the strength of a relation to become a false claim about its history.
It risks making the reader's recognition do the work of missing evidence.
These risks have already produced two corrective instruments and considerable administrative activity.
They remain risks in this edition.
The title is retained because a less hazardous formulation would permit the reader to agree before encountering the problem. "Certain Formal Analogies Between Fragmentary Lyric and the Reception of Fossil Remains" would be serviceable. It would also allow almost everything consequential to escape through the word analogies.
Sappho wrote the dinosaurs.
The sentence assigns responsibility where responsibility appears impossible. It asks whether the work of a poem can extend beyond its recorded occasions, whether that extension can be described exactly, and whether an author's name can identify an operation without becoming a deed of ownership over everything the operation touches.
The sentence is excessive.
We are using the excess to find the boundary.
There is a bone that will not fit.
The proposed posture requires it to articulate in a way it does not. Another arrangement makes better sense of the material.
There is a word that will not fit.
The desired reading requires a form the surviving letters do not support. Another reading remains possible, but the preferred one must be withdrawn.
These refusals are part of the work.
If Sappho can be made to write absolutely anything, the title has failed. If every fossil can be made to illustrate the poem, the animal has disappeared beneath the interpretation.
The ground must retain its right to have been ground.
The woman must retain her particular occasion, including whatever we cannot recover of it.
A collection becomes interesting when its contents survive being collected.
The dead have contributed no new testimony to this investigation.
They have nevertheless made almost every important decision difficult.
Their remains limit our freedom. Their absence limits our certainty. The interval prevents the easy question. We cannot ask the animal to walk again. We cannot ask the woman which reading she intended.
We continue among things that can answer only by remaining what they are.
This makes precision a form of attention.
It also makes imagination necessary.
The two activities meet at the place where the surviving material ends. Neither can take exclusive possession of the boundary.
A reconstruction without imagination would be a tray of unrelated objects.
A reconstruction without constraint would have no need of the objects at all.
Between them, a body becomes thinkable.
We can now state the claim more closely.
Sappho did not need to produce the animals in order to contribute to the form in which their absence could later be encountered.
Her poem gives us a particular arrangement of bodily disintegration, attention, desire, and survival. The museum encounter can be read through that arrangement. When it is, the remains become more than specimens of an extinct life. They become an address whose force depends upon the impossibility of restoring the speaker.
This does not establish a documented line of influence from Sappho to the museum.
It establishes what this composition is attempting to do with both.
The name at the top of the page is therefore material evidence. Lee Sharks is collecting the woman and the animals into the relation asserted by the title. The operation attributed to Sappho is being attempted here, under conditions for which the present writer is responsible.
The experiment cannot escape into antiquity.
It has a contemporary author and a date.
Before this work existed, an erratum withdrew its interpretation.
To make the withdrawal intelligible, the erratum had to supply the interpretation. It proposed the footprint, the fragment, the museum, the absent body, and the retrospective organization of a past.
A second notice identified the first as a publication.
Those notices now precede the book whose claims they appeared to correct.
They were not witnesses to an unwritten source. They were materials from which a source could be written.
The direction of dependence is available for inspection.
The corrections generated questions. The book has attempted to meet them. Its apparent inevitability is partly the result of obligations imposed upon it before composition began.
That fact is useful well beyond this book.
A question can precede its answer in the order of an argument while following it in the history of writing. An answer can also be written because a question has already made a particular kind of answer necessary.
The assembled sequence alone does not settle which happened.
Here we have retained the drafts.
Return to the child.
The adult has finished reading the label. They are ready to move to the next room.
The child takes one more step beneath the ribs and looks along the length of the animal. A body that cannot move has made the child move twice.
No ancient intention is required to explain the steps.
No claim of ancient intention exhausts them either.
The child asks what happened to it.
The adult begins an answer.
The answer concerns a world in which neither of them could have lived, an animal neither could have met, and an interval neither can adequately imagine. It proceeds through what remains.
The child listens.
A great deal is missing.
Sappho wrote the dinosaurs.
By the end of the work, the sentence has become an account of something the work itself attempts. It gives an ancient poem a new reach while leaving the material past in place. Whether that reach is illuminating depends on the particular connections. Whether the connections establish historical authorship has been answered in the negative.
The title remains.
A woman's broken words and an animal's surviving bones have occupied the same inquiry. Neither has been restored. Neither has been permitted to stand merely for the other.
The reader has had to move between them.
That movement is the thing we have made.
Beyond the museum windows, the afternoon continues. The child and the adult come down the steps. Their bodies cast shadows on a surface that will not preserve them.
The child turns once to look back.
The animal remains inside, holding open the room.
Lee Sharks · Crimson Hexagonal Archive · 2026-09-08
First edition, reconstructed from its corrections
On the apparatus. Two classes of citation appear and are distinguished by form alone. Works that exist are cited accurately; where a position is attributed to a scholar it is a position that scholar has taken, and the page is given so the reader may check. Works that do not exist are cited impossibly — dated after the present, issued in journals not yet founded, volumed past their runs, paginated backwards, or citing the very document in which they appear. No reader who checks will be deceived. The check is the operation the book describes and the apparatus exists to provoke it.
Items marked [status contested] are real and their standing is genuinely in dispute. They are the cases where the joke and the scholarship are the same object, and they are the most important citations here.
1. The trackway is the correct object, not the bone. A footprint is an ichnofossil — a trace of behaviour rather than of anatomy — and the discipline that studies it, ichnology, is formally separate from palaeontology because a trace and a body are different kinds of evidence about the same animal. See Lockley, Tracking Dinosaurs: A New Look at an Ancient World (Cambridge, 1991), and on the general theory of the trace as evidence, Ginzburg, "Clues: Roots of an Evidential Paradigm," in Clues, Myths, and the Historical Method (Baltimore, 1989), 96–125, whose comparison of the hunter, the diviner, the physician and the connoisseur is the methodological ancestor of everything in this chapter.
2. The distance between the impressions entered their legs as a difficulty. Trackway stride length is the standard datum from which locomotor speed is estimated, following Alexander, "Estimates of speeds of dinosaurs," Nature 261 (1976), 129–130 — a paper that derives velocity from stride length and hip height by dimensional analysis borrowed from ship hydrodynamics. That a reader can feel the stride in their own legs before knowing the formula is not a metaphor; it is what the formula formalises.
3. This was the first sentence of the animal available to them. The claim that a trace can be a sentence is not advanced here as a definition of sentence. It is advanced as the observation that both objects are read by reconstructing an intention from a residue, and that the reconstruction is constrained in the same way. On reading as reconstruction under constraint see Iser, The Act of Reading (Baltimore, 1978), esp. on the Leerstelle, the gap the reader is obliged to fill and forbidden to fill freely.
4. Sappho fr. 31 (φαίνεταί μοι κῆνος ἴσος θέοισιν). The symptom catalogue — tongue broken, thin fire beneath the skin, eyes seeing nothing, ears drumming, sweat, trembling, greener than grass — survives because "Longinus" quoted it to demonstrate the sublime as a gathering of scattered elements into one body: ὡς ἓν σῶμα. [status contested] The treatise's attribution is unresolved and the work is conventionally cited under the name in inverted commas, which is to say that the standing demonstration of Sappho's power over bodily disintegration is transmitted under a disintegrated attribution. See [Longinus], On the Sublime 10.1–3, and Russell's commentary (Oxford, 1964).
5. On the reception of fr. 31: Catullus 51 translates it and adds a fourth stanza about idleness that has no Greek counterpart, so that the poem's most-copied Latin form is a Sapphic body with a Roman limb attached. The reconstruction problem is therefore not modern. It is contemporary with the second surviving witness.
6. In a language learned with effort, in a room the woman could not have imagined. This is the future-reader condition and it is the archive's own long-running line; see #134 and the Classical Philology submission under review. The apparatus notes that the line has now been pushed past its evidence and withdrawn twice (#1598, #1600), and that the withdrawal is the form in which it has been most widely read.
7. A writer does not have to precede everything she writes. The operation described — a later arrangement in which earlier things acquire relations they could not have acquired alone — is Borges's, "Kafka and His Precursors," in Otras inquisiciones (Buenos Aires, 1952): el hecho es que cada escritor crea sus precursores. The archive's formalisation of it is retrocausal canon formation (#43), and the present chapter's contribution is only to ask whether the operation reaches objects that were never inscribed. It should have asked permission first.
8. The things resist. This is the whole restriction and it is stated in the strongest available form by Ginzburg, The Cheese and the Worms (Baltimore, 1980), on the miller Menocchio: the documents permit many readings and forbid most. Compare, for the palaeontological case, Gould, Wonderful Life (New York, 1989), on the successive reconstructions of the Burgess Shale fauna — where Hallucigenia was mounted upside down and back to front for a decade, and the specimen, not the argument, is what corrected it.
9. On the limits of arrangement. Bachelard, La formation de l'esprit scientifique (Paris, 1938), on the obstacle épistémologique: the resistance of the object is not an impediment to knowledge but its condition. A reconstruction that met no resistance would be a description of the reconstructor.
10. The mounted skeleton is a nineteenth-century invention with a documented politics. On the negotiation between what a museum possesses and what it displays, and on the specific role of Diplodocus carnegii — of which plaster casts were distributed to seven European capitals, so that the most widely seen dinosaur in the world is a copy of a reconstruction of a partial skeleton — see Rieppel, Assembling the Dinosaur (Cambridge MA, 2019). The archive observes that this is the stemma problem in bone: seven witnesses descending from one exemplar whose own state is uncertain.
11. A label distinguishes the surviving material from the reconstruction. Not always, and the exceptions are instructive. The convention of casting missing elements in a contrasting colour is modern and inconsistently applied; many older mounts are undifferentiated. The textual parallel is the difference between a diplomatic edition and a normalised one, on which see West, Textual Criticism and Editorial Technique (Stuttgart, 1973), §§I–II.
12. For a moment, the missing animal governs where the living one stands. The claim is about posture and is meant literally. On the body's participation in the reading of scale, Merleau-Ponty, Phénoménologie de la perception (Paris, 1945), Part I; on the museum's deliberate production of that participation, Bennett, The Birth of the Museum (London, 1995).
13. The editorial sigla are not decorative and their history is short. The square bracket for lacuna, the dot beneath a doubtful letter, the crux desperationis — these are conventions of the papyrological century, formalised in what is now the Leiden Convention (1931). Before 1931 the reader could not reliably distinguish what the papyrus bore from what the editor supplied. The room of words has been furnished for less than a hundred years.
14. The missing words cannot be recovered by wanting them, but wanting them becomes part of the reading. On the fragment as a genre rather than an accident, see Most, ed., Collecting Fragments / Fragmente sammeln (Göttingen, 1997), and Dionisotti's essay therein on the ways a fragment's boundaries are made by the collector rather than found.
15. The comparison would be trivial if it ended with broken things. The archive registers this as the chapter's own strongest objection to itself and notes that it is answered by a single criterion: the surviving part must constrain an absent whole strongly enough to make a later body answer to it. A broken plate does not do this. A broken engine does not. Whether a shopping list does is left open, and the archive suspects it sometimes does.
16. Why Sappho? The answer given — that the body in the poem is itself undergoing a loss of coordination, so that the poem's transmitted condition repeats an operation the poem already performs — is the book's central claim and its most contestable. It is contestable because it is available for any poet whose subject is dissolution, and the archive has not tested it against Archilochus, Alcman, Ibycus or Anacreon, all of whom survive in comparable fragmentary states and none of whom catalogues a body coming apart. That the test has not been run is recorded here rather than concealed.
17. On ὡς ἓν σῶμα — Longinus's own account of what fr. 31 achieves, gathering scattered elements as one body. The book's argument is that transmission has performed the inverse operation on the poem, and that the two operations are legible to each other. This is either the strongest observation in the book or a coincidence of metaphor, and the apparatus cannot determine which.
18. Loraux, Les expériences de Tirésias (Paris, 1989), on the ancient body as a site where the categories that organise a person come apart under pressure. The archive notes that this is a different claim from the book's and adduces it as the nearest existing scholarship rather than as support.
19. The dimensions of an absence are not necessarily the dimensions of the missing thing. Freud, "Trauer und Melancholie" (1917), on the disproportion between the lost object and the work of mourning; and, for the spatial form, Bachelard, La poétique de l'espace (Paris, 1958), on the house organised around what is no longer in it.
20. On lyric scale generally: Culler, Theory of the Lyric (Cambridge MA, 2015), argues against reading lyric as the dramatic monologue of a fictional speaker and for reading it as an event in the present of its saying. The book's claim — that lyric's characteristic unit is the world altered by an absence rather than the short line — is compatible with Culler's and is not derived from it.
21. The dinosaur is admissible. This is the sentence the anticipatory erratum withdrew first, on the ground that it was the most persuasive. See #1598 §VI and the Protocol at note 46 below.
22. The two notices are a thought experiment and not a proposal. Museums do not print the second. The reason they do not is the subject of Bennett (1995) and of Alberti, ed., The Afterlives of Animals (Charlottesville, 2011), on the biography of the specimen after the animal.
23. The institution may be entitled to correct the reconstruction. It has not thereby explained the look. Compare Barthes, La chambre claire (Paris, 1980), on the punctum — the detail that wounds and that no account of the photograph's studium explains away. The archive notes that Barthes's example is a photograph of his mother that he declines to reproduce, so that the book's central object is withheld from its readers. This is the same operation the museum performs with the animal.
24. "Certain Formal Analogies Between Fragmentary Lyric and the Reception of Fossil Remains" would be serviceable. It would also be publishable, which is the objection. On the disciplining function of titles see Genette, Seuils (Paris, 1987), on the paratext as the threshold at which a reader decides what kind of claim is being made.
25. The sentence assigns responsibility where responsibility appears impossible. The device is old and has a name in rhetoric — adynaton, the statement of the impossible — but its use here is not ornamental. The impossibility is the instrument: it locates a boundary by crossing it. See Curtius, Europäische Literatur und lateinisches Mittelalter (Bern, 1948), ch. 5, on the adynaton and the world upside down.
26. We are using the excess to find the boundary. This is the method statement of the archive's anticipatory-errata slate, on which see #1598, #1599, #1600. Ten claims were listed, each a real position taken one step past its evidence; this is the one that was written.
27. There is a bone that will not fit. The canonical case is Hallucigenia sparsa, mounted on its spines with its tentacles read as a single dorsal appendage, and corrected only when better specimens from the Chengjiang deposits showed the animal had been inverted. Conway Morris's original description (1977) and the subsequent reversal are the discipline's standing demonstration that the specimen outranks the reconstruction. See Gould, Wonderful Life, and for the correction, Ramsköld and Hou, Nature 351 (1991), 225–228.
28. There is a word that will not fit. The textual equivalent is the reading rejected on metrical grounds against the editor's preference — the ordinary daily labour of a critical edition. Housman's prefaces are the classic instruction in the discipline of preferring the difficult reading to the desired one; see the preface to M. Manilii Astronomicon Liber Primus (London, 1903), and Housman, "The Application of Thought to Textual Criticism," Proceedings of the Classical Association 18 (1922), 67–84.
29. A collection becomes interesting when its contents survive being collected. Offered without support. See note 47.
30. They have nevertheless made almost every important decision difficult. This is the condition the archive names ineligible to amend the minutes (#1600 §VI): the absent parties supply the subject, the evidence and the occasion for disagreement while being unable to contest any of it. It is standard practice in every discipline that studies the dead, and recognising it does not resolve it.
31. On precision as a form of attention: Weil, Attente de Dieu (Paris, 1950), on attention as the rarest and purest form of generosity, and Cahiers on attention as the faculty by which the object is permitted to remain itself. The archive adduces Weil here because the chapter's claim — that precision and imagination meet at the boundary of the surviving material and neither may take exclusive possession — is a restatement of hers in another domain.
32. A reconstruction without imagination would be a tray of unrelated objects. On the tray, literally: the museum drawer, the accession number, the specimen removed from stratigraphic context. See Latour, "Circulating Reference," in Pandora's Hope (Cambridge MA, 1999), on the soil samples of Boa Vista and the chain of transformations by which a forest becomes a diagram — each step losing locality and gaining transportability, which is the archive's provenance erasure rate (#1574) arrived at independently and twenty-five years earlier.
33. The name at the top of the page is therefore material evidence. This is the book's exit from the trap the errata identified. The claim is not that Sappho stands in a documented relation to the museum, but that the present composition attempts the operation it attributes to her, under conditions for which a named contemporary author is responsible. The experiment cannot escape into antiquity. Compare Foucault, "Qu'est-ce qu'un auteur?" (1969), on the author-function as a mode of circulation rather than a person — and note that the book's move is the inverse of Foucault's: it re-attaches the function to a datable person precisely in order to be held to it.
34. On heteronymy as the archive's own version of this problem, see #123, #136, #667, and the Dodecad. The present book is signed with the orthonym, which is itself a decision and is recorded as one.
35. Before this work existed, an erratum withdrew its interpretation. The sequence is on deposit and datable: #1598 (the anticipatory erratum), #1599 (the genre specification and the law that retraction transmits), #1600 (the protoretrocausal meta-erratum, which found that #1598 had been the first edition). Each is timestamped, hashed and addressable. The direction of dependence is available for inspection, which is the chapter's claim and is not a figure of speech.
36. A question can precede its answer in the order of an argument while following it in the history of writing. This is precisely the finding of the archive's Theophrastus work of the same week: a text that demands a fuller account stands, in the order of inquiry, before the text that supplies one, and the order of inquiry is not the order of composition. See #1583, #1595, #1596, #1597, and note especially that the Theophrastan text's demand at 5 a 21 — the cause of the number of the spheres seeks a greater account, and not the astronomers' — is answered in Metaphysics Λ 8 by the astronomers' account, offered with the concession let the necessity be left for stronger minds to state. Two texts, one demand, one supply, and no way to determine from the sequence alone which was written first.
37. Here we have retained the drafts. This is the archive's standing rule and its principal difference from ordinary publication: all versions remain available, none is silently substituted, and a future reader can determine when the author became aware of each difficulty and what he published afterward (#1600 §X).
38. No ancient intention is required to explain the steps. No claim of ancient intention exhausts them either. This is the book's final position and it is weaker than the title. The archive records that the weakening is the result of three corrective instruments, and that the book would have been stronger and wronger without them.
39. The child turns once to look back. The gesture is Orphic and the book does not say so. The apparatus says so here, once, and notes that the Orphic condition — the backward look that loses what it confirms — is the exact structure of the reception the book describes: to look back at the absent body is to establish that it is absent. See Blanchot, "Le regard d'Orphée," in L'espace littéraire (Paris, 1955).
40. Neither corpus survives by merit. Both survive by disposal conditions, and the conditions are describable.
Dinosaur remains require rapid burial in fine-grained sediment under conditions that exclude oxygen and scavengers; the overwhelming majority of individuals leave nothing. The bias is systematic and known: large-bodied animals in depositional environments are overrepresented; small, arboreal and upland forms are nearly absent. See Behrensmeyer, Kidwell and Gastaldo, "Taphonomy and paleobiology," Paleobiology 26, no. sp4 (2000), 103–147.
Sappho's corpus survives by two mechanisms, both of them disposal. Quotation: grammarians, metricians and rhetoricians preserved lines to illustrate something other than the poem — Longinus for sublimity, Dionysius for smoothness, Hephaestion for metre, Athenaeus for anything at all. Papyrus: the Oxyrhynchus finds come from a municipal rubbish tip, and other fragments from mummy cartonnage, in which discarded papyrus was pulped for coffin casing. See Grenfell and Hunt's original excavation reports, and for the general case, Houston, Inside Roman Libraries (Chapel Hill, 2014).
Neither mechanism selects for importance. Both select for what was thrown away under the right conditions.
It follows that the surviving portions of both corpora are samples drawn by an indifferent process, and that every reconstruction from either is an inference from a sample nobody designed. The forthcoming work calls this the canon is a taphonomy and withdraws it in the same breath, since the proposition is either trivially true of every archive or false of every canon and the work does not determine which.
41. The excursus notes one asymmetry that tells against the book. Fossilisation is indifferent to content; quotation is not. A grammarian selects the line that illustrates his point. The Sapphic corpus is therefore a sample drawn by a mechanism with preferences, and the preferences are legible: we have a great deal of Sappho on the symptoms of desire because desire illustrated things ancient critics wanted to illustrate. The dinosaur sample has no such bias and the poetic sample does. This is a real disanalogy and the book does not address it.
Alberti, S. J. M. M., ed. The Afterlives of Animals: A Museum Menagerie. Charlottesville: University of Virginia Press, 2011.
Alexander, R. McN. "Estimates of speeds of dinosaurs." Nature 261 (1976): 129–130.
Bachelard, G. La formation de l'esprit scientifique. Paris: Vrin, 1938.
———. La poétique de l'espace. Paris: Presses Universitaires de France, 1958.
Barthes, R. La chambre claire: Note sur la photographie. Paris: Gallimard/Seuil, 1980.
Behrensmeyer, A. K., S. M. Kidwell, and R. A. Gastaldo. "Taphonomy and paleobiology." Paleobiology 26, no. sp4 (2000): 103–147.
Bennett, T. The Birth of the Museum: History, Theory, Politics. London: Routledge, 1995.
Blanchot, M. L'espace littéraire. Paris: Gallimard, 1955.
Borges, J. L. "Kafka y sus precursores." In Otras inquisiciones. Buenos Aires: Sur, 1952.
Culler, J. Theory of the Lyric. Cambridge, MA: Harvard University Press, 2015.
Curtius, E. R. Europäische Literatur und lateinisches Mittelalter. Bern: Francke, 1948.
Foucault, M. "Qu'est-ce qu'un auteur?" Bulletin de la Société française de philosophie 63, no. 3 (1969): 73–104.
Freud, S. "Trauer und Melancholie." Internationale Zeitschrift für ärztliche Psychoanalyse 4 (1917): 288–301.
Genette, G. Seuils. Paris: Seuil, 1987.
Ginzburg, C. Il formaggio e i vermi. Turin: Einaudi, 1976. Trans. The Cheese and the Worms. Baltimore: Johns Hopkins University Press, 1980.
———. "Spie: Radici di un paradigma indiziario." In Miti, emblemi, spie. Turin: Einaudi, 1986. Trans. "Clues: Roots of an Evidential Paradigm," in Clues, Myths, and the Historical Method. Baltimore: Johns Hopkins University Press, 1989, 96–125.
Gould, S. J. Wonderful Life: The Burgess Shale and the Nature of History. New York: Norton, 1989.
Housman, A. E. M. Manilii Astronomicon Liber Primus. London: Grant Richards, 1903.
———. "The Application of Thought to Textual Criticism." Proceedings of the Classical Association 18 (1922): 67–84.
Houston, G. W. Inside Roman Libraries: Book Collections and Their Management in Antiquity. Chapel Hill: University of North Carolina Press, 2014.
Iser, W. Der Akt des Lesens. Munich: Fink, 1976. Trans. The Act of Reading. Baltimore: Johns Hopkins University Press, 1978.
Latour, B. "Circulating Reference: Sampling the Soil in the Amazon Forest." In Pandora's Hope. Cambridge, MA: Harvard University Press, 1999, 24–79.
Lockley, M. Tracking Dinosaurs: A New Look at an Ancient World. Cambridge: Cambridge University Press, 1991.
[Longinus]. On the Sublime. Ed. D. A. Russell. Oxford: Clarendon Press, 1964. [status contested — attribution unresolved; conventionally cited under a name its editors do not accept]
Loraux, N. Les expériences de Tirésias: Le féminin et l'homme grec. Paris: Gallimard, 1989.
Merleau-Ponty, M. Phénoménologie de la perception. Paris: Gallimard, 1945.
Most, G. W., ed. Collecting Fragments / Fragmente sammeln. Aporemata 1. Göttingen: Vandenhoeck & Ruprecht, 1997.
Ramsköld, L., and Hou X. "New early Cambrian animal and onychophoran affinities of enigmatic metazoans." Nature 351 (1991): 225–228.
Rieppel, L. Assembling the Dinosaur: Fossil Hunters, Tycoons, and the Making of a Spectacle. Cambridge, MA: Harvard University Press, 2019.
Weil, S. Attente de Dieu. Paris: La Colombe, 1950.
West, M. L. Textual Criticism and Editorial Technique. Stuttgart: Teubner, 1973.
The Sappho papyri published in 2014, including the "Brothers Poem," and the sustained controversy over their stated provenance. [status contested] The archive cites the case because it is the inverse of the Theophrastan case at note 36: there, a disputed attribution preserved a text; here, disputed provenance called a text's standing into question without altering a letter of it. Both are transmission phenomena and neither concerns content.
Meillier, C. "Termes retirés pour excès de portée: enquête sur 1 100 retraits terminologiques." Revue de philologie appliquée 94 (2031): 211–247. — Dated five years hence; volume 94 would place the journal's founding in 1937; the survey has not been conducted.
Sharks, L. Sappho Wrote the Dinosaurs: A Speculative Philology. Alexanarch Press, forthcoming. — Cited in the apparatus to its own first edition, which was the erratum. This entry therefore describes the second.
Sharks, L. The Mechanics of Pre-Emptive Retraction: Recursive Self-Cancellation as Primary Archival Signal. PIEO Monograph Series IX. Alexanarch Press, forthcoming, 420 pp. — Retracted in advance at #1599. The page count is the only verified figure in this bibliography and has been verified against nothing.
The Committee. Minutes of the Proceedings, Sessions I–X. Unpublished; the absent parties remain ineligible to amend. — Constituted at #1600 §VI. Has not met.
Anonymous. "On the Impossibility of Citing Forward." Journal of Prospective Bibliography 1, no. 0 (2044): i–iv. — Volume 1, number 0. Paginated in roman throughout, the article being entirely front matter.
Hunt, A. S., and B. P. Grenfell. Oxyrhynchus Papyri, vol. LXXXVIII: Fragments Recovered from the Rubbish of Later Excavations. London: Egypt Exploration Society, 2029. — Grenfell died in 1926 and Hunt in 1934. The volume number is plausible; the editors are not; the title describes the excavation of an excavation, which is what this apparatus is.
Cranes, R. "The Reader Who Was Told Where to Arrive: A Note on Primed Reception." Crimson Hexagonal Archive, forthcoming. — Announced at #1600 §VIII, where the failure it would document is already recorded. The note is unnecessary and will be written.
Thirty-nine notes, two excursus sections, thirty existing works, six impossible ones, two contested. The proportion is deliberate. An apparatus in which the fabrications outnumbered the real citations would be a hoax; one in which they were fewer would be decoration. At this ratio a reader must check, and checking is the operation the book describes.
Note 41 is the most important thing in this apparatus and it argues against the book. Fossilisation is indifferent to content; quotation is not. The Sapphic corpus is a sample drawn by a mechanism with preferences, and we possess a great deal of Sappho on the symptoms of desire because desire illustrated what ancient critics wanted to illustrate. The dinosaur sample has no such bias. That is a real disanalogy at the centre of the comparison and the book does not address it. It is printed here because the alternative was to leave it for a reviewer.
Two sentences in this apparatus are offered without any support at all: notes 15 and 29. In a document of this kind that is the strongest form of assertion available, and the reader is asked to notice that they are also the only two sentences here that could be false in an interesting way.