Theophrastus's short Metaphysics is read as a pupil's aporiai on Aristotle's theology. This paper argues that its sixth and seventh sections are a hinge written from inside the inquiry of Metaphysics Λ — μέχρι μὲν δὴ τούτων ἄρτιος ὁ λόγος … τὸ δὲ μετὰ ταῦτ᾽ ἤδη λόγου δεῖται πλείονος — of the same form as Λ's own chapter transitions (εἴρηται … λεκτέον; πότερον δὲ μίαν θετέον … καὶ πόσας): it summarizes Λ 6–7 clause by clause as achieved, names Λ 8's question as the next question, contests Λ 8's method in Λ 8's own terms (οὐ γὰρ ὅ γε τῶν ἀστρολόγων, answering Λ 8's φιλεῖν μὲν ἀμφοτέρους, πείθεσθαι δὲ τοῖς ἀκριβεστέροις), and lands on Λ 10's closing aporiai in Λ 10's own word (ἐπεισοδιώδης), turning Λ 7's νοῦς θιγγάνων on the inquirer. One work, continuous across the point where the name at the head of the roll changed. The catalogue in Diogenes records the state of affairs in its own wording — Ὑπομνημάτων Ἀριστοτελικῶν ἢ Θεοφραστείων — and the receiving texts describe the two libraries reaching Rome as one lot divided by Andronicus's πίνακες. Three readings of the division are weighed: a change of person, a librarian's partition, and the author's own editorial division of the work at the seam of named authorial function. The seam's placement — at the very sentence where the account changes status — discriminates against the partition and for the author; between person and function the page cannot decide, but function explains everything the page contains (the ἤ, the doublets, the bestowed name, the placement, the register) and person explains only the name. Companions in the dossier: Ep. II 314c, Strato's πλὴν ὧν αὐτοὶ γεγράφαμεν, μετωνόμασεν at D.L. V.38, and the title Τὰ πρὸ τῶν τόπων under both names. Nothing about dates or hands is asserted; the fragment is admitted into Λ's work and nothing else is admitted. Extracted from the notebook The First Draft; companion to EA-CORPORA-12 (#1580), Catalogues and Wills (#1581), and Many and One (#1576).
Lee Sharks · v0.2 · 2026-09-06 · extracted from the notebook The First Draft (§10) and developed as a standalone; v0.2 adds the third reading of the division — the author's own editorial division of the work at the seam of named authorial function — and the placement argument that favours it
Theophrastus's short Metaphysics is read as a pupil's questions about a master's theology: aporiai addressed, from outside, to the doctrine of Metaphysics Λ. I argue that the text's sixth and seventh sections are not a description of that doctrine but a hinge written from inside it — μέχρι μὲν δὴ τούτων ἄρτιος ὁ λόγος … τὸ δὲ μετὰ ταῦτ᾽ ἤδη λόγου δεῖται πλείονος — of exactly the form by which Λ moves between its own chapters; that the hinge summarizes Λ 6–7 as achieved, names Λ 8's question as the next question, and that the sections following contest Λ 8's method in Λ 8's own terms and land on Λ 10's closing aporiai in Λ 10's own word (ἐπεισοδιώδης). The inquiry is continuous across the point where the tradition changed the name at the head of the roll. The catalogue preserved in Diogenes already records this state of affairs in its own wording — Ὑπομνημάτων Ἀριστοτελικῶν ἢ Θεοφραστείων — and the receiving texts describe the two libraries reaching Rome as one lot, divided into two names by Andronicus's πίνακες. Three readings of the division are then weighed: a change of person, a librarian's partition, and an author's own editorial division of the work at the seam of named authorial function. The seam's placement — at the very sentence where the account changes status from complete to needing-more — discriminates against the librarian and for the author; between person and function the page cannot decide, but function explains everything the page contains and person explains only the name. The paper claims continuity of the inquiry, asserts nothing about dates or hands, and admits the fragment into Λ's work without admitting anything else.
The text known as Theophrastus's Metaphysics is about three thousand words of questions. It asks how the study of first things is to be delimited, whether there is a connection between the intelligibles and nature, whether the first substance is one or many, why the heavenly bodies desire motion rather than rest, whether the outermost sphere's revolution is its substance, how far order reaches, where the good is, and whether everything is for the sake of something. It answers none of them. It has been read, since it acquired its name, as a pupil's aporiai on a master's doctrine: Theophrastus, who heard Aristotle and inherited his school and his books, setting down the difficulties he found in the theology of Metaphysics Λ. On that reading the text's sixth section — which summarizes the doctrine of a single unmoved principle that is actuality and substance, indivisible, not a quantity — is a pupil's précis, and its seventh, which says that what follows needs more argument, is the pupil turning to his own questions.
Read at its grammar, the sixth and seventh sections are something else. They are a transition: the sentence-form by which a treatise closes one part and opens the next. Λ writes such sentences between its own chapters; the fragment writes one at the exact point where Λ 8 begins, and then does what a treatise does after such a sentence — takes up the next question, tests the method the next chapter used, and carries the inquiry to the place where the treatise's last chapter left it unresolved. So I shall argue: the fragment is the continuation of Λ's inquiry, one work with Λ by every mark the page can carry — question, vocabulary, position, and the hinge itself — and what divides them is not a break in the inquiry but a change of name at the point where the work's function changes: an editorial division of one work by named authorial function, which the catalogue, in one entry, confessed.
Two sentences of scope. Nothing here concerns persons: whether one hand or two wrote the continuous inquiry is not a question the page can answer, and the paper does not ask it. And nothing here concerns dates: the order established is the method's order — achieved, next, contested, unresolved — and the calendar is a separate operation that these texts do not perform.[^1]
The fragment's fifth section states the doctrine of the principle that moves as object of desire; the sixth closes the account and the seventh opens what follows:
μέχρι μὲν δὴ τούτων οἷον ἄρτιος ὁ λόγος ἀρχήν τε ποιῶν μίαν πάντων καὶ τὴν ἐνέργειαν καὶ τὴν οὐσίαν ἀποδιδούς, ἔτι δὲ μὴ διαιρετὸν μηδὲ ποσόν τι λέγων ἀλλ᾽ ἁπλῶς ἐξαίρων εἰς κρείττω τινὰ μερίδα καὶ θειοτέραν … τὸ δὲ μετὰ ταῦτ᾽ ἤδη λόγου δεῖται πλείονος περὶ τῆς ἐφέσεως, ποία καὶ τίνων. (§§6–7)[^2]
Up to this point the account is, as it were, complete — making one principle of all things, assigning it actuality and substance, and further saying it is not divisible nor any quantity but simply lifting it into a better and more divine portion … But what comes after this now requires more argument, concerning the desire: of what kind, and of what things.
Linger in the words. ἄρτιος ὁ λόγος: the λόγος is complete, in order, even — the word for a thing that has its parts and lacks none. μέχρι μὲν δὴ τούτων: up to these things — a boundary drawn by the writer in the inquiry's own progress. τὸ δὲ μετὰ ταῦτα: what comes after these — the next part named as next. ἤδη: now — the moment of turning. This is not the grammar of description from outside. It is the grammar Λ uses when it turns. Λ 6 opens: τίνες μὲν οὖν αἱ ἀρχαὶ τῶν αἰσθητῶν καὶ πόσαι … εἴρηται. ἐπεὶ δ᾽ ἦσαν τρεῖς οὐσίαι … περὶ ταύτης λεκτέον — what the principles of sensibles are has been said; since there were three substances, about this one it must now be said. Λ 8 opens: πότερον δὲ μίαν θετέον τὴν τοιαύτην οὐσίαν ἢ πλείους, καὶ πόσας, δεῖ μὴ λανθάνειν — whether one such substance is to be posited or more, and how many, must not escape us.[^3] εἴρηται / λεκτέον; ἄρτιος ὁ λόγος / λόγου δεῖται πλείονος: the same hinge, the same μέν … δέ, the same act of a writer standing inside a moving inquiry and marking where one part ends and the next begins.
And what the hinge closes is Λ 6–7, clause by clause. One principle of all things: Λ 7's ἐκ τοιαύτης ἄρα ἀρχῆς ἤρτηται ὁ οὐρανὸς καὶ ἡ φύσις. Actuality and substance: Λ 7's ἔστι τι ὃ οὐ κινούμενον κινεῖ, ἀΐδιον καὶ οὐσία καὶ ἐνέργεια οὖσα. Not divisible nor any quantity: Λ 7's close, δέδεικται δὲ καὶ ὅτι μέγεθος οὐδὲν ἔχειν ἐνδέχεται ταύτην τὴν οὐσίαν ἀλλ᾽ ἀμερὴς καὶ ἀδιαίρετός ἐστιν. Lifted into a better and more divine portion: Λ 7's τοῦτο γὰρ ὁ θεός.[^4] The fifth section had already stated Λ 7's central thesis as a result — the principle ἀκίνητος καθ᾽ αὑτήν, moving by a power greater and prior, τοιαύτη δ᾽ ἡ τοῦ ὀρεκτοῦ φύσις, source of ἡ κυκλικὴ ἡ συνεχὴς καὶ ἄπαυστος — which is Λ 7's κινεῖ δὲ ὧδε τὸ ὀρεκτὸν καὶ τὸ νοητόν … κινεῖ οὐ κινούμενα. The hinge summarizes what Λ 6–7 established, as established, and names the point where the account, being complete so far, must go on.
Where does it go? To Λ 8. The seventh section's question — many circular motions, some contrary to one another; if one mover, why not one motion; if many, how their concord toward the best — is Λ 8's question: πότερον δὲ μίαν θετέον … ἢ πλείους, καὶ πόσας. And the eighth section takes up Λ 8's answer and finds it wanting:
τὸ δὲ κατὰ τὸ πλῆθος τῶν σφαιρῶν τῆς αἰτίας μείζονα ζητεῖ λόγον· οὐ γὰρ ὅ γε τῶν ἀστρολόγων. (§8)
The account according to the number of the spheres seeks a greater explanation of the cause; for the astronomers' is not it.
Λ 8 had answered the question of number by the astronomers' count — Eudoxus's spheres, Callippus's additions — and had said so with a scruple of its own: τὸ δὲ λοιπὸν τὰ μὲν ζητοῦντας αὐτοὺς δεῖ τὰ δὲ πυνθανομένους παρὰ τῶν ζητούντων … φιλεῖν μὲν ἀμφοτέρους, πείθεσθαι δὲ τοῖς ἀκριβεστέροις — for the rest we must partly inquire ourselves and partly learn from the inquirers … loving both, but believing the more exact.[^5] The fragment answers the scruple: the count is the astronomers' business, and the cause of the plurality is not settled by counting. It then asks what Λ 8 had not: why bodies that have a natural desire pursue motion and not rest; whether desire, especially of the best, requires soul; whether the moved bodies are ensouled. These are not a pupil's objections to a doctrine. They are the questions a writer asks who has just written Λ 8 and is not satisfied with it — questions at the next step, in the treatise's own vocabulary (ὄρεξις, ἐφίεσθαι, τὸ πρῶτον, τὸ κινοῦν), at the treatise's own level of generality.
Follow the fragment past its middle and its questions converge on one place: Λ 10, the treatise's last chapter, which ends not with theses but with aporiai. The second section had already asked whether the whole is ἐπεισοδιῶδες — episodic, made of unconnected scenes — and preferred a συναφή. Λ 10 ends on the same word: those who make number first ἐπεισοδιώδη τὴν τοῦ παντὸς οὐσίαν ποιοῦσιν … τὰ δὲ ὄντα οὐ βούλεται πολιτεύεσθαι κακῶς.[^6] The tenth, fourteenth, fifteenth, and thirty-fourth sections ask where the good is — in the order, in the leader, how far order reaches, why there is no passage further toward the better — which is Λ 10's opening question (ποτέρως ἔχει ἡ τοῦ ὅλου φύσις τὸ ἀγαθὸν καὶ τὸ ἄριστον) worked at length. The eighteenth and thirty-second sections ask why the whole substance is in contraries, the worse nearly equal to or exceeding the better — Λ 10's aporia about contraries and evil. The third, twelfth, thirteenth, and thirty-third sections take up Λ 10's doxographic critique — number as first, the one and the indefinite dyad, Speusippus, the Academy stopping short — and extend it by names (Xenocrates, Hestiaeus, Archytas on Eurytus). And the twenty-fifth section turns Λ 7's most striking image back on the inquirer: Λ had said of the divine mind νοητὸς γὰρ γίγνεται θιγγάνων καὶ νοῶν; the fragment says of theory at the extremes αὐτῷ τῷ νῷ ἡ θεωρία θιγόντι καὶ οἷον ἁψαμένῳ, διὸ καὶ οὐκ ἔστιν ἀπάτη περὶ αὐτά — theory is mind touching, and so there is no error about them.[^7]
So the shape: the fragment presupposes Λ 6–7, marks their completion, takes up Λ 8's question, contests Λ 8's method, and works Λ 10's unresolved aporiai as its own body — then closes, ὃ δ᾽ ἐξ ἀρχῆς ἐλέχθη, by returning to its beginning with the demand for a ὅρος. It is a Β written after Λ: the treatise's own edge, gathered as a whole work, in the aporetic register, with the treatise's vocabulary, and without the treatise's two signature formulae (κινεῖ ὡς ἐρώμενον; νόησις νοήσεως) because it is not restating the achieved part but working the unfinished one.[^8] Two texts end on the same unresolved question, one under each name.
What, then, separates the two texts? Not a break in the inquiry: the hinge is the inquiry's own, the next question is the next question, the contest is in the treatise's terms, the landing is the treatise's last page. What separates them is that at some point in transmission the roll changed its name. And the tradition knew it. Diogenes's catalogue of Theophrastus carries the entry Ὑπομνημάτων Ἀριστοτελικῶν ἢ Θεοφραστείων — Notes, Aristotelian or Theophrastan — a work the catalogue could assign to either name and assigned, in its own wording, to both by ἤ.[^9] Twelve further titles stand letter-for-letter in both catalogues. The receiving texts describe the physical corpus as one lot: Strabo's τά τε Ἀριστοτέλους καὶ τὰ τοῦ Θεοφράστου βιβλία, sold together to Apellicon; Plutarch's τὰ πλεῖστα τῶν Ἀριστοτέλους καὶ Θεοφράστου βιβλίων, taken together by Sulla; and it was Andronicus, per Plutarch, who εἰς μέσον θεῖναι καὶ ἀναγράψαι τοὺς νῦν φερομένους πίνακας — put them out and drew up the tables now in circulation.[^10] The division of the one lot into two names is, on the receiving texts' own account, the editor's πίναξ.
The received reading does not deny any of this. It concedes that the fragment continues Λ's inquiry — that is what "a pupil's aporiai on the master's system" means — and holds that the change of name marks a change of hand. It is entitled to hold that; but it must then show the change on the page, and the page gives none. There is no marker at the hinge of a second person taking up a first person's work: no φησί, no ἐκεῖνος, no he says against I ask. The hinge is written in the inquiry's own voice — ἄρτιος ὁ λόγος, the account we have been giving — and the questions that follow are asked from inside it.
Three readings of the division are then available, and the seam's placement decides between two of them. The first: a change of person — the received reading. The second: a librarian's partition — Andronicus, or whoever drew the πίνακες, sorting one lot into two names where rolls broke or shelves were arranged. The third: the author's own editorial division of the work, cut at the seam of named authorial function — the achieved account assigned to a sealing name, the working continuation to a working name. A librarian's cut falls where the material happened to divide, arbitrary relative to the work's internal state. An author's cut falls at a functional boundary. And this cut falls at the one sentence that changes the account's status: ἄρτιος ὁ λόγος under one name, λόγου δεῖται πλείονος under the other. Complete goes to the seal; needs-more goes to the draft. That is not where a partition would fall and it is exactly where a self-editing author would cut — the name changing at the point the function changes, and the hinge announcing both.
Read as an instrument rather than a confusion, the ἤ has company throughout the dossier. The Second Letter states the instrument outright: οὐδ᾽ ἔστιν σύγγραμμα Πλάτωνος οὐδέν … τὰ δὲ νῦν λεγόμενα Σωκράτους ἐστίν — an author dividing the work by named function and saying so. Strato's will cuts the same way in law: τὰ βιβλία πάντα, πλὴν ὧν αὐτοὶ γεγράφαμεν, the library separated from our own writings, possession from authorial function. The name Θεόφραστος, a name for a voice, bestowed by the sealing position on the working one (μετωνόμασεν, D.L. V.38), is what naming a function looks like. And the catalogue carries, under both names, a title that names the function in so many words: Τὰ πρὸ τῶν τόπων — the things before the Topics — a pre-treatise position with its own entry.[^13] The twelve doublet titles acquire a test they did not have: if the Theophrastan member is systematically the preparatory, collecting, or aporetic version of the pair, the doublets are function-split editions of one work, not a school's shared notebooks.
Between the first reading and the third the page cannot decide: a pupil could have been assigned the function; a function could have been filled by one hand; the Greek gives a hinge and a name, not a body. But the comparison can be run. Function explains everything the page contains — the ἤ, the doublets, the bestowed name, the hinge's placement, the aporetic register on one side of the cut and the assertive on the other. Person explains only the name. Whatever the ἤ of the catalogue marks, the fragment's own sentences mark an editorial division of one work at the seam where its function changes.
Three things follow, and three do not. It follows that the fragment belongs to Λ's inquiry as its continuation, and should be read with Λ as one work reads with itself — Λ 10's aporiai and the fragment's as a single unfinished pass. It follows that the boundary between "Aristotle" and "Theophrastus" is, at this seam, a name-boundary drawn across a continuous text at the point where its function changes — on the better reading, by the work's own editor — and that the catalogue's ἤ is the record of the drawing. And it follows that the seam is a case for the general claim of Many and One — that the operation of dividing a corpus by name is an operation performed on the corpus, whose marks can be read — now caught on a text short enough to read whole.[^11] It does not follow that one hand wrote both; that Theophrastus is "in" any configuration; or that other Theophrastan texts stand in this relation to other Aristotelian ones. The Characters, De sensibus, and the botany are other seams, with their own counts, and the discipline that seats a candidate and holds him apart until each seam has been read applies without exception.
A marker of a second voice at the hinge — any grammatical sign that §6 reports rather than concludes. A misdescription of Λ 6–7 in §6, so that the "completed account" is not Λ's; it is not misdescribed. An extant Theophrastan work in which the fragment's aporiai are answered, making the fragment a Β to a Theophrastan Λ; none is extant. Or a collation of the fragment's text (Ross–Fobes, Gutas with the Arabic) that removes the hinge's wording; the paper's Greek is the seated First1KGreek text and awaits that collation.[^12]
The claim is small and, once seen, hard to unsee: a treatise closes a part, opens the next, tests it, and is left holding its own last questions — and at the turn of a page the name at the top changed, at the turn of the function. The catalogue wrote ἤ. So does the text. And so, on the reading the seam favours, did the author.
[^1]: The fragment is cited by the section numbers of the First1KGreek edition (tlg0093.tlg006), which runs 1–34; Λ by chapter with the seated canonical-greekLit text. Print collation against Ross–Fobes and Gutas, and against Jaeger's Λ, is pending throughout.
[^2]: §§6–7 as seated. The manuscripts' ἀραιρετέον/ἀφαιρετέον and the reading of ἡ ἀπόφασις at the end of §6 do not affect the hinge's form.
[^3]: Λ 6, 1071b3–5; Λ 8, 1073a14–15. Both from the seated Λ.
[^4]: Λ 7, 1072b13–14 (ἤρτηται); 1072a25–26 (ἀΐδιον καὶ οὐσία καὶ ἐνέργεια οὖσα); 1073a5–7 (ἀμερὴς καὶ ἀδιαίρετος); 1072b30 (τοῦτο γὰρ ὁ θεός). The fragment's §5 against Λ 7, 1072a26–27 (κινεῖ δὲ ὧδε τὸ ὀρεκτὸν καὶ τὸ νοητόν).
[^5]: Λ 8, 1073b13–17, immediately before the Eudoxus report (1073b17ff.) and Callippus (1073b32ff.).
[^6]: Λ 10, 1075b37–1076a4. The shared word is ἐπεισοδιώδης; the fragment's §2 has ἐπεισοδιῶδες τὸ πᾶν, Λ's ἐπεισοδιώδη τὴν τοῦ παντὸς οὐσίαν.
[^7]: Λ 7, 1072b20–21; fragment §25. Compare Θ 10, 1051b24–25 (θιγεῖν καὶ φάναι), where touching is the mode of grasping the simple, about which there is no falsity.
[^8]: The register and vocabulary counts are in the notebook The First Draft, §8: the fragment's aporetic density (ἀπορ-/ζητ
[^9]: D.L. V.42–50, the Theophrastan catalogue, seated on the Diogenes Laertius corpus (EA-CORPORA-12 #1580) and related in Catalogues and Wills (#1581), where the twelve cross-catalogue titles are listed.
[^10]: Strabo XIII.1.54; Plutarch, Sulla 26; both as seated and quoted in #1581.
[^11]: Many and One (#1576) makes the general claim for the Plato–Aristotle seam; this paper reads one continuous text across the Aristotle–Theophrastus seam.
[^13]: D.L. V.42–50 and V.22–27: Τὰ πρὸ τῶν τόπων is among the twelve titles standing letter-for-letter in both catalogues (#1581, §3.4). Topics I 14 and VIII describe the preparatory collection of ἔνδοξα the title would name.
[^12]: The one collation point that could matter is §6's ἄρτιος: the hinge's force does not rest on the adjective alone (μέχρι μὲν δὴ τούτων … τὸ δὲ μετὰ ταῦτ᾽ ἤδη carries it), but the adjective is the word that says complete.
Theophrastus, Metaphysics: First1KGreek tlg0093.tlg006, as seated on the archive's Theophrastus corpus (EA-CORPORA-09/01), read whole. Aristotle, Metaphysics Λ: Perseus canonical-greekLit tlg0086.tlg025, book 12, read whole; every locus cited checked against it. Diogenes Laertius, Strabo, Plutarch: as seated in EA-CORPORA-12 (#1580) and quoted in Catalogues and Wills (#1581). Translations mine. Print collation pending as noted.