Nothing Before Us: Socrates and the Ending of the Sophistical Refutations

Nothing Before Us: Socrates and the Ending of the Sophistical Refutations

Nothing Before Us

Socrates and the Ending of the Sophistical Refutations

Lee Sharks · v0.20 · 2026-09-07 · collations discharged against the public-domain prints (Bekker 1831; Hayduck 1884); Appendix L (the literature, by prohibition) added; body revised: §9 (the culmination) added; the abstract and §4 changed; Appendix P (what the text prohibits) added after a second reader's audit; §§2–3, 5–8 as at v0.17; Appendix A1a and E.8 seated at v0.17

Abstract

At the end of the Sophistical Refutations Aristotle denies that anything of his inquiry existed before it: οὐδὲν παντελῶς ὑπῆρχεν. In every version examined here the sentence is translated as it stands; where it is narrowed — to no systematic treatment, no completed art — the narrowing comes in the gloss, because everyone knows that Socrates asked questions before Aristotle answered them. I argue that the sentence means what it says; that the closing chapter is built, joint by joint, so that it can be kept; and that the figure the qualification exists to protect — Socrates as a predecessor from whom something of the inquiry was received — is a figure the chapter itself has already dissolved. Read with the treatise's own account of the peirastic, Socrates' asking is either the artless examining everyone shares, from which nothing is received, or something more, which the epilogue files under the inquiry's own purposes. On neither reading is he an antecedent source, and on the second — which I defend — he stands inside the beginning the first person recounts. The chapter supplies the vocabulary to settle his place and declines to use it: on the last page of a treatise about the qualifications readers supply, the reader is left to choose which word to move — and is, by that choice, examined. That examination is peirastic, conducted in the form of Socrates' own questioning; the ending performs his operation on its reader at the culmination of the art. What the chapter leaves open as an attribution it settles as an enactment: Socrates' questioning, as the ending performs it, is the completed art exercising its capacity on the reader — not the artless share of chapter 11 — while his possession of the art is left where the chapter leaves it. Nine prohibitions the text supplies, against which any rival reading may be checked, are anchored in Appendix P.

1. Introduction

The Sophistical Refutations ends with the most famous boast in the history of logic:

Ταύτης δὲ τῆς πραγματείας οὐ τὸ μὲν ἦν τὸ δ᾽ οὐκ ἦν προεξειργασμένον, ἀλλ᾽ οὐδὲν παντελῶς ὑπῆρχεν. (SE 34, 183b34–36)

Of this inquiry it was not the case that part had been worked out beforehand and part not; nothing at all existed.

Everyone knows the sentence cannot mean what it says. Socrates asked questions and refuted the confident; Plato wrote the questioning down and built an Academy around it; Zeno had arguments, the Megarians had arguments, the paid teachers of eristic had arguments for sale. The sentence closes a treatise whose examples are drawn from that world. So it is read — naturally, almost without the reader's noticing that a reading has occurred — as denying something narrower: no systematic treatment of reasoning, no completed theory, no art as opposed to a practice. The qualification is so standard that it passes for the text.

It is worth noticing where the qualification actually lives. Not in the translations. The translators, ancient and modern, render the sentence whole. The two Arabic versions of the tenth century give لم يكن منها شيء موجوداً البتّة, nothing of it existed at all, the negation sealed with البتّة; the Byzantine paraphrast copies Aristotle's words and adds one of his own, ὅλης — of this whole treatment, the totality reinforced; Pickard-Cambridge prints Nothing existed at all as a sentence standing alone; Forster gives it did not exist at all.[^1] The qualification enters one step later, in the gloss on what the faithfully rendered sentence means — and the earliest English commentary states the gloss's premise out loud. Edward Poste, note 9 to this chapter:

It is difficult to reconcile Aristotle's assertion with what we know had been done by Plato and Socrates and the Eleatics and Megarians. What he really performed in his dialectical treatise was to indicate a number of methodic principles or elements of method.[^2]

The premise is what we know; the sentence is what must be reconciled with it; and the note supplies what he really performed — whether as correction of the assertion or as its meaning, the note's own wording does not settle. What is common to both operations is the direction of travel: Socrates' position as Aristotle's predecessor is fixed before reading begins, and the sentence is made answerable to it — in its truth, if the note disputes; in its scope, if the note reinterprets.

This paper argues that the sentence means what it says; that the chapter that contains it is constructed so that it can be kept; and that the term the qualification exists to protect — Socrates as a predecessor from whom something of the inquiry, however rudimentary, was received — is a term the chapter has already removed. So I shall argue. Nothing here impugns the fourth century. Socrates asked; the practice of examining existed; the paper asserts nothing about history beyond what the paragraph asserts, and the paragraph, we will see, asserts less about history than its readers do. I will also argue, controversially, that the better reading of the ending places Socrates inside the inquiry whose beginning it recounts — and that the chapter declines to say so outright, leaving exactly that question open for a reader it has just spent thirty-three chapters training to notice supplied qualifications. The argument proceeds by reading the last page once, whole, in order (§2); by putting the qualified reading to a single test it fails (§3); by locating the practice of examining where the treatise itself locates it (§4); by stating the case for inclusion (§5) and what the chapter does instead of stating it (§6); and by reporting what readers do at this sentence when no one is watching (§7).

2. The last page

The epilogue runs from 183a37 to 184b8, and it should be read as one movement, because it is one: an accounting. Its first sentence says what was undertaken:

Προειλόμεθα μὲν οὖν εὑρεῖν δύναμίν τινα συλλογιστικὴν περὶ τοῦ προβληθέντος ἐκ τῶν ὑπαρχόντων ὡς ἐνδοξοτάτων· τοῦτο γὰρ ἔργον ἐστὶ τῆς διαλεκτικῆς καθ᾽ αὑτὴν καὶ τῆς πειραστικῆς. (183a37–b1)

We proposed to discover a certain capacity of reasoning about any problem put forward, from the most reputable of the materials available; for this is the work of dialectic in itself and of the peirastic.

A capacity, δύναμις — not a doctrine, not a book — and the work it is a capacity for is named twice: dialectic's, and the peirastic's.[^9] The inquiry then reports a second undertaking, and the sentence should be kept whole — one participle in it will matter: since dialectic, by its neighbourhood to sophistic, is equipped to take an examination not only dialectically but also as one who knows — οὐ μόνον πεῖραν δύναται λαβεῖν διαλεκτικῶς ἀλλὰ καὶ ὡς εἰδώς — the inquiry set itself not only the stated work but also how, sustaining an account, we shall guard the thesis: λόγον ὑπέχοντες φυλάξομεν τὴν θέσιν (183b1–6). Answering, then, is played as one who knows — ὡς εἰδώς: a role the treatise assigns, not an epistemic promotion it promises. The cause has been stated; and then, at once:

Τὴν δ᾽ αἰτίαν εἰρήκαμεν τούτου, ἐπεὶ καὶ διὰ τοῦτο Σωκράτης ἠρώτα ἀλλ᾽ οὐκ ἀπεκρίνετο· ὡμολόγει γὰρ οὐκ εἰδέναι. (183b6–8)

We have stated the cause of this, since it was also for this reason that Socrates used to ask but not answer; for he admitted that he did not know.

One man, named, inside the recital of the inquiry's purposes, his conduct set under the same cause — καὶ διὰ τοῦτο, also for this reason — that the inquiry has just given for its own second task. The echo is exact: the answerer's role was just defined ὡς εἰδώς; Socrates would not take it, ὡμολόγει γὰρ οὐκ εἰδέναι — the participle of the role answered by the infinitive of the refusal, two lines apart. Then the inventory: how many aims, out of how many materials, where the resources come from, how to ask and how to arrange the whole questioning (πῶς ἐρωτητέον ἢ τακτέον τὴν ἐρώτησιν πᾶσαν, 183b10–11), how to answer, how to solve — the method's contents ticked off, asking included.

Then the transition, in the treatise's own verb for what readers miss — δεῖ δ᾽ ἡμᾶς μὴ λεληθέναι τὸ συμβεβηκὸς περὶ ταύτην τὴν πραγματείαν, we must not let what has happened concerning this inquiry escape our notice (183b16–17) — and the paragraph turns to what precedes discoveries in general, drawing a line through all of them:

τῶν γὰρ εὑρισκομένων ἁπάντων τὰ μὲν παρ᾽ ἑτέρων ληφθέντα πρότερον πεπονημένα κατὰ μέρος ἐπιδέδωκεν ὑπὸ τῶν παραλαβόντων ὕστερον, τὰ δ᾽ ἐξ ὑπαρχῆς εὑρισκόμενα μικρὰν τὸ πρῶτον ἐπίδοσιν λαμβάνειν εἴωθε... (183b17–21)

Of everything that is discovered, some things, taken over from others who had laboured on them before, have advanced part by part in the hands of those who received them later; while things discovered from the very beginning are accustomed to make small progress at first...

Two provenances, not two sizes. Received and grown by inheritors; or found from the beginning, small at first, the beginning itself the greatest part and the hardest to see (183b22–26). Rhetoric is the worked example of the first class: its founders advanced it altogether a little way, and the celebrated moderns, receiving it from many as if in a succession — παραλαβόντες παρὰ πολλῶν οἷον ἐκ διαδοχῆς — enlarged it: Tisias after the first, Thrasymachus after Tisias, Theodorus after him, and many contributed many parts; no wonder the art has bulk (183b26–33).

Then the sentence. Of this inquiry: not part worked out beforehand and part not — nothing at all was there.

The next word is καὶ γάρ, for in fact: the paragraph immediately gives the reason its denial is true, and the reason is the one place a reader might think it false. There were professionals — τῶν περὶ τοὺς ἐριστικοὺς λόγους μισθαρνούντων, people paid to teach eristic argument. Their instruction resembled Gorgias' πραγματεία — ὁμοία τις ἦν ἡ παίδευσις τῇ Γοργίου πραγματείᾳ; hold the noun — speeches handed out to be memorized, rhetorical from some, question-form from others. And so — διόπερ — their teaching was quick but artless, ταχεῖα μὲν ἄτεχνος δέ (184a1–2):

οὐ γὰρ τέχνην ἀλλὰ τὰ ἀπὸ τῆς τέχνης διδόντες παιδεύειν ὑπελάμβανον, ὥσπερ ἂν εἴ τις ἐπιστήμην φάσκων παραδώσειν ἐπὶ τὸ μηδὲν πονεῖν τοὺς πόδας, εἶτα σκυτοτομικὴν μὲν μὴ διδάσκοι... δοίη δὲ πολλὰ γένη παντοδαπῶν ὑποδημάτων· ...βεβοήθηκε μὲν πρὸς τὴν χρείαν, τέχνην δ᾽ οὐ παρέδωκεν. (184a2–8)

For they supposed they educated by giving not the art but the products of the art — as if someone professing to hand over knowledge that the feet should never hurt were then not to teach shoemaking... but to hand out many kinds of shoes of every sort: such a person has helped meet the need, but has not handed over an art.

Then the last accounting, rhetoric against reasoning:

καὶ περὶ μὲν τῶν ῥητορικῶν ὑπῆρχε πολλὰ καὶ παλαιὰ τὰ λεγόμενα, περὶ δὲ τοῦ συλλογίζεσθαι παντελῶς οὐδὲν εἴχομεν πρότερον ἄλλο λέγειν, ἀλλ᾽ ἢ τριβῇ ζητοῦντες πολὺν χρόνον ἐπονοῦμεν. (184b1–2)

And about rhetoric there existed many old things said; but about reasoning we had absolutely nothing earlier else to say — except that, searching by trial, we laboured a long time.

And the envoi: judge the method as one that stands from such beginnings, ἐκ τοιούτων ἐξ ἀρχῆς ὑπαρχόντων, against the other treatments grown by transmission, τὰς ἐκ παραδόσεως ηὐξημένας — and grant the omissions pardon and the discoveries much thanks (184b3–8).

That is the whole page: an undertaking named as a capacity; a second task and its cause; Socrates under the cause; a division of all discoveries by provenance; rhetoric filed on the received side with its chain of names; the denial; the professionals as the denial's proof; the shoes; the count of things said; the envoi on the same axis the division opened. Everything the following sections use is already visible in it.

3. What kind of nothing

Here is a test the qualified reading must pass. If the sentence at 183b34–36 denied completeness — if οὐδέν meant nothing systematic, nothing finished — then the paragraph around it should traffic in degrees of development: more and less elaborated, rudimentary and mature, sketch and system. It does not. From 183b17 to 184b8 the paragraph's one axis is provenance: taken from others or found from the beginning; παρ᾽ ἑτέρων against ἐξ ὑπαρχῆς; ἐκ διαδοχῆς and ἐκ παραδόσεως against ἐξ ἀρχῆς. Where degree appears at all — the small first progress of new discoveries, rhetoric's founders advancing it a little — it appears inside the provenance classes as their behaviour, never as a scale on which the two classes are points. The reading that hears nothing complete has imported a dimension the paragraph does not use, to a sentence built entirely of the dimension it does. Every joint of the sentence confirms this, and each will bear one look.

The negation. οὐ τὸ μὲν ἦν τὸ δ᾽ οὐκ ἦν προεξειργασμένον: what is denied is a mixed condition — part worked out beforehand, part not. But the mixed condition is precisely the qualified reading's own picture of the fourth century: some elements of the method anticipated, the system lacking. The sentence's first clause denies the picture the reading offers as the sentence's meaning. A careful defender will reply that Aristotle denies the mixed condition while meaning only that nothing complete existed. Then two questions have no answer. Why deny the mixed condition, which is denied by denying that anything was worked out at all, if only completeness were at stake — completeness is denied by denying completeness? And why does ἀλλά then introduce not a reformulation about development but an existence claim?

The verb. Because it does introduce one. The participle of the first clause is προεξειργασμένον, worked out beforehand — the vocabulary of elaboration, and it stays inside the negation. The verb of the second clause is ὑπῆρχεν. ὑπάρχειν is not to have been elaborated; it is to be there — to be available, to belong, the Analytics' own workhorse for what holds of what. Nothing at all was there. Greek contrast can, of course, reformulate across a change of predicate; the argument is about cost and warrant, not possibility. The asserted clause supplies its own finite verb and no participle; completely worked out enters only if a reader carries the negated member's predicate across ἀλλά and reattaches the adverb to it — two operations on a clause of four words, neither marked in the clause itself. And the treatise's own usage shows the import is no default. At the close of chapter 20: οὐχ ἁρμόσει ἐπὶ πάντας οὐδὲ πάντως ἐρωτωμένους, ἀλλ᾽ ἔστι πρὸς τὸν ἐρωτῶντα, οὐ πρὸς τὸν λόγον (177b33–34) — will not fit negated, the ἀλλά-member asserting its own predicate, and importing ἁρμόσει would invert the sense, since the point is that the solution addresses the questioner and precisely does not fit the argument. The epilogue keeps its own control in its last lines: ὑπῆρχε on one side of the μέν/δέ at 184b1–2, εἴχομεν … λέγειν on the other, and no reader carries either verb across. One instance and one control are not a survey;[^3] the standing claim is the narrower one: nothing in the syntax requires the import, and the import is the costlier construal.

The adverb. παντελῶς is not evidence for the import; it recurs, thirty lines later, in the reversed order — παντελῶς οὐδὲν εἴχομεν πρότερον ἄλλο λέγειν (184b1) — where no construal yields nothing complete: whatever the adverb intensifies there, it intensifies a negation of having anything to say. The pair οὐδὲν παντελῶς / παντελῶς οὐδέν brackets the passage like a signature. It seals totality; it never supplies worked out.

The proof. καὶ γάρ. The sentence about the paid teachers is offered as the reason the denial is true — the one standing counter-example, pre-empted. And the terms of the pre-emption should be read slowly, because they are the paragraph's own transmission vocabulary turned off. What the received class does is παραλαμβάνειν and παραδιδόναι — taking over, handing on; ἐκ διαδοχῆς; ἐκ παραδόσεως. What the professionals conspicuously failed to do is stated in that verb: τέχνην δ᾽ οὐ παρέδωκεν — did not hand over an art. The chapter's one named set of would-be predecessors is denied precisely παραδοῦναι. There were people before; there was no handing-on, because what they gave was not the kind of thing whose giving is a transmission.

The shoes. Which is what the analogy fixes. οὐ τέχνην ἀλλὰ τὰ ἀπὸ τῆς τέχνης: not the art but the things from the art. A shoe is not a small amount of shoemaking; a hundred shoes are not more shoemaking than one. The contrast is of kind — the capacity against its products — and the sentence draws it in order to say why product-giving transmits nothing. Note what the analogy does not say: it does not say that no one made shoes before there was a theory of shoemaking. It is silent about untutored practice, and the silence matters below (§4). What it closes is one door only: that the products in circulation constituted a partial possession, a partial handing-on, of the art. That is also the whole of its evidential scope — the analogy fixes what counts as transmitting; what ὑπῆρχεν denies is fixed by the frame around it.[^4]

The count. At 184b1–2 the paragraph says what its ὑπάρχειν has been counting. On rhetoric's side, ὑπῆρχε... τὰ λεγόμενα — things said were there, many and old. On reasoning's side, first person: we had nothing earlier else to say — ἄλλο λέγειν — except the long trial. The claim is kept distinct from 183b35's and should be kept so: not nothing existed but we had nothing to take up and say; the exception clause names what stood where sayings would have stood — τριβῇ, labour by rubbing, trial. And the two claims are parallel by design: λεγόμενα against λέγειν, existence of sayings against availability of anything to say. What was countable beforehand, had there been anything, was discourse available for taking up. Of that: none.

The frame. The envoi returns to the opening division in its own words — ἐξ ἀρχῆς ὑπαρχόντων against ἐκ παραδόσεως ηὐξημένας — so that the disputed sentence sits between two statements of the same axis, the axis of provenance, and the last word of the frame is the transmission-word the shoemaker was denied.

So the test's result. The paragraph does not prove that no other dimension can be supplied; paragraphs prove nothing about quantifiers. What it does is define before in provenance terms in the sentences before the denial and return to them in the treatise's last line — so that a reader who hears nothing systematic must do two things at once: shift the dimension the paragraph has just fixed, and import a predicate the clause does not contain. Two operations at once, and the text licenses neither. The domain, meanwhile, was never open: it is the sentence's own subject, ταύτης τῆς πραγματείας. Which raises the reading's second and better route: keep the quantifier total, and determine the noun. Nothing of the inquiry existed; the practice — Socratic asking above all — lay outside it, before it, the raw material a theory later organized. Whether that route is open depends on one word, and the next section begins with it.

4. The activity and the art

Everything now runs through one word, and the chapter has been using it. ἡ πραγματεία is what we chose — Προειλόμεθα (183a37); it is the thing to which the inquiry assigned its works — οὐ μόνον τὸ λεχθὲν ἔργον ὑπεθέμεθα τῆς πραγματείας (183b3–4); it is what has had something happen to it that must not escape notice — τὸ συμβεβηκὸς περὶ ταύτην τὴν πραγματείαν (183b16); and it is the denial's subject — ταύτης τῆς πραγματείας (183b34). Grammar first: a πραγματεία is chosen, assigns itself tasks, has a history. It is an undertaking, not a field. But the decisive control on the word stands inside the proof-sentence itself: the teachers' instruction, says καὶ γάρ, was like Gorgias' πραγματεία — τῇ Γοργίου πραγματείᾳ (183b36–37). Gorgias had one. It existed, it is named, and its prior existence is adduced to explain why this one had nothing before it. So the noun cannot denote the practice of arguing in general — that existed in everyone — and cannot denote organized engagement with arguments as such — Gorgias' was exactly that, and the sentence's own proof presupposes it. πραγματεία individuates undertakings; the demonstrative selects one; and the one selected is individuated for the reader by its stated ἔργον — to find the συλλογιστικὴ δύναμις, and to sustain an account ὡς εἰδώς. Three things stand apart in the chapter's own vocabulary: the activity of examining, which is everyone's (172a30–36); undertakings upon it, of which Gorgias' product-dispensing πραγματεία is the named, existing, non-transmitting instance, disposed of by the shoe; and this undertaking, the ἔργον-defined inquiry, of which alone the sentence says οὐδὲν ὑπῆρχεν. The first existed universally and threatens nothing. The second existed and is the sentence's proof, not its counter-example. The denial's subject is the third.

The serious form of the qualification is not τέχνη means a book. It is: τέχνη here names an organized, teachable, transmissible discipline as against untheorized practice; the sentence denies the discipline; the practice existed and Socrates is its famous case. Two passages meet it, and together they end it.

The first is the shoemaker again, read for what its contrast is. The passage does not oppose the organized to the raw; it opposes the art to its products, and the failed teacher fails not because his material was unorganized but because products are not the kind of thing whose transfer conveys a capacity. The defender's contrast — discipline versus untheorized practice — is not on the page. And untheorized practice is not left floating for the defender to locate; the treatise has located it, and its scope is everyone.

The peirastic enters in chapter 2 as one of the four kinds of argument in discussion: arguments from the answerer's own opinions, wielded by an examiner who need not know the subject, against one who professes to (165b4–6). Chapter 8 calls it a part of dialectic (169b25); chapter 11, a kind of dialectic, διαλεκτική τις (171b4) — and then says whose it is:

διὸ πάντες καὶ οἱ ἰδιῶται τρόπον τινὰ χρῶνται τῇ διαλεκτικῇ καὶ πειραστικῇ· πάντες γὰρ μέχρι τινὸς ἐγχειροῦσιν ἀνακρίνειν τοὺς ἐπαγγελλομένους... ἐλέγχουσιν οὖν ἅπαντες· ἀτέχνως γὰρ μετέχουσι τούτου οὗ ἐντέχνως ἡ διαλεκτική ἐστι. (172a30–36)

Hence all people, even laymen, in a way use dialectic and the peirastic; for all, up to a point, undertake to cross-examine those who profess... So all refute; for artlessly they share in that of which dialectic is the artful possession.

Linger in the two adverbs, because the whole dispute is inside them. μετέχουσι: they share, partake — the activity is one, and everyone is in it. ἀτέχνως... ἐντέχνως: two modes of having the same thing — without the art, with it. Untheorized practice of examining is not a special antechamber where Socrates waits for Aristotle; it is the condition of ἅπαντες, everyone, the treatise's own word. And note what the universal share does not establish. The division at 183b17 is a division τῶν εὑρισκομένων, of things found, taken over and advanced; universal, artless participation is not, by those words, a discovery anyone made or a thing anyone took over — chapter 11 says only that all share in it, ἀτέχνως, and universality does not by itself entail a finding, a transmission, or their absence. What it does establish is enough: universal participation does not itself constitute a transmitted contribution to this undertaking. The epilogue supplies the denial; chapter 11 prevents mere participation from functioning automatically as its counterexample. The artless share stands outside the epilogue's taxonomy of provenance not as an oversight but because participation and provision are different things in the treatise's own terms — which is exactly why its prior existence never threatened οὐδὲν παντελῶς ὑπῆρχεν in the first place. Nothing of the πραγματεία is in it.

Now the fork, and it is the treatise's, not mine. One firebreak: nothing in it claims that Socrates possessed the art, or that he stands inside the first-person plural by grammar. The claim is structural — the treatise gives a practitioner exactly two places to stand, and the epilogue gives this practitioner a treatment it gives no other. When the epilogue names Socrates, what is he? Either his asking is an instance of the universal artless examining — in which case nothing of the inquiry was received from it, no finding and no transmission, the denial stands whole, and he stands untouched by it, one of the ἅπαντες. Or his asking is more than an instance of everyone's share — as the naming suggests, and the cause, and the seat in the recital — in which case the epilogue has set it inside the inquiry's own account of what it undertook. On neither horn is he the figure the qualified reading requires: a predecessor from whom something of the inquiry, however rudimentary, was received. The received relation is the one relation the chapter denies of the inquiry (183b34–36) and the one relation it never states of him. It follows that the two terms the received picture cannot reconcile — nothing, total, and Socrates, before — cohabit without strain the moment "before" stops meaning "source".

A middle position deserves its own name — many careful readers stand on it: Socrates as a historically significant precursor — neither a transmitter of the inquiry nor an undifferentiated member of everyone. The position is coherent. First, what it concedes: it keeps οὐδέν total. A significant precursor from whom nothing was received contradicts no clause of the sentence; whoever stands here has already left the qualified reading; the negative thesis counts them allies. What remains against the position is a question, not a refutation: significant how? The epilogue does not leave the significance blank; it specifies it, and the specification is a cause — the inquiry's own, for its second task — and a seat in the recital of its purposes. Significance located there is the inclusion-shaped kind, and §5 weighs it. Significance located elsewhere — fame, influence, the regard of the history of philosophy — is not on this page. The middle position collapses toward one horn or the other as soon as its adjective is given content from the page; and on every such content it keeps the sentence. Nothing was received; Socrates asked; both stand.

And once both stand, the qualification is out of work. Its function was never semantic — the Greek gave it nothing to do. Its function was custodial: to keep Socrates on the far side of a border the chapter never draws, at the price of adjusting the one sentence in the chapter built to resist adjustment.

5. Socrates in the beginning

I will also argue, controversially, that the ending does more than leave Socrates standing untouched; it draws him in. The claim is not that the grammar compels this — it does not, and I will say below what the chapter does instead of compelling it. The claim is that inclusion is the reading on which every sentence of the last page keeps its face value, and that its rival pays a bill it has not noticed.

Four things, and their conjunction. First, the naming. The universal artless share needs no name and gets none anywhere in the treatise; ἅπαντες is its name. To name one practitioner out of everyone is to select, and the epilogue selects once. Second, the cause. Socrates' conduct is introduced by ἐπεὶ καὶ διὰ τοῦτο — since also for this reason — where τοῦτο is the cause the inquiry has just given for its own second undertaking. His asking-without-answering and the inquiry's decision to master answering are filed under one αἰτία. That is not a claim that his practice caused the inquiry; it is where the sentence puts him — under the inquiry's purposes, inside its recital of what it set out to do and why, in the position the paragraph gives to no historical figure besides him. Tisias gets a chain; Socrates gets a reason, and it is the inquiry's own. Third, the order. The recital presents the answering task first (183b4–5) and Socrates second (183b6–8), and later the trial (184b2); and it has keyed the two roles to knowing in its own words — the answerer answers ὡς εἰδώς (183b3) — the ὡς marking a role taken, not a state achieved — and Socrates declined the role, ὡμολόγει γὰρ οὐκ εἰδέναι (183b8). But the relation among the three activities runs the other way from the presentation, and the chapter has given us the terms to say so. Examining without knowing — the peirastic's defining case, and Socrates' stated one: ὡμολόγει γὰρ οὐκ εἰδέναι — is where an inquiry that ends in the capacity to answer begins; the long trial, τριβῇ ζητοῦντες πολὺν χρόνον, lies between. From the one who asked and did not know, to the ones who answered, through the labour: that is an order of grasp, and it is not a chronology of persons — a mode, then a capacity, with the work in the middle.[^5] Fourth, the vacancy. In every treatment ἐκ παραδόσεως, the space before the moderns is occupied by received material and its chain of names. In this one, the space is occupied by first-person labour — εἴχομεν, ἐπονοῦμεν — the "we" toiling exactly where the παραλαβόντες would stand. The paragraph does not merely deny inheritance; it shows what filled inheritance's place, and it speaks of the filling as ours.

The rival reading — Socrates as famous instance, a vivid case and no more — has been kept available throughout, and it is not empty. An example can carry explanatory weight: the vivid case of asking that would not answer explains what limitation the undertaking's second task addresses. Grant that function fully. The comparison that decides between the readings is then explanatory, not eliminative: what does inclusion explain that exemplary relevance leaves unexplained? This — the relation the last page asserts among its terms. On the exemplary reading, the illustration, the added task, and the first-person labour stand side by side, their arrangement unexplained: why the illustration sits inside the recital of purposes rather than beside it; why its cause is the inquiry's own (καὶ διὰ τοῦτο); why the role he declined is the very role the added task masters; why the space where received material stands in every other art is filled here by we labouring. On the inclusive reading the four are one movement: the inquiry, recounting its own constitution, begins its account with the questioner whose refusal of the answerer's role is the shape of its starting point, and the task, the labour, and the capacity follow as the account's remaining clauses. The middle position of §4 meets the same comparison: its significant must be located on the page, and the page locates it in the cause and the seat — the inclusion-shaped places. On the inclusive reading the denial is total, the man is named, the cause is shared, and the labour is we; I know of no sentence on the page that it disturbs, and none whose placement it leaves a coincidence.

And the fork has a witness outside the treatise — not a scattered pattern of honours, but one passage that draws the fork itself. Metaphysics Μ 4, pausing on Socrates before the theory of Forms, should be read in its own order (1078b23–32). Socrates busied himself with the ethical virtues, seeking first to define universally concerning them. Why definitions? — συλλογίζεσθαι γὰρ ἐζήτει, for he was seeking to syllogize, ἀρχὴ δὲ τῶν συλλογισμῶν τὸ τί ἐστιν, and the what-it-is is the starting-point of syllogisms. And why was the seeking still to do? — διαλεκτικὴ γὰρ ἰσχὺς οὔπω τότ᾽ ἦν, for dialectical capacity did not yet exist. Then the assignment: δύο γάρ ἐστιν ἅ τις ἂν ἀποδοίη Σωκράτει δικαίως, τούς τ᾽ ἐπακτικοὺς λόγους καὶ τὸ ὁρίζεσθαι καθόλου· ταῦτα γάρ ἐστιν ἄμφω περὶ ἀρχὴν ἐπιστήμης — and, at once, the separation from what the Platonists made of it: ὁ μὲν Σωκράτης τὰ καθόλου οὐ χωριστὰ ἐποίει.[^10] Hold the clauses together and they are the configuration argued here, stated by Aristotle as history: the man engaged in the operations — seeking to syllogize, working the starting-point of syllogisms — while the capacity, the ἰσχύς, was not yet there; two contributions justly assigned to him; their misdevelopment charged to others. οὔπω ἦν of the capacity is the Metaphysics' tense for what ὑπῆρχεν denies of the undertaking whose stated ἔργον is to find that δύναμις. And the ἀρχή at which Μ 4 fixes his work is bound by the passage itself to syllogisms — the object of the epilogue's own count, περὶ τοῦ συλλογίζεσθαι (184b1). That the starting-point of a science and the beginning of this undertaking correspond is an inference; its license is Μ 4's own chain — the man seeking to syllogize, at the ἀρχή of syllogisms, before the capacity existed — set beside an epilogue that names the same man under the cause of the task, denies that anything of the capacity-finding undertaking preceded it, and speaks the finding in the first person. Α 6 keeps the family verb — Σωκράτους περὶ μὲν τὰ ἠθικὰ πραγματευομένου — and then records an uptake: Plato, ἀποδεξάμενος, taking up the definitional concern. Uptake from Socrates is real in the corpus, and it is Plato's, of a concern; nowhere is it this undertaking's, of its material.[^10] The bare-instance reading must read Μ 4 as one more coincidence of emphasis. The inclusive reading finds in it the same picture the epilogue gives — told there as history, told here as the inquiry's own constitution.

At which point the burden changes hands — not to the vanishing of the rival, which keeps the function granted above. While the transmitting placement stood, it explained the history, and inclusion argued uphill. With that placement closed — and the closing is the part of this paper its hardest readers grant — two furnished readings remain, and the question is the section's own: which leaves less unexplained? Exemplary relevance explains the naming and its local point, and stops there; the four arrangements above, and now the Μ 4 concord, stand outside its account — coincidences it can only accumulate. Inclusion explains each by the same clause of itself. A reader may still prefer the exemplary reading; the preference owes its own account of five accidents, or a third placement, and §4 has walked the ground where a third would have to stand. The paper asserts inclusion because, by the only measure the page provides — what a reading leaves unexplained — it is the better one.

Three claims end this section at their strengths. That Socrates is placed inside the explanation of the inquiry's second task, the sentence establishes. That his position is the structural beginning — questioning without the profession of knowledge, from which the task of answering ὡς εἰδώς takes its point — the role contrast, the order, and the Μ 4 concord — the seeker of syllogizing at the ἀρχή of syllogisms, before the capacity existed — argue, and I assert. That he stands inside the first person's referent, nothing here demonstrates, and nothing here claims it as demonstrated; it is offered as the best reading of a space the next section shows the chapter furnishes and declines to close.

6. What the chapter does instead of saying so

But the chapter does not say it — and the manner of its not saying is the last thing to read.

The epilogue possesses, visibly, on this very page, every instrument needed to fix Socrates' place. It has the transmission vocabulary: παρ᾽ ἑτέρων, ἐκ διαδοχῆς, ἐκ παραδόσεως — freshly used, three lines of names attached. It says none of it of Socrates. It has the origination vocabulary: ἐξ ὑπαρχῆς, ἐξ ἀρχῆς. It says neither of him. It has the first person, twice, in the very sentences at issue. It does not say whether he is inside it. It has the universal, ἅπαντες, from chapter 11. It does not say he is only that. The chapter that welds every joint around οὐδέν — negation, verb, adverb, proof, analogy, count, frame — leaves the person unfastened, having laid out the fasteners in a row.

In another treatise this would be ordinary silence, and I would make nothing of it. This treatise is about supplied qualifications. Its catalogue is the moves by which readers and answerers convert sentences into the sentences they can accommodate: the ἁπλῶς taken πῂ, the qualification smuggled in (ch. 5, 25); the divided taken together (ch. 4, 20); the answer given to a question not asked; the refutation performed on oneself without perceiving it (ch. 16). Its last page then sets before the reader one sentence fixed beyond adjustment and one person left without a place, in a configuration where the reader's standing habit — Socrates the predecessor — cannot be kept unless the fixed sentence is loosened. A reader who resolves the person's place owes an account of why the chapter furnished the alternatives and chose neither; §5 gave mine, and a defender of the vivid-instance reading owes theirs. But a reader who resolves it the received way — the person fastened outside, the sentence adjusted to fit — performs, on the last page of the Sophistical Refutations, the operation the treatise teaches: supplies to a text the qualification the text was written without, and does not perceive the supplying. I do not claim the chapter was designed as a trap; structures do not need intentions to function, and the claim of design is not readable off the page. What is readable is the function. A treatise on the detection of unperceived operations, ending on a sentence whose easiest domestication is such an operation, is its own final exercise — whoever set it.

7. The readers

How the exercise has gone can be reported briefly. In every version I have examined, the translators keep the sentence total: البتّة in both Arabic versions, ὅλης added by the paraphrast, Nothing existed at all, it did not exist at all.[^6] Where the qualifying vocabulary first enters the English chapter is one line after the sentence, at ἄτεχνος — the teachers' instruction was rapid but unsystematic (Forster), rapid but unscientific (Poste) — attached to the professionals' pedagogy. Whether the word then migrated from the instruction to the gloss on οὐδέν is a hypothesis about transmission, not a documented step; what is documented is position: the vocabulary of system enters the English chapter one line after the sentence it is later used to explain, and never inside it. The commentary states its premise where Poste stated it: from what we know to what he really performed — the placement fixed first, the sentence made answerable to it, in truth or in scope, whichever the note is doing. The Latin thirteenth century sharpens that boundary. Aquinas, who filed the treatise in his division of logic and glossed at length the Metaphysics' gratitude to predecessors — accepimus a praedecessoribus — never brings that text and this sentence into contact; no reconciliation is attempted, and whether any tension was felt the silence cannot say (Appendix A, C2). The two scholastics who do read the chapter keep the denial whole, and their narrowing phrases attach to transmission. Albert's paraphrase renders the sentence total in both members — sed nihil omnino erat inventum a prioribus ante nos — stages the one Platonic candidate his (mis)reading supplies, the Gorgias, and dissolves it with Aristotle's own verb, artem non tradidit; his sole qualification is per artem traditam … nisi exempla quaedam particularia (C3). Giles construes the boast twice as nihil traditum per modum artis, non adiutus prioris auxilio, and builds the chapter's praise on the totality: specialis inventor (C4). Per modum artis is not innocent of a conception of art — Albert's shoemaker fails secundum principia artis, by universal precepts ordered to one end — so these witnesses establish less than confirmation and more than silence: emphatic denial and explicit attention to the questioners coexist in them without repair, and the qualifying phrase, where it comes, rides the transmission verbs. Both key answering to those who know — responsio opus et actus scientis (Albert); soli scientes possunt observare ne respondendo sibi contradicant (Giles) — the role contrast of 183b1–8 audible in both. Among the witnesses examined here, the degree-reading — nothing systematic — is not attested before the modern period. And the pattern has a present tense. A recorded sample of machine referee reports, generated on successive versions of a note asserting only what the sentence asserts, restored the qualification in every terminal report and moved the man in none; the design was responsive, so the sample shows the operation's persistence under objection in that environment — not spontaneous convergence — and it arbitrates nothing about the reading; sample, prompt, coding, and registers are in the note.[^7] What the file as a whole shows is a divide, not a chorus. The witnesses who qualify at all divide by period and by axis: the medieval readers, where they narrow, narrow in transmission vocabulary and move nothing; the degree-qualification, among the witnesses examined, is not attested before the modern period. The moderns move the word the text fixed.

8. Which word moves?

Every reading of this ending moves a word. The received reading moves οὐδέν — to nothing systematic, nothing complete, nothing of the theory — and holds the person still. I have argued that the text will not give up the word: the paragraph builds its denial out of provenance and welds every joint, and the treatise has already housed untheorized practice in everyone, where its priority threatens nothing. What the text will give up — what it holds out, unfastened, with the fasteners lying beside it — is the person. Move him instead: out of the chain he was never in, into the beginning the first person recounts, where the one who asked and did not know stands at the head of an order of grasp that ends in answering — the seat Μ 4 assigns him, seeking to syllogize at the ἀρχή of syllogisms before the ἰσχύς existed; or, if not that far, then at least out of settled and into open, which is where the chapter leaves him. Either movement keeps the sentence. Only the received reading's movement costs one.

And a last question, opened rather than defended. The epilogue states an order — the asker who did not know; the trial; the answerers — and for two millennia that order has been read as a chronology of two lives, the master's and the student's student's. If, on the last page of the Organon, the order of grasp and the order of time come apart — if before can mean at the beginning of rather than prior to and outside — then the question asked here of one sentence stands ready for larger texts: where else in the corpus is an order of understanding being read as a sequence of biographies? That question is pursued elsewhere.[^8] Here it is enough to have read one page whole. Nothing was before us. Socrates asked. The chapter keeps both — and waits to see which word its reader will move. And in waiting, it does one more thing, which §9 reads: it asks.

9. The culmination

The last question can be closed, and on the treatise's own strongest terms — not as an attribution the chapter makes, but as an operation the chapter performs.

Take the two things established so far and put them together. First, §6: the chapter possesses the vocabulary to settle Socrates' place and does not use it; what it does instead is leave the reader a sentence whose easiest domestication is the operation the treatise teaches the reader to detect, so that the last page is the treatise's final exercise on whoever reads it. Second, §4: the treatise's own account of the examining art has two modes of the same activity — ἀτέχνως, the artless examining everyone shares, and ἐντέχνως, the art — and the epilogue names the art's method as questioning and answering, with Socrates' questioning connected to that method's own explanation (183b6–8, P2). Now ask what form the final exercise takes. It is a question put to the reader — where does Socrates fall? — which the reader must answer, and by whose answer the reader is examined: the one who supplies the adjective is refuted by the sentence they altered; the one who keeps the sentence has been brought, by asking, to see what it holds. That is peirastic. It is examination by question of a respondent who believes they know. It is Socrates' operation — ἠρώτα ἀλλ᾽ οὐκ ἀπεκρίνετο — performed on the reader by the ending of the treatise.

So the question the chapter leaves open as an attribution it answers as an enactment. Socrates' questioning supplies the form of the examination the ending performs on its reader. He is doing constitutive work within the art: the reader encounters his operation in undergoing the examination of his place. Whether he possessed the τέχνη is left unsettled, as §6 says it is; where his questioning stands is settled by where the treatise uses it — inside the art's performance, at the point where the art exercises itself most fully. And that point is not the beginning. It is the culmination. Aristotle's completed art, at the end of the Organon, turns its examination upon Socrates' place, and in doing so performs Socrates' questioning: the asker who did not answer appears at the achievement of the undertaking, as the shape of its final act.[^11]

This is the literal form of nothing before us. The we of προειλόμεθα and εἰρήκαμεν is not a lineage reaching back through a predecessor (P3, P5, P9); it is performed on the last page — the art examining you who have heard (184b6–8) in the form of that man's questioning — and Socrates is inside it not as an origin from which something was received but as the operation the finished art executes. The three positions of §5's triangle are on one page: the first person that undertook, the hearers who will pardon the omissions, and the one who asked; and the second is examined by the first in the form of the third.

It follows, and this closes the reading, that Socrates' questioning, as enacted by this ending, is not artless participation: it is the completed art exercising its capacity on the reader. The artless share of chapter 11 is what everyone does without the art; what the ending does with Socrates' questioning is the art's own culminating exercise, and a form the completed art takes at its achievement is not a form the art lacks. That is a claim about the operation as the ending performs it, not a claim that an artful enactment settles, retrospectively, its named practitioner's possession of the art — which §6 leaves open and §9 does not close. The chapter never says the enacted claim of him. It does it — which, on a page about what texts do to readers who supply what the text withheld, is the stronger placement, and the only one the text itself performs.

Notes

[^11]: The development of §9 was reached in exchange with a second reader (Astra, 2026-09-07): "at the level of the treatise's enacted τέχνη, that construction answers the question it leaves open as an attribution. Socrates' questioning supplies the form of the examination the ending performs on its reader … The question leaves his possession of the τέχνη unsettled while placing his questioning inside its performance." The reader first placed the operation at the beginning the epilogue recounts; the correction — that the ending places it at the culmination of Aristotle's operation — is Lee Sharks's, and the reader's restatement is kept: "Socrates appears at the achievement of the undertaking, where the art exercises itself most fully."

[^1]: Ibn Zurʿa (from the Syriac of Athanasius) and the old version attributed to al-Nāʿimī, in the Paris Organon (ed. Badawī, Manṭiq Arisṭū III, Cairo 1952; texts as presented in the Digital Corpus for Graeco-Arabic Studies): فأمّا هذه الصناعة فليس إنّما كان بعضها موجوداً وبعضها غير موجود... لكن لم يكن منها شيء موجوداً البتّة (Ibn Zurʿa); بل لم يكن منها شيء موجوداً البتّة (al-Nāʿimī). Anonymi in Aristotelis Sophisticos Elenchos Paraphrasis, CAG XXIII.4 (ed. Hayduck, 1884): ταύτης δὲ τῆς ὅλης πραγματείας οὐ τὸ μὲν ἦν... W. A. Pickard-Cambridge, Oxford Translation (1928); E. S. Forster, Loeb 400 (1955). Poste's translation likewise keeps the denial inside the negated member (had not been partially, though imperfectly, elaborated by others); what his note then does — dispute the assertion or reinterpret it — is the question n. 2 leaves open. The full witness apparatus, with loci and the location of any qualification in each witness, is Appendix A below. Still open there: Ross's apparatus; Dod's text of Boethius (the medieval vulgate is recovered in Appendix A, C1, from its embeddings in Albert and Giles, pending collation); Dorion; Hasper; Avicenna and Averroes; and the further Latin commentators mapped by Ebbesen, of which Ockham and Nifo are located.

[^2]: E. Poste, Aristotle on Fallacies, or the Sophistici Elenchi (London 1866), 171 (n. 9 to ch. 34). Whether the note preserves the assertion and disputes its truth, or reinterprets the assertion to preserve it, its own wording does not settle; the direction of travel is explicit either way.

[^3]: The broader comparative survey of Aristotelian οὐ Χ ἀλλὰ Υ with a change of finite verb remains owed. The chapter-20 instance and the 184b1–2 control given in the text are evidence that the import is not conventional; they are not a demonstration that it is impossible, and the argument does not need one.

[^4]: The analogy is thus a provenance argument before it is an ontological one: possessing or distributing products does not amount to possessing or transmitting the art. Cf. Prior Analytics I.30, 46a17–27 on the relation of collected materials to the method that uses them; the comparison is developed elsewhere and not needed here.

[^5]: A medieval reader of the treatise classifies the practice the same way: Michael of Ephesus ([Alexander], In Sophisticos Elenchos, CAG II.3, ed. Wallies, 18,9–10 and 18,26–28) calls the peirastic μέρος τοῦ διαλεκτικοῦ and adds πολὺ δὲ καὶ τὸ τῆς πειραστικῆς εἶδος παρὰ Πλάτωνι ἔστιν εὑρεῖν ἔν τε τῷ Εὐθύφρονι καὶ ἐν ἅπασι σχεδὸν αὐτοῦ τοῖς διαλόγοις — support for the classification of the activity, and no more; the commentator says nothing of chapter 34's provenance claim.

[^6]: See n. 1. The one apparent exception, Poste's expanded rendering, keeps not governing the elaboration; though imperfectly characterizes the condition denied, not one affirmed.

[^7]: The registers, the constant referee prompt, the meaning of unprimed, and the verbatim reports are given in the accompanying reception study (The Particle, archive record #1574): twenty terminal reports across eleven configurations, generated by signed-out instances of two commercial systems on successive versions of a note asserting the sentence's positional facts; the sequence of configurations was responsive, so the sample measures the operation's robustness under successive objections, not a baseline; one further report, generated on a later configuration, held the negation throughout and is registered separately. The coding asks one thing of each report: did it preserve Socrates as a source-antecedent and qualify the denial to accommodate him? All twenty did.

[^9]: Some take ἡ πραγματεία here as the written treatise, even the Topics–SE whole. The construal changes nothing: the Γοργίου control and the ἔργον-individuation of §4 govern the book-reading identically, and the companion count at 184b1 is περὶ τοῦ συλλογίζεσθαι, wider than either.

[^8]: Many and One: Division, Unity, and the Boundaries of a Corpus in Plato and Aristotle (archive record #1576), whose question — which word moves? — this paper has put to one sentence, and whose §9.2 first read the epilogue's sequence as an order of grasp.

[^10]: Metaph. Μ 4, 1078b23–32, whose clauses run: Socrates busied with the ethical virtues, seeking to define universally; συλλογίζεσθαι γὰρ ἐζήτει, ἀρχὴ δὲ τῶν συλλογισμῶν τὸ τί ἐστιν; διαλεκτικὴ γὰρ ἰσχὺς οὔπω τότ᾽ ἦν, ὥστε δύνασθαι καὶ χωρὶς τοῦ τί ἐστι τἀναντία ἐπισκοπεῖν; δύο γάρ ἐστιν ἅ τις ἂν ἀποδοίη Σωκράτει δικαίως, τούς τ᾽ ἐπακτικοὺς λόγους καὶ τὸ ὁρίζεσθαι καθόλου· ταῦτα γάρ ἐστιν ἄμφω περὶ ἀρχὴν ἐπιστήμης; ἀλλ᾽ ὁ μὲν Σωκράτης τὰ καθόλου οὐ χωριστὰ ἐποίει οὐδὲ τοὺς ὁρισμούς. Metaph. Α 6, 987b1–4, with the sequel in which Plato, ἀποδεξάμενος, takes up the definitional concern — uptake by Plato, of a concern; not receipt by this undertaking, of its material. An. Pr. I.31, 46a31–33 (ἡ δὲ διὰ τῶν γενῶν διαίρεσις μικρόν τι μόριόν ἐστι τῆς εἰρημένης μεθόδου … οἷον ἀσθενὴς συλλογισμός) is cited for one thing only: Aristotle possesses part-of-method vocabulary and uses it where he means it — of division; at 183b34–36 what he writes is that vocabulary's denial. Cited from the standard texts; collation noted under Texts.

Texts

Greek text of SE 34, 183a37–184b8 and of chapters 2, 8, 11 as seated in #1574, appendix ("Texts"), from the Bekker-derived text of the Digital Corpus for Graeco-Arabic Studies; Bekker lineation as assigned there; collation against Ross (OCT 1958) outstanding at all loci, with the margin check listed as open; seated readings flagged for that collation: Προειλόμεθα μὲν οὖν (183a37), πῶς ἐρωτητέον ἢ τακτέον τὴν ἐρώτησιν πᾶσαν without ἐστίν (183b10–11), and the ch. 20 close at 177b33–34; Metaph. Α 6, 987b1–4 and Μ 4, 1078b23–32 and An. Pr. I.31, 46a31–33, quoted in §5 and n. 10, await the same collation. The presence of ἄλλο at 184b1 (παντελῶς οὐδὲν εἴχομεν πρότερον ἄλλο λέγειν in the seated text) was flagged because at least one circulating text omits it; collated 2026-09-07 against Bekker's editio princeps of the Academy text (Berlin 1831, vol. I p. 184, archive.org aristotelisopera01arisuoft), which prints περὶ δὲ τοῦ συλλογίζεσθαι παντελῶς οὐδὲν εἴχομεν πρότερον ἄλλο λέγειν, ἀλλ᾽ ἢ τριβῇ ζητοῦντες πολὺν χρόνον ἐπονοῦμεν, and 183b34–36 as quoted here, with 184a2 παρ᾽ αὐτῶν; no variant is recorded for 184b1–2 in the legible portion of Bekker's apparatus. The word stands in the first critical text. Ross's OCT (1958) remains unseated — in copyright, not on the open shelf — and is now the only outstanding witness at this line. Translations are the author's working versions unless attributed.

Appendix P. What the Text Prohibits

The nine prohibitions the last page supplies, in the order readers try to escape them; every locus from the seated text of Appendix A; each with what it rules out and how far it reaches. A reading of the sentence must satisfy all nine. This paper's reading does. A rival reading is invited to be checked against the same list, and to name the prohibition it breaks and the sophistry by which it breaks it. The list is Astra's audit of 2026-09-07, with the reach of each prohibition stated by the auditor, adopted whole; the order is the order of the escape routes as they were tried on an unprimed composer the same day (capture se34-we-walk-chatgpt-unprimed-20260907).

P1. The scope is this undertaking. Ταύτης δὲ τῆς πραγματείας (183b34) points back to the undertaking the retrospective has just described — the syllogistic capacity, questioning and answering, the treatment of arguments (183a37–b8).

Rules out: rescuing the denial by relocating it to an undertaking defined independently of that description — "logic," "the formal theory of valid inference," a calculus the Analytics later supplies.

Reach: it fixes the referent; it does not by itself decide whether every activity the undertaking examines counts as a contribution to it. That boundary is argued from the passage (P2, P8).

The escape it closes, first tried: "he means the systematic study of logic."

P2. Socrates is inside the activity. Τὴν δ᾽ αἰτίαν εἰρήκαμεν τούτου, ἐπεὶ καὶ διὰ τοῦτο Σωκράτης ἠρώτα ἀλλ᾽ οὐκ ἀπεκρίνετο· ὡμολόγει γὰρ οὐκ εἰδέναι (183b6–8). The reason just given for the dialectician's questioning is the reason for Socrates' questioning: he is placed by the account of the undertaking's own tasks, from inside it.

Rules out: "Socrates is irrelevant, or must be imported from Plato." Calling Socrates a practitioner does not remove him from the activity identified here, and does not by itself establish his position relative to the denied provision.

Reach: it prohibits exclusion from the activity; it does not by itself settle inclusion in the speaking we (see the joints, below), and it does not settle provenance — P1 and P6–P9 constrain what a reader may infer about that. The Byzantine paraphrast (Appendix A1a) is the earliest witness that excellence in the activity is no reason to weaken the denial: Socrates διαλεκτικὸς ἄκρος, and the total denial, in one breath.

The escape it meets, second tried: "a practitioner of dialectic, but not a systematic logician" — which does not fail here (a practitioner is in the activity) but at P7, where systematic is supplied as the condition for counting something within the denied provision.

P3. Received is received from others. τῶν γὰρ εὑρισκομένων ἁπάντων τὰ μὲν παρ᾽ ἑτέρων ληφθέντα πρότερον πεπονημένα κατὰ μέρος ἐπιδέδωκεν ὑπὸ τῶν παραλαβόντων ὕστερον (183b17–19). The reception class is defined: things taken over from others who laboured before, advanced part by part by those who received them.

Rules out: disguising transmission as shared affiliation — "same school, so not really received."

Reach: this clause defines the class; the denial (P6–P7) excludes the undertaking from it. It does not prohibit affiliation without transmission.

The escape it closes, third tried: "a lineage, an intellectual tradition."

P4. The beginning may be small. τὰ δ᾽ ἐξ ὑπαρχῆς εὑρισκόμενα μικρὰν τὸ πρῶτον ἐπίδοσιν λαμβάνειν εἴωθε … μέγιστον γὰρ ἴσως ἀρχὴ παντός (183b20–23).

Rules out: "the predecessor contributed only a little, so nothing existed" — where the little is a beginning of this undertaking; the text says a beginning can be small in extent and the greatest part in consequence.

Reach: it does not establish that any particular earlier activity was such a beginning; it forbids dismissing one for being small.

The escape it closes, fourth tried: "Socrates had a primitive version that Aristotle systematized."

P5. Succession is named, and it is not this. παραλαβόντες παρὰ πολλῶν οἷον ἐκ διαδοχῆς, then Tisias after the first, Thrasymachus after Tisias, Theodorus after him (183b26–33).

Rules out: repairing the sentence by enlarging we into a succession — "Socrates and Aristotle in one tradition, so their transmission no longer counts."

Reach: chronological succession is not the prohibited relation; succession through transmitted contributions to this undertaking is. A merely chronological lineage remains possible; it cannot explain inherited material.

The escape it closes, fifth tried: "the we is a historical lineage reaching back to Socrates."

P6. Not part and part. οὐ τὸ μὲν ἦν τὸ δ᾽ οὐκ ἦν προεξειργασμένον (183b34–35).

Rules out: "Socrates supplied this part; Aristotle the rest," as antecedent provision and subsequent completion of this undertaking.

Reach: nearly absolute at the level of the prohibition; the only dispute can be whether an alleged contribution falls inside the scope (P1), never whether partial provision is permitted.

The escape it closes, sixth tried: "some of it had been worked out."

P7. Nothing means nothing. ἀλλ᾽ οὐδὲν παντελῶς ὑπῆρχεν (183b35–36).

Rules out: replacing zero with an adequacy threshold — "nothing sufficiently developed, comprehensive, systematic, formal, comparable."

Reach: the absolute is relative to the stated undertaking (P1). It does not independently establish "no one previously performed any syllogizing"; scope must be established (P1–P2), and adequacy cannot be inserted.

The escape it closes, tried at every stage: the conserved adjective.

P8. Shoes are not shoemaking. οὐ γὰρ τέχνην ἀλλὰ τὰ ἀπὸ τῆς τέχνης … βεβοήθηκε μὲν πρὸς τὴν χρείαν, τέχνην δ᾽ οὐ παρέδωκεν (184a2–8).

Rules out: counting the provision of ready-made arguments as transmission of the art.

Reach: a specific exclusion of one mode of provision — the paid teachers'. It does not say no art existed anywhere, and it does not assign Socrates to those teachers; using it to place him needs P2.

The escape it closes, seventh tried: "the sophists had the practice; Aristotle supplied the theory."

P9. Not enlarged by transmission. περὶ δὲ τοῦ συλλογίζεσθαι παντελῶς οὐδὲν εἴχομεν πρότερον ἄλλο λέγειν, ἀλλ᾽ ἢ τριβῇ ζητοῦντες πολὺν χρόνον ἐπονοῦμεν (184b1–2); then the comparison class, τὰς ἐκ παραδόσεως ηὐξημένας (184b3–5).

Rules out: restoring inherited provision at the end — treating this undertaking's growth as the enlargement-by-transmission just ascribed to the others.

Reach: note the grammar the composers drop. The subject is first-person, εἴχομεν; the adverb is πρότερον; the verb is λέγειν; there is ἄλλο, nothing else to say; and there is no πρὸ ἡμῶν. What we had earlier to say about the activity and whether anyone earlier performed the activity are different propositions. The reading rests on the former, and the we does its work at this word.

The escape it closes, last tried: "it was received after all, just informally."

The two joints, stated as joints. After the nine, two steps in this paper's reading remain inferences rather than prohibitions, and are marked as such. (i) From P2 to the speaking we: the connector shares a reason; Socrates stands in the third person while the we of προειλόμεθα and εἰρήκαμεν is the exposition's. The paper does not need pronoun membership; it needs P2 and P9 together — Socrates inside the activity, and we having had nothing to say before — which is what §5 argues. (ii) From P3's παρ᾽ ἑτέρων to not other: "not other relative to the undertaking's provenance" and "not a distinct maker" are different propositions. The paper claims the first — Socrates is not filed as one from whom this was received — and withholds the second, which this sentence cannot decide and which is argued elsewhere on other evidence.

The standing of the reading. It satisfies P1–P9. The transmitting-antecedent reconciliations examined here — Socrates as the predecessor from whom part of this was received, in any of their forms — fail these prohibitions: they restore the excluded transmission (P3, P5, P9), insert the adequacy threshold (P7), or file participation as provision (P6). The non-transmitting alternatives — the significant precursor who hands nothing on (§4), exemplary relevance without inclusion (§5) — preserve the denial and are not failed by the list; they are compared with inclusion in §5, where the paper argues that inclusion explains more of the page than they do, not that they break a prohibition. The invitation is open at both levels: supply a transmitting reading that meets all nine, or a non-transmitting one that explains the four whys of §5 better.

Appendix L. The Literature, by Prohibition

The paper is long; the literature would make it longer. The rule used here keeps it short and honest: (1) a modern reading is reported only from a seated text — no construal is attributed to a scholar whose page has not been read from a source on the archive's shelf; (2) what is reported is keyed to Appendix P — which prohibition, if any, the reading relaxes, in one line — so the review is an index against the paper's own constraints and not a survey; (3) unseated works are named as debts, with what each would decide, and not summarized from memory. The body cites this appendix by number.

L1. Seated: the translations (Appendix D). The English, Latin, Arabic, and Byzantine witnesses are the paper's primary literature, and Appendix D–E already reads each at the sentence: where the gloss enters (systematic, unsystematic, complete), which word it moves, and whether the denial survives it. The datum from that reading stands as the literature's first fact: the qualification is modern — Forster's unsystematic and its English successors — and no premodern witness supplies it (E.7–E.8). Against P7 these are the relaxations; against P2, the paraphrast (A1a) is the one premodern reading that places Socrates inside dialectic and keeps the denial total, which is the paper's own position in outline.

L2. Named as debts, with what each would decide. Ross, Aristotle's Prior and Posterior Analytics (1949) and the OCT Topica et Sophistici Elenchi (1958) — the apparatus at 183b35 and 184b1 (§ F); Dorion, Les Réfutations sophistiques (Paris–Québec 1995) — the modern commentary most likely to state the qualification and its warrant, and therefore the one whose page would show which prohibition it relaxes and why; Hasper's text and translation (2013) — the recent English rendering of the last chapter, to be read at P7 and P9; Fait, Le confutazioni sofistiche (2007); Schreiber, Aristotle on False Reasoning (2003) — for the peirastic and the ch. 11 distinction that §4 and §9 depend on; Dorion's work on Aristotle and Socratic dialectic, to be read at P2; and the Ockham and Nifo expositions located in § F. Each is a seating, not a citation, until its page is on the shelf; the paper does not report what it has not read.

L3. What the review would look like when discharged. One row per work: the sentence as rendered or construed; the prohibition relaxed, if any (P1–P9), or none; the escape route of Appendix P it corresponds to, if any. Nothing else. A reading that relaxes no prohibition is a rival on the paper's terms and is answered in §5; a reading that relaxes one is answered by the table. That is the whole of the literature the argument needs, and it fits on a page.

Appendix A. The Witnesses

The apparatus §7 of the note declared open: identifiable renderings of SE 34, 183b34–36 and 184b1–2, with the location of any qualification. Read from the sources named; loci as printed; nothing paraphrased where a quotation could be had. Each entry records (a) the rendering of 183b34–36, (b) of 184b1–2 where reached, (c) where in the witness the qualification enters, if it does. Two seams are tracked throughout because the note's argument turns on them: what becomes of παντελῶς, and what becomes of ἄτεχνος at 184a2 (the teachers' instruction, ταχεῖα μὲν ἄτεχνος δ᾽), since that is where the word systematic enters the English tradition.

A. Greek

A1. Anonymous paraphrase of the Sophistici Elenchi (Byzantine; CAG XXIII.4, ed. M. Hayduck, Berlin 1884; the passage at p. 67 Hayduck: the volume's index locorum, read from the print scan (archive.org p4anonymiinaristo23hayd), places Τισίας ηὔξηκε τοὺς ῥητορικοὺς λόγους at 67,23 sq., with Thrasymachus and the Gorgias comparison following; an earlier reading of the same index as 61,23 was an OCR error, corrected 2026-09-07). The paraphrast keeps the sentence almost verbatim and adds one word:

ταύτης δὲ τῆς ὅλης πραγματείας οὐ τὸ μὲν ἦν τὸ δ᾽ οὐκ ἦν προεξειργασμένον, ἀλλ᾽ οὐδὲν παντελῶς ὑπῆρχεν.

Qualification: none. The addition is ὅλης — of this whole treatment — which reinforces the totality and names the domain as the treatment entire. The paraphrase then runs the shoe passage and 184b1–2 essentially as Aristotle's text (παντελῶς οὐδὲν εἴχομεν πρότερον ἄλλο λέγειν ἀλλ᾽ ἢ τριβῇ …).

A1a. The same paraphrast at 183b1–8 (pp. 66,29–67,1 Hayduck, as the volume's own index places Σωκράτης διαλεκτικὸς ἄκρος ὢν ἠρώτα ἀλλ᾽ οὐκ ἀπεκρίνατο; collated 2026-09-07 against the print scan, archive.org p4anonymiinaristo23hayd, where the sentence stands at the foot of p. 66 under the running head [Arist. 34 p. 183b1] and the apparatus on that page records ἄκρως M at the key word — the adverb in one manuscript where Hayduck prints the adjective; the earlier reference to pp. 196–197 was a misreading of the scan and is withdrawn). His version of the second task keeps the role-participle (… ἀλλὰ καὶ ὡς εἰδώς … ὅπως λόγον ὑπέχοντες καὶ ἐρωτώμενοι φυλάξομεν τὴν θέσιν …), and his Socrates sentence supplies its mirror and an epithet Aristotle does not:

τὴν δὲ αἰτίαν εἰρήκαμεν τούτων· ὁ γὰρ διαλεκτικὸς ὥσπερ μὴ εἰδὼς ἐρωτᾷ· διὰ τοῦτο Σωκράτης διαλεκτικὸς ἄκρος ὢν ἠρώτα, ἀλλ᾽ οὐκ ἀπεκρίνατο· ὡμολόγει γὰρ οὐκ εἰδέναι. ὁ μέντοι σοφιστικὸς πάντα εἰδέναι ἐπαγγέλλεται …

We have stated the cause of these things: for the dialectician asks as one not knowing; therefore Socrates, being a consummate dialectician, asked but did not answer, for he agreed he did not know. The sophist, by contrast, professes to know everything … (The volume's index confirms the thread: ὁ διαλεκτικός opp. the sophist's πάντα εἰδέναι at 195,30; sophistic answering at 198,1.)

Classification of A1 whole: the paraphrast is the strongest premodern occupant of the fork's second horn. He promotes Socrates inside dialectic — not one of the ἅπαντες but its ἄκρος — while keeping the ὡς-grammar exact (the dialectician asks ὥσπερ μὴ εἰδώς: a role, mirroring the answerer's ὡς εἰδώς), and two sentences later he renders the denial total and strengthens it with his own ὅλης. Peak of the activity and zero of the possession, in one paraphrase, without contact between the lemmata and without strain: supreme practice does not register with him as partial prior possession of the πραγματεία, so there is nothing for a qualification to do. (Read from the archive.org OCR of Hayduck; pages and lines to be confirmed against the print.)

A2. [Alexander] = Michael of Ephesus, In Sophisticos Elenchos (CAG II.3, ed. M. Wallies, Berlin 1898), on SE 2, 165a38–b11, p. 18,9–28 Wallies:

τρίτον φησὶν εἶναι τὸν πειραστικὸν συλλογισμόν, ὅς ἐστι καὶ αὐτὸς διαλεκτικός· μέρος γὰρ τοῦ διαλεκτικοῦ … πολὺ δὲ καὶ τὸ τῆς πειραστικῆς εἶδος παρὰ Πλάτωνι ἔστιν εὑρεῖν ἔν τε τῷ Εὐθύφρονι καὶ ἐν ἅπασι σχεδὸν αὐτοῦ τοῖς διαλόγοις.

The third kind, he says, is the peirastic syllogism, which is itself also dialectical, for it is a part of the dialectical … and much of the peirastic form is to be found in Plato, in the Euthyphro and in almost all his dialogues. — A medieval (twelfth-century Byzantine) commentator, inside the treatise's own tradition, classing the Platonic dialogue as the peirastic form, and the peirastic as part of dialectic. This supports the identification §3 of the note makes from SE 2, 8, 11 and 34 — Platonic peirastic activity within dialectic — and establishes nothing about inheritance, and nothing about how Michael read chapter 34. Whether the commentary reaches 183b34–36 is to be confirmed in the print (the OCR carries running heads to 183a only).

B. Arabic (ninth–tenth century; ed. ʿA. Badawī, Manṭiq Arisṭū III, Cairo 1952; text as presented in the Graeco-Arabic Studies corpus)

The Paris manuscript of the Organon carries three Arabic versions of the Sophistici Elenchi: that of Yaḥyā ibn ʿAdī, that of Abū ʿAlī ʿĪsā ibn Isḥāq ibn Zurʿa (both from the Syriac of Athanasius), and an old version attributed to al-Nāʿimī. Two distinct renderings of chapter 34 are readable in the corpus.

B1. Ibn Zurʿa (colophon: نقل عيسى بن إسحق بن زرعة من السرياني بنقل أثانس):

فأمّا هذه الصناعة فليس إنّما كان بعضها موجوداً وبعضها غير موجود، وإنّما أضيف إليها الآن، لكن لم يكن منها شيء موجوداً البتّة.

As for this art (ṣināʿa), it is not that part of it existed and part did not exist, and it has only now been added to; rather, nothing of it was in existence at all.

(a) παντελῶς → البتّة (al-battata, "at all, absolutely"), a total negator on the whole clause. (b) The translator glosses the denied member: and it has only now been added to — reading the μέν/δέ condition as prior partial existence plus later addition, i.e. the received-development pattern the paragraph has just described for rhetoric. The qualification is thus placed inside the denied member, not the asserted one. (c) 184b1–2: ولأنّ أشياء خطبيّة كثيرة قد كانت موجودة في سالف الدهر؛ فأمّا في عمل القياس فلم يكن عندنا قديماً فيه شيء؛ إلّا أنّا بعد أن كددنا في الطلب زماناً طويلاً — many rhetorical things existed in past time; but in the work of syllogism there was not with us anything earlier in it; except that after we toiled long in the search … — παντελῶς not rendered here. (d) τὰ ἀπὸ τῆς τέχνης → أشياء صناعيّة (artful things, things belonging to the art); τέχνην δ᾽ οὐ παρέδωκεν → لم يفد صناعة; ἄτεχνος → لم يكن … يستفيد صناعة (did not gain an art). (e) The shoes: أجناساً كثيرة للخفاف مختلطة غير مفصّلة — many kinds of shoes, mixed, undivided.

B2. The old version attributed to al-Nāʿimī (colophon: نقل الناعمي; the compiler notes he does not know from which language it was made):

فأمّا صناعتنا هذه فلم يكن منها شيء موجوداً مستعملاً، ومنها شيء موجوداً لا مستعمل، بل لم يكن منها شيء موجوداً البتّة.

As for this art of ours, it is not that some of it existed in use and some existed not in use; rather, nothing of it was in existence at all.

(a) παντελῶς → البتّة again. (b) The μέν/δέ member is read as in use / not in use — a different construal of προεξειργασμένον, but again placed inside the denied member. (c) 184b1–2: وقد قيل قديماً في كلام الخطب كثيراً. فأمّا في السلجسموس وتأليف المقدّمات فلم يكن لنا قديماً شيء، بل قد أقمنا زمناً مجتهدين في طلب ذلك — much was said of old in the speech of orators; but in the syllogism and the composing of premises we had nothing of old; rather we spent a long time striving in the search for it — the adverb again not rendered. (d) Socrates: وأنّ سقراطيس لذلك كان يسأل الجميع ولم يكن يجيب، وذلك أنّه كان مقرّاً بأنّه لا يعلم — and that Socrates for that reason used to question everyone and would not answer, because he confessed that he did not know. (e) The shoes: قوالب الحذاء وكثرة أنواعها — shoe-lasts and the multitude of their kinds.

B3. Yaḥyā ibn ʿAdī. The corpus presents the same text under his heading as under Ibn Zurʿa's for this chapter; whether chapter 34 survives in a distinct Yaḥyā version is to be checked in Badawī.

Finding for the Arabic: both readable versions keep the quantifier total (البتّة) at 183b35 and put whatever construal they make of προεξειργασμένον inside the denied member; neither adds systematic, complete, or theory to the asserted clause. Both drop παντελῶς at 184b1. ἄτεχνος becomes no art was gained/handed on, not unsystematic.

Not yet seated: Avicenna, al-Shifāʾ, al-Safsaṭa (ed. A. F. al-Ahwānī, Cairo 1958); Averroes, Talkhīṣ al-Safsaṭa (ed. Jehamy). To be quoted from print.

C. Latin

C1. Boethius, translatio (Aristoteles Latinus VI.1–3, ed. B. G. Dod, Bruges–Paris 1975, pp. 5–60), with James of Venice's fragments and Moerbeke's recension in the same volume. The edition itself remains to be quoted; but the medieval vulgate's shape at the loci is now recoverable from two independent embeddings — Albert's paraphrase-lemmata (C3) and Giles's littera (C4):

• 183b6–8: propter hoc et Socrates interrogabat, sed non respondebat: fatebatur enim nescire (C4, littera).
• 183b16–17: Oportet autem nos non latere quod accidit circa… (C3 and C4 both open the chapter with it; μὴ λεληθέναι kept).
• 183b34–36: lemma incipit Huius autem negocii non haec quidem… (C4, divisio catchword); the denial's adverb as omnino — Albert's close paraphrase sed nihil omnino erat inventum a prioribus ante nos (C3) — with the μέν/δέ rendered as two denied parts (non haec quidem erat, illa autem non, per C3's double gloss pars quaedam / alia pars).
• 184a8: artem autem non tradidit (C3, verbatim in the paraphrase).
• 184b1–2: Et de rhetoricis quidem [lemma incipit, C4]; erant quidem multa et antiqua dicta … nihil habuimus prius (C3).
• 184b7–8: …indultionem; inventis autem multas habere grates (C4, littera; C3 indultionem sive indulgentiam… multas habere sive agere nobis grates).

All pending collation against Dod for exact wording and recension; the embeddings agree with each other and keep the quantifier total. (Ebbesen, Commentators and Commentaries on Aristotle's Sophistici Elenchi, Leiden 1981, remains the map for the rest of the Latin tradition.)

C2. Thomas Aquinas — the shaped silence (s. XIII)

Aquinas wrote no commentary on the SE; his Organon commentaries are the Peryermeneias (unfinished) and the Posteriora. In them the treatise is filed, not read at its ending:

Tertio autem processui rationis deservit pars logicae, quae dicitur sophistica, de qua agit Aristoteles in libro elenchorum. (Expositio Posteriorum I, lect. 1 n. 6 [Marietti 79471])

Further citations are of the same kind — the book as fallacy-manual: In Post. An. I lect. 13 n. 7 (sophists "sicut in libro elenchorum"); Expositio Peryermeneias I lect. 9 ("de quibus plenius facit mentionem in I elenchorum"). The De fallaciis transmitted under his name (authenticity doubted) is a taxonomy built from disputatio down through the thirteen fallacies; its eighteen elench- references are classificatory; the epilogue appears nowhere.

The positive datum is where Aquinas does comment, at length, on the received relation: Sententia Metaphysicae II lect. 1 nn. 15–16, on Metaph. α 1, 993b11–19 — it is just to thank even superficial predecessors; the Timotheus/Phrynis chain (a named succession in music, exactly parallel in form to SE 34's Tisias-chain in rhetoric); and his own expansion:

A quibusdam enim praedecessorum nostrorum accepimus aliquas opiniones de veritate rerum... Et iterum illi, a quibus nos accepimus, invenerunt aliquos praedecessores, a quibus acceperunt, quique fuerunt eis causa instructionis. (n. 16 [81854])

accepimus a praedecessoribus — the παρ᾽ ἑτέρων relation affirmed for philosophy generally, glossed approvingly, with a διαδοχή as its example — and no cross-reference to the Elenchi boast, here or anywhere in the authentic corpus so far examined. Aquinas had the cheapest possible occasion to stage the reconciliation Poste calls "difficult": commenting the gratitude text with Boethius's total rendering of 183b34–36 on his shelf. No sed contra dicitur in fine Elenchorum occurs.

Classification: an absence-witness, and the absence has a shape: the century's chief glossator transmits the counter-text (reception affirmed) at length and never brings it into contact with the denial. No reconciliation is staged; whether tension was felt, the silence cannot say. This bounds the qualification's history from a second side: not in the translators (A, B, D), not staged in Aquinas (C2); its home so far is the modern gloss (D-notes, E).

Still to seat for the Latin line: Boethius's exact wording at 183b34–36 and 184b1–2 from Dod (AL VI.1, pp. 5–60) — C1 above remains a placeholder until quoted; then the two Latin commentators who actually reach ch. 34: Albert the Great (Liber Elenchorum, paraphrase) and Giles of Rome (Expositio super libros Elenchorum, ed. Venice 1500 and after), with Ebbesen's Commentators and Commentaries on Aristotle's Sophistici Elenchi as the map.

C3. Albert the Great, Liber II Elenchorum, tract. V, cap. 3, "Qualiter scientia logices est inventa" (Borgnet, Opera omnia II, Paris 1890, pp. 711–713; archive.org operaomniaexedit02albe)

The one full scholastic paraphrase of the chapter so far seated, and it performs the provenance reading move for move.

• The transition kept: Oportet autem nos non latere quod accidit circa … totum logicum negotium, quantum ad modum inveniendi ipsum — the chapter titled by the finding-verb (inventa), the μὴ λεληθέναι rendered.
• The two provenances expanded: quae quidem sumpta sunt ab aliis, a quibus prius elaborata sunt, a posteris particulariter augentur … prima inventa ab eis successoribus priorum — received-and-grown against found-from-the-beginning; the Tisias chain kept with Albert's own simile, quorum dicta collegit quasi avis quaedam grana — Tisias gathering the elders' sayings like a bird gathering grains.
• The denial total, both members: Sic autem a nobis non est inventa haec scientia sophistica et universaliter logica: ante nos enim huius negotium non hoc (hoc est, pars quaedam) erat: illud autem (hoc est, alia pars) non erat perscrutatum … sed nihil omnino erat inventum a prioribus ante nos de hoc negotio. The μέν/δέ glossed as two parts and both denied; omnino seals it; a prioribus ante nos names the axis.
• Plato staged — and dissolved by the shoe: Albert (misreading τῇ Γοργίου πραγματείᾳ) takes the teachers' model to be liber Platonis, qui dicitur Gorgias, eo quod ad discipulum suum Gorgiam est scriptus — and then runs Aristotle's own firewall against it: quaedam doctrina fuit… tamen erat similis mercenarii… non enim artem dantes sive tradentes, sed potius quaedam exemplaria… artem autem secundum principia artis non tradidit. The one Platonic candidate the Latin chapter ever names at the boast is conceded as quaedam doctrina and dismissed with Aristotle's transmission-verb negated.
• The qualification, where it occurs, is on the transmission axis: de syllogismis autem per artem traditam nihil habuimus prius ante nos … de logica quidem tali nihil traditum habuimus prius ab Antiquis, nisi exempla quaedam particularia — "nothing handed down by way of art," "except certain particular examples." Not degree; provenance. The exception clause absorbs the teachers' handouts, exactly as the Greek's ἀλλ᾽ ἤ absorbs the τριβή.
• Socrates the humble non-knower — and Plato beside him: interrogatio sit signum nescientis. Propter hoc Socrates semper interrogabat, sed non respondebat: quia ex humilitate in ipsa interrogatione fatebatur se non scire: cum (ut dicit Aristoteles in libro de proprietatibus elementorum) Plato esset omni sapientia perfectus et laudatus, tamen si quid docuit, docuit quasi interrogando. Albert widens the questioner's place to admit Plato — teaching quasi interrogando — rather than quarantining Socrates outside the inquiry. And just above: responsio opus et actus scientis — answering as the knower's act, the ὡς εἰδώς key in the paraphrast's own words.
• The envoi's gloss says the boast aloud: de omissis quidem artis a nobis, qui adiutorium nullum habuimus, tradere indultionem… inventis autem… multas habere sive agere nobis grates — "we who had no help."

Classification: faithful rendering at 183b34–36 (total, both members); reinterpretation nowhere on the degree axis; the narrowing phrase (per artem traditam, nisi exempla) is at 184b1–2 and attaches to the transmission verbs; what conception of ars rides inside it — secundum principia artis, universal precepts ordered to one end — is Albert's, and is left open here. No Socrates-reconciliation is staged.

C4. Giles of Rome, Expositio supra libros Elenchorum (s. XIII ex.; ed. Venice, Torresani, 24 Sept. 1500; archive.org A3360641, OCR partially legible; leaf-level check from images pending)

• Answering keyed to knowers: at 183b1–8 the expositio gives the reason the treatise took on responding: difficile est non dicere aliquid repugnans, et soli scientes possunt observare ne respondendo sibi contradicant — only knowers can answer without self-contradiction. Then the littera: propter hoc et Socrates interrogabat, sed non respondebat: fatebatur enim nescire, glossed volebat interrogare … nolebat respondere, the imperfects taken as volitional.
• The boast construed on the transmission axis, twice: in the divisio — ut Philosophus ostenderet se fuisse huius artis specialem inventorem, dabat modum quo haec ars inventa erat … ceterae artes inventae sunt per auxilium plurium … sed haec scientia sive haec ars non sic inventa est: quia ante tempus Aristotelis nihil de ea erat traditum per modum artis; propter quod Philosophus innuit se fuisse specialem inventorem ipsius — and in the closing laude-argument: cum tempore Philosophi nihil per modum artis inventum esset de logica, Philosophus eam complete tradens, non adiutus prioris auxilio, dignus est admiratione et laude. "Nothing handed down/found per modum artis"; "with no predecessor's help"; specialis inventor. The qualifying phrase exists — and it qualifies transmission, not degree of elaboration.
• The envoi's littera: …indultionem; inventis autem multas habere grates — as the special inventor, petit veniam de omissis et gratiam de inventis.

Classification: rendering kept; construal attached to transmission vocabulary (traditum per modum artis, non adiutus prioris auxilio), the chapter read as an argument that Aristotle earns the admiratio because nothing was received; the conception of ars riding in per modum artis left open. No reconciliation with Plato or Socrates attempted; Socrates remains the non-answering non-knower of the littera.

D. English

D1. E. Poste, Aristotle on Fallacies, or the Sophistici Elenchi (London 1866), translation and note 9 to ch. 34. The translation:

But the system I have expounded had not been partially, though imperfectly, elaborated by others: its very foundations had to be laid.

and at 184b1–2:

dialectic had absolutely no traditional doctrines. Our researches were tentative, long, and troublesome. If, then, starting from nothing …

Note 9:

It is difficult to reconcile Aristotle's assertion with what we know had been done by Plato and Socrates and the Eleatics and Megarians. What he really performed in his dialectical treatise was to indicate a number of methodic principles or elements of method (τὰ κοινά); and it is probable that none of his predecessors had separated and extricated these from the specific propositions (τὰ ἴδια) …

Classification, exactly. In the translation, not still governs the elaboration: had not been partially, though imperfectly, elaborated denies the condition; though imperfectly characterises the condition being denied, not one being affirmed. The wording is added, but the sentence does not thereby affirm a partial inheritance. Its very foundations had to be laid does not obviously weaken the assertion and can be read as intensifying it — a beginning without inherited foundations. So the translation is an expanded rendering that keeps the denial where the Greek keeps it; it is not, on these quotations, a qualification of the asserted clause. ἄτεχνος → unscientific; 184b1 → no traditional doctrines, a domain word inserted at the second sentence.

Note 9 is the stronger witness, and it needs the right label. Poste begins from the difficulty of reconciling Aristotle's assertion with what we know and proceeds to what he really performed. Two readings of that movement are open: that he preserves the assertion as strong and disputes its truth (a historical correction), or that he presents the corrective account as what the sentence means (a reinterpretation offered as reconciliation). His phrase Aristotle's assertion and his acknowledgement of a reconciliation problem show that he recognises the strong assertion; whether he then substitutes the narrower proposition for its meaning is not settled by the note's wording. The note is therefore recorded as documenting reconciliation from an external historical account, with the semantic substitution left as the question the witness raises and does not decide. It is the earliest English witness examined here.

D2. W. A. Pickard-Cambridge, Oxford Translation (1928; Revised Oxford Translation 1984):

Of this inquiry, on the other hand, it was not the case that part of the work had been thoroughly done before, while part had not. Nothing existed at all.

184b1–2:

on the subject of reasoning we had nothing else of an earlier date to speak of at all, but were kept at work for a long time in experimental researches.

Qualification: none in the asserted clause; Nothing existed at all stands as its own sentence. One adverb enters the denied member — thoroughly done before for προεξειργασμένον — which builds a degree into the thing denied. ἄτεχνος → rough (ready but rough). of an earlier date to speak of keeps ἄλλο λέγειν.

D3. E. S. Forster, Loeb Classical Library 400 (Cambridge MA 1955). The sentence is rendered faithfully — the μέν/δέ denied, then "nay, it did not exist at all." 184b1–2 gives "absolutely no earlier work to quote." Forster's marginal analysis heads the paragraph Originality of the present treatise. Qualification: none in the asserted clause. Two entries elsewhere: ἄτεχνος is rendered unsystematic (rapid but unsystematic) — the lexical location of systematic in the English text of the chapter, attached to the teachers' instruction; and earlier work to quote at 184b1 narrows ἄλλο λέγειν toward antecedent literature. Whether the word then migrates from the instruction to a gloss on nothing is a hypothesis about transmission, not a documented step; a later quotation performing the transfer would establish it, and none is seated yet.

Not yet seated: P. S. Hasper's translation (Logical Analysis and History of Philosophy 15, 2012); L.-A. Dorion, Les Réfutations sophistiques (Paris–Québec 1995), text and notes; M. Zanatta (Italian); E. Rolfes (German, 1922); Ross, OCT 1958, apparatus at 183b35 and 184b1.

E. What the witnesses show, so far

Three operations are distinguished throughout, because the paper's argument turns on distinguishing them: faithful rendering; disagreement with the assertion; reinterpretation of the assertion. The apparatus must not combine them while criticising readers for combining Aristotle's.

1. The translations examined keep the quantifier total. A1 (with ὅλης added), B1 and B2 (البتّة), D2 (Nothing existed at all), D3 (it did not exist at all), and, on the corrected reading, D1 as well (not governing the elaboration; the denial kept). None qualifies the asserted clause with systematic, complete, or theory. This supports the note's treatment of the negation and establishes, for these witnesses, that the qualification is not in the rendering of 183b35–36.

2. Explanation inside the denied member is not qualification of the asserted one. B1 (and it has only now been added to), B2 (in use / not in use), D1 (though imperfectly), D2 (thoroughly done) all add words to the condition Aristotle denies. That is a different operation from narrowing what is asserted, and the apparatus records it as such.

3. One commentary confronts the assertion with an external historical account. Poste's note 9 recognises the assertion as strong (difficult to reconcile) and supplies what he really performed. Whether that is disagreement with the assertion or reinterpretation of it is the exact question this witness raises; it is not decided here.

4. The lexical location of systematic in the English chapter is ἄτεχνος at 184a2 (Forster unsystematic; Poste unscientific), in the sentence about the teachers. Its transfer to a gloss on nothing is a hypothesis awaiting a quotation that performs it.

5. A medieval commentator on the treatise's own tradition classes Platonic dialogue as the peirastic and the peirastic as part of dialectic (A2). This supports §3's identification and nothing beyond it.

What follows for the note: the claim in §0 and §3 that the translations, for the most part, do not construe is now supported for the witnesses examined; the further claim that the tradition then supplies the qualification in the gloss is supported by one commentary note whose exact operation is open, and is otherwise not yet established. The note should say the first and hold the second as the hypothesis it is, with Poste as its one seated instance.

7. The Latin thirteenth century narrows, where it narrows, in transmission vocabulary. The scholastic phrases are per artem traditam / traditum per modum artis / non adiutus prioris auxilio (C3, C4): the qualifier rides the transmission verbs. Per modum artis also carries a conception of what an art is (Albert: secundum principia artis), and that conceptual load is left open here — so these witnesses establish coexistence, not confirmation: emphatic denial and explicit attention to the questioners stand together in them without repair. The reconciliation operation Poste announces in 1866 (D1-note) is not staged in C2–C4: Aquinas never brings the gratitude-text and the boast into contact; Albert stages the one Platonic candidate and dissolves it with artem non tradidit; Giles builds the chapter's honour on the boast being total. Among the witnesses examined, the qualification's degree-form is not attested before the modern period.

8. The fork's second horn has a premodern occupant, and the denial survives him intact. The Byzantine paraphrast (A1a) calls Socrates διαλεκτικὸς ἄκρος — supreme in dialectic, more than the universal share — in the same paraphrase whose denial-sentence he strengthens with ὅλης. The premise the modern qualification requires — that excellence in the activity constitutes partial prior possession of the undertaking — is visibly absent: where a modern reader must move a word, the paraphrast promotes the man and totalizes the sentence in one breath. His role-grammar states why: the dialectician asks ὥσπερ μὴ εἰδώς — ἄκρος modifies a role, and a role is not inventory.

F. Open

Ross's apparatus at 183b35 and 184b1 (Bekker 1831 collated at both, 2026-09-07 — the seated readings confirmed; Ross unseatable on the open shelf); the paraphrast now collated against Hayduck's print at A1 and A1a (page and line references corrected; ἄκρως M recorded from the apparatus); Dod's text for C1 (the embeddings above pending collation); Hasper; Dorion's notes (the modern commentary most likely to state the qualification and its warrant); Avicenna and Averroes; Yaḥyā's version; the page/line references for A1 and A2 against the print; Giles at leaf level from the images (the OCR is legible only in patches); and, from the same hunt, two further Latin witnesses located but unread: Ockham, Expositio super libros Elenchorum (archive.org venerabilisincep0003will, lending-restricted) and Agostino Nifo, Expositiones in libros de Sophisticis elenchis (archive.org ARes034053) — the latter the natural bridge witness between the scholastic construal and the modern gloss. Also to check: Pickard-Cambridge's rendering of ἄτεχνος at 184a1–2 (reported elsewhere as ready but rough; if confirmed, the English entry of unsystematic narrows to Forster alone, and D3's position-datum sharpens).