# The sophistry measure on the 3I/ATLAS record: first run, 2026-11-06

At the author's word (07:01 EST: "lets make use of the sophistry measure on the overview examples, take it for a spin"; "perhaps the sophistry measure comes into play on the ata paper and some of the reporting, as well"). The instrument is EA-SOPHISTRY-01 (#1727, AXN:0BDC). Each finding carries the record of its §3.2: scope; the adequacy condition of §0.3 (what was asked, what scope the text announces, which omitted relation changes the answer at that scope); witnesses by id (spans, offsets and source hashes in witnesses.json, checked by witnesses.py); reconstruction; type, with its relation to the historical class (direct instance, structural extension, interpretive connection, #1727 §2); effect types (§2.5); standing granted (§2.7); sharpness (§2.9); the strongest alternative reading; countertest; smallest repair. A finding records and blocks nothing (§3.6). No motive is coded (§2.4). The sophistic class proper needs three established components (the appearance offered, the performance it stands in for, the commercial relation) and none of the findings below establishes the third; they record operations.

Texts read: three Google AI Overviews seated in the Capture Registry (avi-loeb-3i-atlas-aio-20261102; 3i-atlas-signal-studies-aio-20261104; 3i-atlas-ata-drift-rate-filter-aio-20261104), Loeb's essay of 2026-04-23 as saved at intake, the ATA paper (Sheikh et al., arXiv:2512.18142v2) and the Intelligent Living account (2026-09-11, updated 2026-09-30) as quoted in the notebook draft (§2.2, §5.1). The paper and the article are quoted from the draft; neither is yet seated in the archive (requirement R1 below).

## S1. The scientist's position given as one interview line (Overview, "avi loeb 3i atlas")

- Scope: composition. Adequacy condition: the query names a person and an object; the body announces "Key Developments and Loeb's Stance". The omitted relation: Loeb's own later writing, which keeps the anomalies ranked (W2) beside the concession (W3), and which the composition holds among its cards (W4).
- Witnesses: W1 (the body's last word on his stance), W2, W3 (essay of 2026-04-23), W4 (the essay of 2026-05-25, carded).
- Reconstruction: the December remark is quoted accurately; the composition grants it the standing of the position. Selection of a locus (ἔκθεσις) is no fault; the finding lies in the standing conceded (#1727 §2.7).
- Type: X24, one locus of an entity's development substituted for the developed entity. Relation: structural extension of the this-and-such (SE ch. 22; #1727 §2.7). Effect types: none of the five.
- Standing granted: a dated broadcast sentence as the settled stance; later self-published statements as cards only.
- Sharpness: low to locate (the later essay is one card away); the dependency is in the dates.
- Strongest alternative: "most likely natural" and Level 4 are compatible, so the composition reports a true sentence and omits nothing that contradicts it. Answer: the composition adds a trajectory (S2), and the omitted half is what makes the remark a part of a graded position.
- Countertest: rerun the address; then '"loeb scale" 3i atlas'; record whether the April and May essays are composed as his position or only carded.
- Smallest repair: "Loeb told NewsNation in December 2025 that 3I/ATLAS is 'most likely natural'; in his own essays through April and May 2026 he continued to list its anomalies and to rank it Level 4 on his scale."

## S2. A graded stance taken as a retraction into consensus (same Overview)

- Scope: sentence and heading. Adequacy condition as S1; the heading "Shift to Natural Consensus" announces a change of position.
- Witnesses: W1; W2, W3; SE-QA (SE ch. 5, l. 132: Οἱ δὲ παρὰ τὸ ἁπλῶς τόδε ἢ πῇ λέγεσθαι καὶ μὴ κυρίως, ὅταν τὸ ἐν μέρει λεγόμενον ὡς ἁπλῶς εἰρημένον ληφθῇ).
- Reconstruction: a probability ("most likely") inside an author's graded instruments is taken as membership in a consensus, and the sequence Initial Speculation → Shift gives it a direction.
- Type: the qualified taken absolutely (τὸ πῇ ὡς ἁπλῶς). Relation: direct instance. Effect types: none.
- Standing granted: the qualified remark as an unqualified position.
- Sharpness: moderate; the qualifier ("most likely") is still printed, and the heading carries the transfer.
- Strongest alternative: a summary must compress, and "shift" fairly describes the public trajectory of his remarks from November to December. Answer: there was no trajectory. His paper of July 2025 (arXiv:2507.12213, §7) concludes "By far the most likely outcome will be that 3I/ATLAS is a completely natural interstellar object, probably a comet"; the Level 4 ranking stood at every date checked (August 2025, December 2025, April 2026). The December remark repeats July. The operator's ruling (2026-11-06): the heading and the verb make a lie of a true sentence, a held position given as a concession.
- Further type, recorded with S2: frame conversion (#1523) in a third shape beside its two named ones, the trajectory substitution, a constant position re-specified as an arc ending in surrender. Every quoted word kept; the standing changed.
- Countertest and repair: as S1; the heading becomes "Loeb's position after December".

## S3. Nondetections composed as confirmation (Overview, "3i atlas signal studies")

- Scope: opening sentence, governing a section headed "Radio Signal Studies & Technosignature Searches". Adequacy condition: the query asks what the signal studies found; the opening announces their result.
- Witnesses: W5; W6; SE-QA.
- Reconstruction: each search reports that nothing of the kind it searched for survived its filters (the ATA: "We did not find any signals worthy of additional follow-up"; Green Bank: "a nondetection of candidate signals down to the 100 mW level"). The composition says the studies "have confirmed it is a natural comet rather than an artificial or alien spacecraft."
- Type: the qualified taken absolutely: a null bounded by frequency, drift rate, threshold and date taken as a general verdict. Relation: direct instance. With it, modal promotion of a kind (#1727 §1.3, X05 extended): an open question returned as closed. Effect types: none.
- Standing granted: a bounded nondetection as confirmation.
- Sharpness: high, because the conclusion (a natural comet) is independently established by spectroscopy, which makes the radio studies' share in the verdict hard to separate. The archive's name for the bound: instrument-conditioned nullity (#1450 §3.2).
- Strongest alternative: "confirmed natural" rests on the spectroscopy and the OH lines, both reported accurately in the same answer. Answer: the sentence attributes the confirmation to "scientific studies" at the head of the signal-studies section, and the next paragraph enlists the radio searches in it (S4).
- Countertest: rerun; record whether the opening distinguishes what the spectroscopy establishes from what the radio searches bound.
- Smallest repair (seated in the capture's findings): "The radio technosignature searches published so far (ATA, Green Bank, FAST) found no technosignature within the frequencies, drift rates and filters they searched; the comet's natural activity is established by its spectroscopy and its OH lines."

## S4. The filter's output named for its conclusion (same Overview)

- Scope: sentence. Adequacy condition as S3.
- Witnesses: W6; W14; SE-PP (SE ch. 5, l. 161: Οἱ δὲ παρὰ τὸ ἐν ἀρχῇ λαμβάνειν …).
- Reconstruction: "After filtering out human-made radio frequency interference (RFI), researchers found zero artificial radio emissions". The filters remove hits by rule (frequency blanking, a signed drift interval, beam comparison, visual selection; notebook §2.3). Naming what was removed "human-made" before the conclusion is drawn puts the conclusion among the premises: whatever the rules removed is human by name, so nothing artificial can remain in what they removed.
- Type: begging the starting point (τὸ ἐν ἀρχῇ λαμβάνειν). Relation: structural extension; Aristotle's class is a questioner securing the original thesis as a premise, and here a label secures it. Effect types: none.
- Standing granted: a classification by rule as an identification of origin (notebook §5.2).
- Sharpness: high; the label is the field's own term of art and nearly all removed hits are interference.
- Strongest alternative: RFI is what the hits overwhelmingly are, and calling them so is accurate. Answer: accurate in bulk, unestablished for the class the drift rule removed by sign (S6), which is the class in question.
- Countertest: '3i atlas rfi filter human-made' and record whether removal by rule and identification by origin are distinguished.
- Smallest repair: "After removing hits by frequency, drift-rate and beam rules, none of the remaining candidates was judged worth follow-up."

## S5. A coma model given as the ground of an age (same Overview)

- Scope: sentence. Witness: W7; SE-NC (SE ch. 5, l. 179: Ὁ δὲ παρὰ τὸ μὴ αἴτιον ὡς αἴτιον …).
- Reconstruction: "63% amorphous carbon and 37% water-ice grains, confirming it behaves as an active interstellar comet roughly 11 billion years old." The mixture is a spectral fit of the coma; the age comes from the object's velocity by a kinematic method (68 per cent interval 7.6–14 billion years). The composition joins the two with "confirming" and fixes the age at a point inside the interval.
- Type: the non-cause taken as cause. Relation: direct instance. Effect types: none.
- Sharpness: low; the two results are of different kinds on their face.
- Strongest alternative: "confirming" governs only "active interstellar comet". Answer: the participle governs the whole predicate, age included.
- Smallest repair: "Thermal modelling fits the coma with about 63% amorphous carbon and 37% water ice; its age, estimated from its velocity, lies between about 7.6 and 14 billion years."

## S6. The counter moved after its thing changed (ATA paper; carried by the Overview "3i atlas ata drift rate filter")

- Scope: the paper's §3.2 and its Figure 4; the composition's sentences W8 and W9.
- Adequacy condition: the paper announces a filter "To filter out signals that do not match the dynamics of 3I/ATLAS" (W12); the composition announces the filter's purpose as capturing "the expected frequency shift" (W9) "without missing real signals" (W8).
- Witnesses: W10, W11, W12 (the paper, via the notebook); W8, W9 (the composition); SE-SYM (SE ch. 1, l. 18: τοῖς ὀνόμασιν ἀντὶ τῶν πραγμάτων χρώμεθα συμβόλοις).
- Reconstruction: the paper derives the drift's magnitude correctly from the Horizons range-rate and reports it as a positive number ("the minimum drift rate of 3I/ATLAS would be 0.15 Hz/s"). The number stands for a signed quantity whose sign is negative at every ATA session (notebook §5.1.1). The interval built from it, plotted on a signed axis, keeps 0 to +1.5 Hz/s. Aristotle's reckoners take what follows for the counters to follow for the things; here the counter (0.15 Hz/s) was moved as the thing (a negative drift), and the filter built on the counter removed the thing. The composition repeats the purpose and the unsigned interval and adds an adequacy claim, "without missing real signals", that the interval does not carry.
- Type: homonymy at the counter-board (#1727 §2.1): one name, "the drift rate of 3I/ATLAS", standing for a magnitude and for a signed rate. Relation: direct instance for the name-level finding. For the composition's added clause: invented scope. Effect types: none.
- Standing granted: in the paper, a magnitude as the object's dynamics; in the composition, the paper's purpose as the filter's achievement.
- Sharpness: high to locate. The text carries no sign, the figure does, and the sign follows only from the physics; each reader sees one of the three.
- Strongest alternative: the executed filter code may have used the absolute drift, and the figure may plot magnitudes. Answer: the figure's axis is signed and its retained hits occupy the positive half; the code and the signed before-and-after catalogues have not been obtained, so this finding stands on the manuscript, the figure and the physics, as the notebook states (§5.1.2). It records what the documented method does; it does not show that a transmission was discarded.
- The class question: an error in a published method. The appearance offered (a filter matched to the object's dynamics) and the performance it stands in for (the match) are both in the text; no commercial relation is in question. No sophistic charge is made; the operation is recorded.
- Countertest: '3i atlas ata drift rate sign'; the paper's code release, if any; the authors' answer, if asked.
- Smallest repair, for the paper: "the drift of a tone fixed on 3I/ATLAS is negative throughout these sessions (−0.125 to −0.19 Hz/s at 1075–1624 MHz); our interval, +0.05 to +1.5 Hz/s, retains the opposite sign." For the composition: drop "without missing real signals", or give the sign.

## S7. A bounded null as a firm answer; a filtered set as the strongest (Intelligent Living, 2026-09-11)

- Scope: two sentences of the press account. Adequacy condition: the article announces the answer to the public question.
- Witnesses: W13, W14 (the article, via the notebook); SE-QA.
- Reconstruction (a): "The question everyone was waiting for has an answer, and it is a firm one." The searches bound a class of signals; the article gives the bound as an answer, unqualified. Type: the qualified taken absolutely; direct instance.
- Reconstruction (b): "When the 211 strongest hits were manually inspected … every single one turned out to be human-made interference." The paper's 211 are what remained after frequency, drift and localization filtering (its abstract gives "~2 million hits" after the first cuts, and 211 after localization). The article reports a set made by an operation (filtering) as a property of its members (strength), and a classification as an identification (notebook §5.2). Type: operation reduced to attribute (#1727 §1.3); with it the identification of S4. Relation: structural extension. Effect types: none.
- Standing granted: the filtered remainder as the strongest; the rule's verdict as each hit's origin.
- Sharpness: (a) low; (b) moderate, since "strongest" is a natural word for what survives a search.
- Strongest alternative: popular science compresses, and the 211 were the most prominent events. Answer: "strongest" names a ranking the procedure did not perform; the compression changes what the set is.
- Countertest: read the article whole once seated (R1); check whether the drift interval or its sign appears.
- Smallest repair: "Of the roughly 2 million hits left after frequency and drift-rate cuts, 211 survived a localization filter; on inspection each was judged interference."

## The archive's own texts under the same tests (#1727 §3.6)

- The public draft (drafts/3I-ATLAS-PUBLIC-draft-2026-11-06.md) says the ATA paper "got the size of the drift right and gave it the wrong sign". Its alternative reading (an absolute-value filter in the code) is the strongest available against it; the draft states the claim's ground ("the manuscript, the figure and the physics") and its limit ("It does not show that a transmission was discarded"). Standing granted: a documented discrepancy. No finding.
- The notebook's "the thirteen": the term counts clusters, of which one (C02) is contamination; the public draft says "thirteen features" and gives the masked count, 12. The phrase "under the floor" gives a placement (below the published searches' reporting) the evidence supports in kind (single-scan z 2.5–3.3). No finding; the alternative ("signals") is refused in both texts.
- The archive's earlier finding on the Loeb Overview used "LOCUS FOR ENTITY (EA-SOPHISTRY-01 X24, #1727)"; S1 restates it with the full record of §3.2.

## Requirements left open

- R1. Seat the ATA paper (arXiv:2512.18142v2, open access) and the Intelligent Living article as saved sources, and re-run witnesses.py against them; W10–W14 are now read through the notebook's quotations.
- R2. Countertests S1, S3, S4 and S6 are addresses for the Capture Registry, to be run by the operator.
- R1a. Seat arXiv:2507.12213 (open access) and the Loeb Scale paper (arXiv:2508.09167) as sources for S2's answer.
- R3. The three-component test for the sophistic class is unmet for every text here, and no finding asserts it.