{"id":"b274cd0c-cb3c-4a79-b12f-d3d1b00fc9ab","arxiv_id":"2506.07131","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"A computer-assisted reading of Wittgenstein's Nachlass is presented as evidence that 'meaning as use' persists continuously from his 1914 notebooks to his final manuscripts, supporting a proof-theoretic 'meaning as use' semantics.","lead":"Using the online Wittgenstein Nachlass archive, this paper argues that a thread connecting meaning with use, purpose, and application runs through Wittgenstein's writings from 1914 to 1951. It then links this reading to a program that treats proof-reduction rules in logic and the lambda calculus as the formal counterpart of 'meaning as use'.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Continuity thesis depends on unreported sampling: no queries, counts, or negative cases are given, so the quoted Nachlass passages may illustrate rather than establish a corpus-wide meaning-as-use thread.","rationale":"I focused on the sampling premise because it is the pivot of the abstract's strongest claim. The formal section (§3) restates standard β-reduction rules and the Curry-Howard mapping correctly; whatever one thinks of 'most basic settlement,' the technical content is not where the paper can be shown to fail. The historical claim, by contrast, is an existential generalization from a sample, and no sampling information is reported. The reader's weakest_assumption already names representativeness and the footnote-1 looping risk, and I agree. Independently of that, a systematic count would either rescue or falsify the continuity thesis, so the conditional verdict is the right one; I would not tighten to ACCEPT or relax to REJECT on the present evidence.","tokens_in":27912,"tokens_out":3602,"duration_ms":40313,"concrete_test":"Run a systematic collocation search over the WAB Nachlass: for each manuscript, count occurrences of Gebrauch, Anwendung, Verwendung, and Zweck within a fixed window (say ±5 tokens) of Bedeutung or Sinn, normalize by manuscript length, plot by year, and compare with the co-occurrence rate of Bedeutung/Sinn with Bild, Wahrheit, Vorstellung, or psychischer Vorgang. Publish the queries and raw counts. If the meaning-use collocation is not well above baseline and spread across the 1910s–1950s, the continuity claim is unsupported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract and §1 assert that the meaning/use association 'shows itself from the very beginning through to the very late writings' and that the Nachlass 'strongly support[s] and augment[s]' the meaning-as-use reading. The evidence offered is a curated set of quotations, each plausibly transcribed and dated, but the paper never specifies the search queries, occurrence counts, inclusion/exclusion criteria, or negative cases that would let a reader check that the quotes are representative. Footnote 1 says the manuscript 'substantially reinforces earlier arguments,' which makes the selection mechanism especially load-bearing: if the quoted passages are atypical, or if terms like Gebrauch and Verwendung typically appear in non-semantic contexts or alongside rival meaning-fixing notions, the continuous-thread thesis loses its evidence base. The paper's own quotations contain counter-currents, such as Ms-107's 'Can one say: the meaning of a sentence is its purpose?' posed as an open question, and 'logic cannot address the natural history of the use of a word,' so a corpus-wide distribution is genuinely needed rather than decorative. Because the conclusion leans on the archive 'at large,' unmeasured selection is the weakest load-bearing premise.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper argues that Wittgenstein's Nachlass, from the WW1 Notebooks through very late manuscripts of 1950–51, exhibits a continuous association of meaning with use, application, employment, purpose, and usefulness, thereby challenging the sharp early/late Wittgenstein divide. The evidence is a series of quoted passages, each cited to a specific manuscript page with an access date, together with a proposed formal counterpart: reduction rules in the lambda calculus and in natural deduction are presented as the proof-theoretic/dialogical realization of 'meaning as use'. The paper also claims that Wittgenstein never abandoned the view of language as a calculus, and that the later focus on ordinary language is a return to his early roots rather than a rejection of them.","tokens_in":28033,"tokens_out":4659,"duration_ms":49145,"significance":"If the continuity thesis is accepted, the paper offers a significant revision of standard Wittgenstein scholarship and gives proof-theoretic semantics a direct Wittgensteinian pedigree. The archival excerpts are valuable and precisely located, and the formal material on β-reduction and natural-deduction reduction rules is standard and correctly stated. The strength of the historical conclusion, however, depends on an unreported selection of passages; the paper does not yet establish that the quoted material is representative of the Nachlass as a whole. The significance is therefore real but conditional on a load-bearing methodological point.","major_comments":[{"comment":"The central claim that the meaning/use association 'shows itself from the very beginning through to the very late writings' and that the Nachlass 'strongly support[s] and augment[s]' the meaning-as-use reading is supported only by a curated set of quotations. The paper reports no search queries, no occurrence counts, no inclusion/exclusion criteria, and no negative cases that would let a reader judge whether the passages are representative of the corpus. Footnote 1's admission that the text 'substantially reinforces earlier arguments' makes the selection mechanism especially load-bearing. Given that §4 concludes from the Nachlass 'at large', the paper should either supply a documented search methodology or explicitly restrict the conclusion to the passages quoted.","section":"Abstract; §1; §4"},{"comment":"The same quoted passage that is offered as early second-phase evidence contains two counter-currents to the thesis: 'Kann man sagen: der Sinn eines Satzes ist sein Zweck?' is posed as an open question, and 'Die Logik kann aber nicht die Naturgeschichte des Gebrauchs eines Worts angehen' explicitly limits what logic can do with use. The paper does not discuss these tensions, so the page reads as more univocal than the source text warrants. The argument needs to address these counter-currents explicitly, or the historical claim should be weakened to say that the association appears in, rather than uniformly governs, these texts.","section":"§1, Ms-107, p. 7"},{"comment":"The assertion that reduction rules constitute 'the most basic settlement of meaning for predicate logic' is not derived from the quoted Wittgenstein texts, which say nothing about introduction/elimination rules or Attacker/Defender roles. The mapping of constructors/destructors to Assertion/Attack/Defense is a proposal from the author's earlier functional/dialogical framework, as the dense self-citation in footnote 35 indicates. The paper should clearly mark this as a proposed formal counterpart rather than a reading directly evidenced by the Nachlass, and should justify why these specific reduction rules are the privileged formalization among other proof-theoretic options.","section":"§3.2"}],"minor_comments":[{"comment":"The section heading 'Predicated logic and rules of (proof-)reduction' contains a typo; it should read 'Predicate logic'.","section":"§3.2"},{"comment":"The access dates for Ms-117 appear twice as '16 Jaan 2025'; both should read '16 Jan 2025'.","section":"§1, Ms-117"},{"comment":"The in-text reference 'Martin-L¨or 2022' and the reference-list entry 'Corrections of Assertion and Validity of Inference' should be 'Martin-L¨of 2022' and 'Correctness of Assertion and Validity of Inference', respectively.","section":"§3.2; References"},{"comment":"The name 'Ehrenfecht-Fra¨ıss´e' in the text and the reference 'V¨an¨a¨anen' contain typographical errors; they should be 'Ehrenfeucht-Fraïssé' and 'Väänänen'.","section":"§3.2; References"},{"comment":"The paper would be easier to verify if every quoted German passage identified the transcription layer used (e.g., user-filtered vs. diplomatic); currently only the Ms-131 quotation carries this information.","section":"§1 and §2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is part of a long, heavily self-cited research program, and this is not by itself disqualifying. However, the absence of systematic search reporting makes it difficult to separate the historical finding from the framework's assumptions. The editor may also wish to consider whether a journal in the history of philosophy expects corpus-level claims to be supported by quantitative or at least systematically described archival evidence when the conclusion is framed as being about the Nachlass 'at large'."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things about this paper. First, it puts real primary-source evidence on the table: a set of dated Nachlass passages, from the 1914-16 Notebooks through 1950-51, that tie meaning to Gebrauch, Anwendung, Verwendung, and Zweck. Each quote comes with manuscript number, folio, and an access date from the Bergen archive. If you care about the early/late Wittgenstein question, this is a useful collection. Second, the paper's central claim, that this association 'shows itself from the very beginning through to the very late writings,' is stronger than the reported method can support. The paper says it used computer-assisted search but reports no queries, no occurrence counts, no inclusion/exclusion criteria, and no negative cases. So the quotes may illustrate the author's thesis rather than establish that the theme runs through the whole corpus. Footnote 1 admits the text 'substantially reinforces earlier arguments,' which makes the selection pressure explicit.\n\nWhat's actually new: the late-manuscript passages (Ms-137, 173, 175, 176, 1948-51) and the cross-corpus mapping of the four German terms. The author has made this argument for decades, but this paper extends the evidence base into the very late period. The formal section restates standard beta-reduction and natural-deduction rules correctly; the pairing of constructors/destructors with Assertion/Attack/Defense is a genuine interpretive proposal, but it's an analogy, not a derivation. Calling it 'the most basic settlement of meaning for predicate logic' overreaches; that's a contestable philosophical claim, not a result.\n\nThe soft spots are real but not fatal to the historical core. The sampling problem is the load-bearing one: the paper itself quotes passages that cut against the smooth continuity story (Ms-107's 'Can one say: the meaning of a sentence is its purpose?' as an open question; 'logic cannot address the natural history of the use of a word'), and it doesn't tell us how common these terms are in non-semantic contexts. A reader can't check representativeness. The self-citation pattern is heavy, but the evidence is independent primary text, so I don't hold that against the substance.\n\nWho it's for: scholars of Wittgenstein's Nachlass, proof-theoretic semanticists, and historians of logic. It deserves a serious referee. I'd send it to someone with both Wittgenstein and formal logic competence, and ask for revision that reports the search methodology and addresses the counter-currents.","headline":"A useful collection of dated Nachlass passages and a serious, but methodologically under-reported, continuity thesis.","tokens_in":28715,"tokens_out":2116,"would_cite":true,"duration_ms":21720,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["03A05","03B40","03F05"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper argues that a continuous 'meaning as use' thread runs through Wittgenstein's whole Nachlass, from 1914 to 1951, and that reduction rules in natural deduction and the lambda calculus are its formal counterpart.","keywords":["meaning as use","Wittgenstein's Nachlass","Gebrauch","Anwendung","Verwendung","Zweck","reduction rules","proof-theoretic semantics"],"falsifier":"A systematic corpus count of Gebrauch, Anwendung, Verwendung, and Zweck co-occurring with Bedeutung or Sinn across all dated Nachlass manuscripts, with dates and manuscript identifiers, would settle the claim: if the co-occurrence clusters in only some periods, or if a late manuscript (1948–51) explicitly denies that a word's use fixes its meaning, the continuous-thread reading would be falsified.","tokens_in":27525,"feed_emoji":"📜","tokens_out":14855,"duration_ms":128466,"temperature":0.7,"pith_summary":"This paper argues that a single thread runs through all of Wittgenstein's writings, from the 1914–1916 notebooks to the manuscripts of 1950–51: meaning ties to the use, application, purpose, and usefulness of signs. Drawing on a searchable online edition of the Nachlass, the paper quotes passages where the German terms Gebrauch, Anwendung, Verwendung, and Zweck explain meaning and sense, and claims these passages support a 'meaning-as-use' semantics for predicate logic. It then presents the reduction rules of natural deduction and the $\\beta$-reduction rule of the $\\lambda$-calculus as the formal counterpart of that semantics, so that meaning is settled by consequences before any set-theoretic model is consulted. The payoff, if the claim holds, is that the standard early/late Wittgenstein divide collapses and proof-theoretic/dialogical semantics gains a direct Wittgensteinian pedigree.","feed_headline":"Wittgenstein tied meaning to use from 1914 to 1951, archive shows","feed_subtitle":"If right, the early/late crack in Wittgenstein closes and meaning-as-use gets a formal home.","key_machinery":"The machinery pairs a philological lever with a formal lever. The philological lever is the cluster of German terms Gebrauch (use), Anwendung (application), Verwendung (employment), and Zweck (purpose) as they recur in the searchable Nachlass; their presence across six decades carries the continuity claim. The formal lever is the reduction rule — $\\beta$-reduction in Church's $\\lambda$-calculus and proof-reduction in Gentzen-style natural deduction — treated as the formal face of meaning-as-use: introduction rules are assertion moves, elimination rules are attack moves, and reductions are the defenses that reveal consequences.","core_discovery":"The central discovery claimed is that the association of meaning with use, application, purpose, and usefulness 'shows itself from the very beginning through to the very late writings.' The Nachlass passages gathered here run from the WW1 Notebooks (1914–1916), through the 1929–1933 manuscripts in which Wittgenstein asks whether the sense of a sentence is its purpose and says a sign has meaning only when it finds use, to the 1948–1951 'Last Writings' manuscripts connecting meaning with consequences and application. The paper reads even Tractatus remarks 3.326–3.328 as already demanding attention to significant use and employment, and a 1912 letter to Russell as tying the meaning of the generality sign to its inferential use. On the formal side, the paper's claim is that the most basic settlement of meaning for predicate logic is given by reduction rules: the elimination rules acting on the results of introduction rules display a proposition's consequences, and this is what purpose and usefulness look like in a formal system.","pith_inferences":["A corpus-wide frequency analysis of Gebrauch, Anwendung, Verwendung, and Zweck co-occurring with Bedeutung and Sinn across all dated manuscripts would test the continuity claim directly and independently of the quoted sample.","If the continuity claim is right, the same cluster of terms should also organize Wittgenstein's remarks on mathematics, color, and certainty, offering a unified entry point for teaching the later philosophy.","The dialogue-game reading of reduction rules suggests a testable correspondence between normalization theorems (for example, strong normalization of simply typed $\\lambda$-calculus) and the requirement that every play in a suitable semantic game terminates.","The same search method could be turned on other word clusters such as Regel, Spiel, and Kriterium to check whether the use-terms are uniquely central or merely one of several threads."],"forward_implications":["The early/late distinction collapses: the meaning-as-use reading is not a second-phase invention but a constant of Wittgenstein's philosophy.","Tractatus remarks 3.326–3.328 already commit Wittgenstein to a use-based account of symbols, complicating the picture-theory reading of the early work.","Proof-theoretic semantics and dialogical/game-theoretical semantics acquire a direct Wittgensteinian lineage, with reduction rules as the formal rendering of meaning-as-use.","The meaning of logical constants in predicate logic is fixed by introduction and elimination rules together with their reductions, before any model-theoretic semantics is invoked.","In the $\\lambda$-calculus, $\\beta$-reduction is the primary meaning-constituting operation, prior to set-theoretic or denotational interpretations."],"supporting_citations":[{"why":"Supplies the searchable corpus from which the quoted Nachlass passages are drawn.","marker":"Wittgenstein 2016"},{"why":"Provides the WW1 Notebooks, the earliest dated occurrences of the use-terms in relation to meaning.","marker":"Wittgenstein 1961"},{"why":"The Tractatus remarks 3.326–3.328 anchor the early-period use-based reading.","marker":"Wittgenstein 1922"},{"why":"A 1912 letter to Russell ties the meaning of the generality sign to its inferential use.","marker":"Wittgenstein 1974"},{"why":"Its rule of application is presented as the meaning-giving rule of the $\\lambda$-calculus.","marker":"Church 1936"},{"why":"Its introduction and elimination rules and inversion principle underpin the reduction-rule account of logical constants.","marker":"Gentzen 1935"},{"why":"Its account of assertion as speaker-hearer interaction frames meaning as the purpose of an utterance.","marker":"Martin-Löf 2022"},{"why":"The Utterer/Interpreter distinction supplies the dialogical vocabulary mapped onto introduction and elimination rules.","marker":"Peirce 1932"},{"why":"Game-theoretical semantics for quantifiers supplies the Myself/Nature pairing used in the same mapping.","marker":"Hintikka 1968"}],"fun_headline_variants":["Wittgenstein's meaning-as-use spans his entire archive","From 1914 to 1951, Wittgenstein tied meaning to use","Wittgenstein's early and late writings agree: meaning is use","Nachlass shows meaning-as-use in Wittgenstein from start to finish","Reduction rules formalize Wittgenstein's meaning-as-use"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the quoted passages are representative of the whole Nachlass: because the paper reports no search queries, occurrence counts, inclusion or exclusion criteria, or negative cases, the continuity claim stands or falls on the representativeness of its selection.","fun_headline_variants_meta":{"raw":{"variants":["Wittgenstein's meaning-as-use spans his entire archive","From 1914 to 1951, Wittgenstein tied meaning to use","Wittgenstein's early and late writings agree: meaning is use","Nachlass shows meaning-as-use in Wittgenstein from start to finish","Reduction rules formalize Wittgenstein's meaning-as-use"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000562,"raw_usage":{"total_tokens":2731,"prompt_tokens":1068,"completion_tokens":1663,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":684,"completion_tokens_details":{"reasoning_tokens":1577}},"tokens_in":684,"tokens_out":1663,"duration_ms":14002,"temperature":1.0,"reasoning_tokens":1577,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:42:40.387161+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A systematic corpus count of Gebrauch, Anwendung, Verwendung, and Zweck co-occurring with Bedeutung or Sinn across all dated Nachlass manuscripts, with dates and manuscript identifiers, would settle the claim: if the co-occurrence clusters in only some periods, or if a late manuscript (1948–51) explicitly denies that a word's use fixes its meaning, the continuous-thread reading would be falsified.","supporting_citations":[{"cited_title":"Edited by the Wittgenstein Archives at the University of Bergen (W AB) under the direction of Alois Pichler","cited_arxiv_id":null,"evidence_quote":"Supplies the searchable corpus from which the quoted Nachlass passages are drawn."}],"review_version":1}