Abstract
This corpus places a scope note beginning “this paper does not” in almost every paper. What this paper shows is that those notes fall into two groups whose effects are opposite: one aims the refutation, the other shrinks the claim. Which is better is not adjudicated here; the numbers are given and the judgement is returned to the author. No new theorem or law is claimed.
First, $\mathbf{516}$ of $\mathbf{524}$ papers, or $\mathbf{98.5}$ per cent, carry a scope note——with $\mathbf{3649}$ declarations ending in a negation, a minimum of $\mathbf{1}$, a median of $\mathbf{8}$ and a maximum of $\mathbf{14}$ (Section 2).
Second, and this is the core. The $\mathbf{8}$ papers without one all come from the earliest period——Papers 1, 2, 8, 12, 19, 20, 61 and 62, so the habit was adopted along the way rather than added in answer to criticism (Section 3).
Third, the papers with only $\mathbf{1}$ declaration are the renumbered ones and those filling gaps——Papers 23, 25, 31, 36 and 37, while the maximum of $\mathbf{14}$ belongs to Papers 269, 275, 281, 301 and 314 (Section 3).
Fourth, by word family there are $\mathbf{978}$ declarations narrowing the target, $\mathbf{728}$ lowering the strength of the claim and $\mathbf{283}$ declining a procedure——the first aims, the second shrinks (Section 4).
Fifth, the corpus calls itself falsifiable in $\mathbf{7}$ papers while naming neither Popper nor Lakatos in any——“protective belt” appears in exactly $\mathbf{1}$, Paper 481, which returns the judgement to the author (Section 4).
Sixth, the separator is whether removing the note would enlarge the claim——if it would, the note was aiming; if not, that was all that was ever said (Section 5).
Keywords: scope note, falsifiability, corpus census, Popper, Lakatos.
Scope (what is not claimed)
No new theorem or law is claimed——Popper’s falsifiability and Lakatos’s hard core and protective belt are standard terms. The corpus is not defended——this paper goes as far as the numbers; whether the scope notes form a protective belt or a hard core is returned to the author and is not adjudicated here. No dispute in the philosophy of science is settled——whether Popper or Lakatos is right is not the subject. No automatic classifier is built——the word-family counts propose candidates; classification item by item must be done by reading. The three families here are a guide, not a decision procedure. No comparison with other authors is made——only this corpus was counted, and the general frequency of scope notes in the literature was not measured. The quality of the notes is not assessed——how many there are and of what type, not how well written. Revision history is not traced——whether a note stood in the first version or was added later is inferred only from presence and absence across papers. Relation to earlier papers: Paper 481 showed that a broken law can be blamed on any assumption, and stated explicitly that whether the scope notes are a protective belt or a hard core is returned to the author——this paper takes that seat and supplies the numbers, while still returning the judgement. Paper 200 measured the corpus as its own data——the same gesture is used here, with the habit itself as the object measured. Paper 400 showed that the instrument broke before the thing it was counting——which is exactly why the word-family counts here must not be made into a decision procedure. Paper 500 showed that the corpus’s method is not one habit and that each was adopted at its own time——the $\mathbf{8}$ papers here are a concrete instance. Paper 300 showed that sharing a root is decidable——the two kinds of note are separate roots under the criterion of what a quantity carries. What is added is counting the presence of scope notes across the corpus, naming the $8$ that lack one, giving the distribution of declarations, counting the word families, and placing the separator on whether removal would enlarge the claim.
1. Notation and the Question
Setting: Popper held that falsifiability separates science from what is not science. Lakatos extended this to research programmes, distinguishing a hard core, which is not to be altered, from a protective belt, which absorbs refutations and may be rewritten.
The habit: almost every paper in this corpus places, immediately after the abstract, a short note beginning “this paper does not treat” or “no claim is made”. Call it a scope note.
Observation 1.1 one sentence form can do two opposite jobs. “This case is not treated” sharpens the claim about the rest——while “no claim is made that far” shrinks the claim itself.
Observation 1.2 in Lakatos’s terms the first exposes the hard core and the second thickens the protective belt. Both end in a negation and share a sentence form——appearance does not separate them.
Question: which are the $\mathbf{3649}$ declarations of this corpus?
2. First, Count
Computation 2.1: scanning every numbered paper as of $10$ September $2026$.
| quantity | value |
|---|---|
| papers scanned | $\mathbf{524}$ |
| papers carrying a scope note | $\mathbf{516}$ ($\mathbf{98.5}$ per cent) |
| declarations ending in a negation | $\mathbf{3649}$ |
| minimum per paper | $\mathbf{1}$ |
| median per paper | $\mathbf{8}$ |
| maximum per paper | $\mathbf{14}$ |
Observation 2.2 nearly every paper has one. $\mathbf{98.5}$ per cent——this is a habit of the corpus rather than a decision taken paper by paper, and a habit is not a response to a particular criticism.
Observation 2.3 the distribution has width. From $\mathbf{1}$ to $\mathbf{14}$——the papers are not reciting the same list; the median is $\mathbf{8}$.
Observation 2.4 only the Japanese versions were counted. The English versions use a different sentence form, so the same count does not apply——comparing them would need another method, which is not attempted here.
3. What the Eight Without One Reveal
Computation 3.1: naming the $\mathbf{8}$ papers with no scope note.
| papers | subject |
|---|---|
| numbers 1 and 2 | $n=2$ and $4\pi$ (the foundation; the two that cite nothing) |
| number 8 | the generalisation to $n=d-1$ |
| number 12 | $E=4\pi c^{2}$ |
| numbers 19 and 20 | the pressure and temperature side |
| numbers 61 and 62 | the light sphere and the volume duality |
Observation 3.2 this is the core. All $\mathbf{8}$ come from the earliest period——the scope note is a habit adopted along the way, not something added in answer to a particular criticism.
Observation 3.3 in Lakatos’s sense this bears on the distinction. Rewriting a protective belt happens after a refutation is met——a note placed uniformly is placed before. The direction of time differs.
Observation 3.4 this does not establish that they are not a protective belt. A net cast wide in advance can still let refutations through——one necessary condition has been eliminated, which is not a sufficient one.
Observation 3.5 the sparsest papers show a pattern too. Only Papers 23, 25, 31, 36 and 37 carry a single declaration——the renumbered ones and those filling gaps. Papers written at a turn in the habit have thinner notes.
Observation 3.6 the maximum is $\mathbf{14}$. Papers 269, 275, 281, 301 and 314——it stops there rather than growing without bound.
Observation 3.7 this instantiates Paper 500. There the corpus’s method was shown to be several habits, each adopted at its own time——the scope note is one of them, and the time of adoption survives as a paper number.
4. Two Word Families
Computation 4.1: counting the declarations by word family. This is a way of proposing candidates, not a decision procedure.
| family | count | in Lakatos’s terms |
|---|---|---|
| narrowing the target (not treated, not entered) | $\mathbf{978}$ | aiming |
| lowering the strength (no claim, not called) | $\mathbf{728}$ | shrinking |
| declining a procedure (not discussed, not adjudicated) | $\mathbf{283}$ | either |
Observation 4.2 it is not one or the other. $\mathbf{978}$ against $\mathbf{728}$——neither “all protective belt” nor “all hard core” fits these numbers.
Observation 4.3 the effects run opposite. A declaration that narrows the target strengthens the claim over what remains——write “only for small $\beta$” and a failure at large $\beta$ is no refutation, but a failure at small $\beta$ leaves nowhere to go.
Observation 4.4 a declaration that lowers the strength works the other way. “No new theorem or law is claimed” appears in almost all $\mathbf{524}$——this shrinks the target. The less there is that could be wrong, the less can be shown wrong.
Observation 4.5 the corpus calls itself falsifiable in $\mathbf{7}$ papers. Popper and Lakatos appear in $\mathbf{0}$, “protective belt” in $\mathbf{1}$——that one is Paper 481, which returns the judgement to the author. This paper takes the seat and still returns it.
Observation 4.6 the family counts are only an instrument. One sentence can do both jobs at once——as Paper 400 put it, the instrument broke before the thing it was counting, and this threefold split would move if read item by item. It must not be made into a decision procedure.
5. The Separator
Observation 5.1 the separator is whether removing the note enlarges the claim. If it does, the note was aiming——if it does not, that was all that was ever said.
Observation 5.2 item by item, this can actually be tried. Delete the note and see whether the remaining text becomes inconsistent or stays put——a machine cannot do it, but a reader can. This paper goes as far as proposing the procedure.
Observation 5.3 separate roots under the criterion of Paper 300. An aiming note needs the belonging “which range”; a shrinking note needs none——the same sentence form, different belongings.
Observation 5.4 no judgement is given. The numbers $\mathbf{978}$ against $\mathbf{728}$ and the fact that the $\mathbf{8}$ are the earliest are set side by side——whether these scope notes are a protective belt or a hard core is for the author to decide. This paper lays out the material.
Observation 5.5 what is not said. No comparison with other authors was made——the general frequency of scope notes was not measured, so whether $\mathbf{98.5}$ per cent is high or low cannot be stated.
6. Roll Call and Closing
- $\mathbf{516}$ of $\mathbf{524}$ carry one (Computation 2.1): $\mathbf{98.5}$ per cent.
- $\mathbf{3649}$ declarations (Computation 2.1): minimum $1$, median $8$, maximum $14$.
- the $\mathbf{8}$ without one are earliest (Computation 3.1): the core.
- numbers 1 and 2 also cite nothing (Computation 3.1): the two foundations.
- single-declaration papers are renumbered or gap-filling (Observation 3.5): a turn in the habit.
- the maximum is $\mathbf{14}$ (Observation 3.6): not unbounded.
- $\mathbf{978}$ narrow the target (Computation 4.1): aiming.
- $\mathbf{728}$ lower the strength (Computation 4.1): shrinking.
- falsifiable in $\mathbf{7}$, Popper in $\mathbf{0}$ (Observation 4.5): the word absent, the substance present.
- the separator is removal (Observation 5.1): a reader decides.
- the judgement is returned (Observation 5.4): material only.
Closing: This corpus places a scope note beginning “this paper does not” in almost every paper. First, the counting——as of $10$ September $2026$, $\mathbf{516}$ of $\mathbf{524}$ papers, or $\mathbf{98.5}$ per cent, carry one, with $\mathbf{3649}$ declarations ending in a negation, a minimum of $\mathbf{1}$, a median of $\mathbf{8}$ and a maximum of $\mathbf{14}$. The papers are not reciting one list; the distribution has width. Only the Japanese versions were counted, since the English versions use a different sentence form. The eight without one give a date——they are Papers 1, 2, 8, 12, 19, 20, 61 and 62, all from the earliest period, so the note is a habit adopted along the way and not an answer to a particular criticism. In Lakatos’s sense this bears on the distinction, since rewriting a protective belt happens after a refutation is met while a uniform note is placed before: the direction of time differs. It does not settle the matter, because a net cast wide in advance can still let refutations through; one necessary condition has been eliminated, not a sufficient one. The papers with a single declaration, 23, 25, 31, 36 and 37, are the renumbered ones and those filling gaps, and the maximum stops at $\mathbf{14}$ rather than growing without bound. By family, two groups pull opposite ways——$\mathbf{978}$ declarations narrow the target, $\mathbf{728}$ lower the strength of the claim, and $\mathbf{283}$ decline a procedure. Narrowing aims: state the range and a failure inside it leaves nowhere to go. Lowering shrinks: “no new theorem or law is claimed” appears in almost every paper and reduces what could be shown wrong. Neither “all protective belt” nor “all hard core” fits $\mathbf{978}$ against $\mathbf{728}$. The separator is whether removing the note enlarges the claim——if it does, the note was aiming; if not, that was all that was ever said, and item by item a reader can test this by deleting the note and seeing whether the text becomes inconsistent. Under the criterion of Paper 300, an aiming note needs the belonging “which range” and a shrinking note needs none, so the same sentence form carries different belongings. The corpus calls itself falsifiable in $\mathbf{7}$ papers while naming Popper and Lakatos in none, and “protective belt” appears only in Paper 481, which returns the judgement to the author. That is left unchanged here. To be honest——no judgement was given. The numbers $\mathbf{978}$ against $\mathbf{728}$ and the earliness of the $\mathbf{8}$ were set side by side, and nothing more. The family counts propose candidates and are not a decision procedure, since one sentence can do both jobs; as Paper 400 recorded, the instrument breaks before the thing it counts. No comparison with other authors was made, so whether $\mathbf{98.5}$ per cent is high or low cannot be said. The quality of the notes was not assessed, and whether each stood in a first version was inferred only from presence and absence. What can be said is that one sentence form covers two groups with opposite effects in the ratio $\mathbf{978}$ to $\mathbf{728}$, and no further.
References
- Duhem, P. (1906). La Theorie physique: son objet et sa structure. Chevalier et Riviere.
- Lakatos, I. (1970). Falsification and the Methodology of Scientific Research Programmes. In I. Lakatos & A. Musgrave (Eds.), Criticism and the Growth of Knowledge. Cambridge University Press.
- Popper, K. R. (1959). The Logic of Scientific Discovery. Hutchinson.
- Yamagishi, Y. (2026). The Two Hundredth Census: Measuring the Corpus as Its Own Data [Paper 200]. Zenodo. 10.5281/zenodo.22108829
- Yamagishi, Y. (2026). The Three Hundredth Census: Whether Two Things Share a Root Is Decidable [Paper 300]. Zenodo. 10.5281/zenodo.22171914
- Yamagishi, Y. (2026). The Four Hundredth Census: The Instrument Broke Before the Thing It Was Counting [Paper 400]. Zenodo. 10.5281/zenodo.22699343
- Yamagishi, Y. (2026). A Broken Law Can Be Blamed on Any of the Assumptions [Paper 481]. Zenodo. 10.5281/zenodo.22733927
- Yamagishi, Y. (2026). The Five Hundredth Census: The Corpus’s Method Is Not One Habit, and Each Was Adopted at Its Own Time [Paper 500]. Zenodo. 10.5281/zenodo.22760262
Reference
Yamagishi, Y. (2026). Does Writing “This Paper Does Not” Evade Refutation, or Aim It? [Paper 526]. YOI: https://www.yamagishiyuuki.com/yoi/526
DOI: 10.5281/zenodo.22867648
ORCID: 0009-0006-8587-3450
PDF (English) · PDF (Japanese) · Date proof of the English PDF (OpenTimestamps .ots) · SHA256SUMS
The record page for this paper (in Japanese, with the PDFs in each language and the date proof): YOI 526. This paper is a preprint and has not been peer reviewed.