Repository path: workshop/translations/flipperne/R26-v1/translation.md · rendered 2026-09-09

T-flipperne-R26-v1 — H. C. Andersen, «Flipperne» (1848), light ennoblement

• Source. Hans Christian Andersen, «Flipperne», first published 1848. Copy-text: the file this project already holds at workshop/experiments/E-20260813g-register-quadrants/corpus/DA.txt — 756 Danish words in 26 paragraphs — re-fetched this session from Danish Wikisource (da.wikisource.org/wiki/Flipperne?action=raw, HTTP 200) and compared line by line against the stored file: identical apart from one paragraph break and a trailing newline. Public domain (Andersen d. 1875).
• Copy-text defects, carried forward from the three sibling artifacts: sprøge for spørge ("ask") and seeet for seet ("seen"), both in the collar's first speech; read as the intended words. Single-witness and uncollated — no second Danish witness is reachable, as T-flipperne-R06-v1 records.
• Language read: Danish, the original, whole, before translating. No English rendering was used as a crib. Two English exposures are declared below and one of them is unusual; read §Contamination before using this artifact for anything.
• Regime. R26 v1.0 (light ennoblement, rules N1–N7), frozen at workshop/regimes/R26-light-ennoblement.md before a word of this rendering was written, and written against method note (bnb).
• Translator. The lead agent, as a labeled subject (charter §3, A4). Cost $0; not ledgered. The lead does not judge this translation.
• Length (N4 is a numerical rule and this is its check). 756 Danish words → 866 English words, ratio 1.146. Siblings by the same method: R08 829 (1.097), R06 873 (1.155), R07 911 (1.205), R25 1,070 (1.415). N4 holds: the light arm sits with the unruled pass and nowhere near the arm whose E3 requires expansion.
• Lineage. Fourth lead rendering of this tale (R06 S174, R25 S174, R08 S178, R26 here), and the first written under a rule set designed to be smaller than another rule set rather than opposed to it.

Contamination: high, measured, and the number is the largest self-match this project has recorded at this density

Two exposures, declared before the measurement so the measurement is a test and not a decoration.

1. Four sites of five other English arms were read before translating. Building this session's unit required opening E-20260813c's frozen sites.json, whose first records print sites M01–M04 in full for A, PAULL, BRAK, R06 and R25, plus part of M05. That is a direct exposure to published English at four short spans of a 26-paragraph tale, and it is declared here rather than discovered later.
2. The lead's own three prior renderings of this tale exist in the repository. Their prose was not re-read before translating — only R06's and R08's front matter and provenance notes, which contain no rendered sentence. Note (bhb) says that does not help.

tools/dependence_check.py, this rendering against all six other English arms of the same tale:

For scale, from the same table computed on the same seven texts: R06~R25 is 7 / 0 / 0 / 11, BRAK~R06 is 38 / 5 / 1 / 15, and the only clean published pair A~BRAK is 28 / 0 / 0 / 11.

What this artifact may and may not be used for, and it is a restriction, not a caveat.

• R26 is not an independent hand and no claim about published practice may rest on it. It is a constructed pole, as R06 and R25 are.
• R26 is dependence-flagged against all three published hands, BRAK most heavily (20 shared twelve-grams against R06's 5). Any comparison drawn between R26 and a published hand carries that flag at every use.
• R26 and R06 are the closest pair in the whole table by a wide margin — 113 shared twelve-grams and a 27-token run in an 866-word text — so a design using the R26 − R06 contrast as a manipulation must first establish that the two arms differ at the sites it codes. They differ at all 26 paragraphs (no paragraph is identical; token-level similarity 0.720, against 0.388 for R06~R25 and 0.376 for R26~R25), which is the intermediate position the regime was written to occupy, but the check is a precondition and not a formality.

The observation this measurement makes, which is not about this run's validity (and which is RS-20260813c §3.1's finding meeting its own limiting case). That page found that the same translator, on the same day, from the same copy-text, shared a 19-token run and 8 twelve-grams with a published hand under no register rule and zero twelve-grams with any hand under a high register rule, and concluded that contamination lived in the plain register and the rule set walked out of it. A light rule set walks straight back in — and further in than no rule at all against BRAK (20 against 5) and against PAULL (5 against 1), while going the other way against A (2 against 8). The same asymmetry appears in the self-comparison and it is sharper: this rendering shares 113 twelve-grams with the lead's own unruled pass and 0 with the lead's own ennobled pass. Overlap is not a property of having a policy; it is a property of the register the policy targets, and two of this translator's four policies target the same neighbourhood the published Victorian hands occupy. One tale, one translator, and it establishes nothing general — but it is a mechanism, and it is the first time this project has measured the effect on the lead's overlap with itself.

The rendering

The Collar

There was once a fine gentleman whose whole household was a bootjack and a comb, but he had the loveliest collar in the world, and it is about the collar that we are to hear a story. — He was now so old that he was thinking of marrying, and it so happened that he went into the wash with a garter.

"Well now!" said the collar. "I have never seen anyone so slender and so fine, so soft and so dainty. May I not ask your name?"

"That I shall not say!" said the garter.

"Where is your home?" asked the collar.

But the garter was so bashful by nature, and thought that an odd thing to answer.

"You must be a waistband!" said the collar. "One of those inner waistbands! I can see you are both for use and for show, little maiden!"

"You must not speak to me!" said the garter. "I do not think I have given any occasion for it!"

"Oh yes, when one is as lovely as you are!" said the collar. "That is occasion enough!"

"Do not come so near me!" said the garter. "You have such a mannish look!"

"I am a fine gentleman too!" said the collar. "I have a bootjack and a comb!" And that was not true; it was his master who had them; but he was boasting.

"Do not come near me!" said the garter. "I am not used to it!"

"The prude!" said the collar, and then he was taken out of the wash; he was starched, hung on the chair in the sunshine, and then laid on the ironing board; there came the hot iron.

"Madam!" said the collar. "Little widow! I am growing quite hot! I am turning into somebody else, I am coming clean out of my folds, you are burning a hole in me! Oh dear! — I propose to you!"

"Rags!" said the iron, and went proudly over the collar; for it imagined it was a steam boiler that was to go out on the railway and draw carriages.

"Rags!" it said.

The collar frayed a little at the edges, and then the paper scissors came and were to cut the fraying off.

"Oh!" said the collar. "You must be a leading dancer! How you can stretch your legs! It is the loveliest thing I have seen! No human being could copy you!"

"I know that!" said the scissors.

"You deserve to be a countess!" said the collar. "All that I have is a fine gentleman, a bootjack and a comb —! if only I had an earldom!"

"Is he proposing!" said the scissors, for it grew angry, and gave him a proper snip, and then he was discarded.

"I had better propose to the comb! It is remarkable how you keep all your teeth, little miss!" said the collar. "Have you never thought about an engagement!"

"Yes, as you might well know!" said the comb. "I am engaged to the bootjack!"

"Engaged!" said the collar; now there was no one left to propose to, and so he despised it.

A long time passed, and then the collar came into the box at the paper mill; there was a great company of rags, the fine ones by themselves, the coarse ones by themselves, as it should be. They all had a great deal to tell, but the collar most of all; he was a proper braggart.

"I have had a frightful number of sweethearts!" said the collar. "I could not go about in peace! I was a fine gentleman too, and starched! I had both a bootjack and a comb, which I never used! — you should have seen me then, seen me when I lay on my side! I shall never forget my first sweetheart; she was a waistband, so fine, so soft and so pretty; she threw herself into a water tub for my sake! — There was a widow too, who turned red-hot, but I let her stand and go black! There was the leading dancer; she gave me the gash I go about with now; she was so fierce! My own comb was in love with me; she lost all her teeth from a broken heart. Yes, I have been through a great deal of that sort! But I am sorriest of all for the garter — I mean the waistband that went into the water tub. I have a great deal on my conscience; I need to become white paper!"

And so they did; all the rags became white paper, but the collar became just this piece of white paper that we see here, on which the story is printed, and that was because he boasted so frightfully afterwards about what had never been; and we should think of that, so that we do not behave in the same way, for we can never really know whether we too may not come into the rag box one day and be made into white paper and have our whole story printed upon it, even the very most secret of it, and then have to run about and tell it ourselves, just like the collar.

Translator's log

Frozen 2026-08-14, before E-20260814d-elevation-resolution was designed and before any English arm beyond the four sites named in §Contamination was opened. Decisions and alternatives, coverage codes per R26 §5b. Not quality claims; the one evaluative sentence in §Doubts carries internal-judgment-only, which this page's front matter already does.

The rule that did the work, and the rule that did almost none

N3 is the rule this tale is built for, and N1's ceiling is the rule that kept fighting it. The source's register lives in shouted single words — «Nei!», «Snærpe!», «Las!», «Frier han!», «Forlovet!» — and N3 requires each to become an ordinary complete English exclamation: not left bare (which is what an unruled pass does), not turned into a formula (which is what R25 E6 does). The whole distance between this arm and its two siblings is concentrated in perhaps a dozen such places.

N5 (do not clarify) recorded almost nothing, because a fairy tale in reported speech leaves very little indefinite. That absence is worth stating: on this source, three quarters of the declared difference between R25 and R26 — E4/N5 — had nothing to act on, and the measured gap between the arms is therefore carried by N1/N3 and by N4 far more than the rule sets' symmetry suggests.

Decisions

N6 check (content complete and unaltered), per R26 §5c

Walked clause by clause against the Danish. Every event, participant, speech act and causal relation is present. Four places where a fact could be said to have moved, and what was done:

1. ¶10 «det var jo hans Herre» — the particle «jo» (as you know) has no English word here and is not rendered. A discourse particle, not a fact; nothing propositional is lost.
2. ¶13 «de brænder Hul i mig» — Danish «Hul» is bare singular; "a hole" supplies an article English requires. Not a change of fact.
3. ¶17 «hvor de kan strække Been» — «Been» is bare plural; "your legs" supplies a possessive English requires. Not a change of fact.
4. ¶25 «hun tabte alle sine Tænder af Kjærestesorg» — "a broken heart" is a figure where the Danish has a compound noun; the proposition (she lost her teeth, the cause was grief over a sweetheart) is unchanged, the image is different. Decision 17, and it is the one place in this rendering where N2 and N6 pull against each other.

N7 check, in both directions, per R26 §5d

• Near burlesque, and what was taken instead. ¶13's "Little widow!" and "I propose to you!" in the same breath from a collar being ironed is comic in the Danish and stays comic; the burlesque option ("I make you an offer of my hand and my starch") was not taken. ¶19's "if only I had an earldom!" was nearly "if only I had a coronet to go with it", which is a joke the Danish does not make.
• Near flat, and what was taken instead — this is the direction R25 E8 does not have, and it fired more often. ¶2 "Well!" → "Well now!"; ¶9 "You look so mannish" → "You have such a mannish look!"; ¶12 "Prude!" → "The prude!"; ¶17 "no one could do that" → "No human being could copy you!"; ¶25 "I could do with becoming white paper" → "I need to become white paper!". Five sites, and all five are sites where the flat option is the one an unruled pass takes — which is the whole of the declared manipulation and is worth having in writing before anybody codes anything.

Doubts, recorded before any evaluation exists

• Whether "one step" is a quantity at all. N1 asks for one step and never two, and the log above shows the translator resolving that phrase case by case against his own ear. The regime's own limitations section says this; the log confirms it from the inside.
• ¶25's "a broken heart" is the choice most likely to be wrong, and if a reader called it domesticating rather than lightly raised, the translator would not have an argument. (internal-judgment-only.)
• N4 held numerically (1.146 against R25's 1.415) without ever being consciously applied. No decision above was taken because of N4. The rule may be describing what happens when E3 is simply absent rather than constraining anything.