Repository path: workshop/translations/kalila-ibn-urs/R36-v1/translation.md · rendered 2026-09-09

T-kalila-ibn-urs-R36-v1 — «باب الناسك وابن عرس» (Kalīla wa-Dimna), rendered whole under R36, the answering hand

Work. ʿAbd Allāh ibn al-Muqaffaʿ (d. c. 757), «كليلة ودمنة» — the chapter «باب الناسك وابن عرس», The Hermit and the Ichneumon: the tale of the faithful mongoose killed on a false inference, with the jar of ghee and honey — the "milkmaid's pail" — set inside it as the wife's counter-fable. The third chapter of this work the project has rendered whole, after «باب السائح والصائغ» (R30, sound-plain) and «باب الناسك والضيف» (R34, ornamentalist).

Copy-text. Arabic Wikisource, كليلة ودمنة/الناسك وابن عرس, retrieved 2026-08-16; ../source-wikisource-ibn-urs.txt, 558 words of body text in 19 sentences. Collated whole against the Amīriyya (Būlāq) 1937 printing, صفحة:كليلة ودمنة.pdf/242–246, in ../collation.md: eleven substantive adoptions, all from witness 2 — including a whole clause the copy-text drops. The adopted text is ../copy-text-adopted.txt and it, not the raw copy-text, is what was translated.

Comparator. Wyndham Knatchbull, Kalila and Dimna, or The Fables of Bidpai, Oxford 1819, The Religious Man and the Ichneumon — the only freely reachable published English of this chapter. Extracted programmatically by chapter boundary and not opened until this rendering and its log were committed. Contamination measured after the freeze: contamination: none — see §5.

Regime. R36 the answering hand (../../../regimes/R36-answering-hand.md): the source's sound figures enumerated and frozen first, each answered at the place it falls, and no sound device anywhere else. The disciplined counterpart of the R34 licence, and it exists to find out whether the line can be held.

Extent. The chapter whole: 570 Arabic words of adopted copy-text (558 in the raw copy-text, before the restored clause and the ten smaller adoptions) → 1,075 English words.

Why this chapter and this regime. T-kalila-nasik-R34-v1 §4 printed the project's first census of an ornamentalist's hand and it was not flattering: ten figures answered, twelve further sites invented, eight of them on nothing in the source at all. RS-20260816b then established that a source-blind reader cannot tell the two apart. Before asking whether that matters — the study limb's question — it is worth knowing whether a hand told to answer and not invent can actually do it. This rendering is that measurement, on a chapter of the same work, by the same hand, one day later.

1. The rendering

King Dabshalīm said to Baydabā the philosopher: I have heard this fable. Now set me a fable of the man who is hasty in his affair, without deliberation or regard for what comes after.

The philosopher said: He who is never firm in his affair is never free of regret, and his affair comes to what the hermit's came to, when he killed the ichneumon that had been loving to him.

The king said: And how was that?

The philosopher said: They say that there was a hermit of the hermits in the land of Jurjān, and he had a beautiful wife; and they remained a long while and were granted no child. Then she conceived by him after hope had gone, and the woman was gladdened and the hermit was gladdened by it; and he praised God, be He exalted, and asked of Him that what she carried might be male. And he said to his wife: Good news — for I hope it will be a boy, and there will be profit for us in him, and comfort of the eye. I shall choose him the choicest of names, and summon to him all the scholars there are.

The woman said: What moves you, man, to speak of what you do not know whether it will be or not? Whoever does that, there befalls him what befell the hermit who spilled the ghee and the honey over his own head.

He said to her: And how was that?

She said: They say that a hermit had, from the house of a merchant, a daily allowance of ghee and honey; and he would eat of it his food and his need, and put by what remained and set it in a jar, and hang the jar on a peg in a corner of the house, until it was full. And one day, as the hermit lay on his back with his staff in his hand and the jar hanging above his head, he fell to thinking about the dearness of ghee and honey, and he said: I shall sell what is in this jar for a dinar, and buy with it ten she-goats; and they will breed and they will bear a litter every five months, and it will not be long before they are a great flock, once their young bear young of their own. Then he reckoned it out at that rate over some years, and found it came to more than four hundred goats. And he said: I shall buy with them a hundred head of cattle, a bull or a cow for every four goats; and I shall buy land and seed, and I shall hire ploughmen, and I shall sow with the oxen, and I shall profit by the milk of the females and their increase; and five years will not pass but I shall have got great money from the sowing. Then I shall build a fine house, and I shall buy handmaids and slaves, and I shall marry a beautiful woman of good looks; and she will bear me a boy of birth and of breeding; and I shall choose for him the best of names; and when he grows up I shall school him, and school him well, and be hard on him in it; and if he takes it from me, good — and if not, I shall strike him with this staff. And he pointed with his hand at the jar, and broke it, and what was in it ran down over his face. And I have set you this fable only that you should not be hasty in the speaking of what should not be spoken, and of what you do not know whether it holds or does not hold. And the hermit took the lesson of what his wife had told.

Then the woman bore a beautiful boy, and his father rejoiced in him; and after some days the time came for her to purify herself.

So the woman said to the hermit: Sit by your son until I go to the bath-house, to bathe and be back. Then she set off for the bath-house, and left behind her boy and her spouse. It was not long before the king's messenger came to him, summoning him; and he found no one to leave with his son but an ichneumon he kept about the house, which he had reared from a little thing, so that it stood with him in the place of his own child. So the hermit left it beside the infant, and shut the house upon the two of them, and went off with the messenger.

And there came out of the stonework of the house a black snake, and it drew near the boy; and the ichneumon struck it, then leapt upon it and killed it, and cut it in pieces, and its mouth was filled with the blood of it.

Then the hermit came and opened the door, and the ichneumon met him like one bringing him good news of what it had done in the killing of the snake.

And when he saw it fouled with blood, and frightened, his wits flew from him, and he thought it had throttled his child; and he neither steadied himself in the affair nor stayed over it, until he should know the truth of the case and go by something other than what he supposed; but he was hasty with the ichneumon, and struck it with a staff that was in his hand upon the crown of its head, and it died.

And the hermit went in and saw the boy safe and alive, and beside him a black snake cut in pieces.

And when he knew the story, and the evil of what he had done in his haste became plain to him, he beat upon his head.

And he said: Would that I had never been granted this child, and never done this treachery! And his wife came in and found him in that state.

And she said to him: What ails you? And he told her the news — the good deed of the ichneumon, and the ill deed done to it.

And she said: This is the fruit of haste. And this is the likeness of the man who is not steady in his affair, but does what he is after in hurry and in haste.

2. Translator's log — decisions

Frozen with the rendering. Decisions and alternatives; no quality claims.

E1 What the regime forbids, and why the log is longer at the plain places than at the ornamented ones. Under R34 the interesting decisions were where to put ornament. Under R36 they are where to withhold it, because the hand's default reach is now the thing being measured. §4.3 lists every place a device appeared in draft and was taken out.

E2 The frame formulas are this hand's own, carried from the two neighbouring chapters. King Dabshalīm said to Baydabā the philosopher and And how was that? are the wordings of T-kalila-saih-R30-v1 and T-kalila-nasik-R34-v1. Deliberate self-consistency inside one work, and also lead–lead repetition, which method note (bhb) records this project measuring at up to 37 contiguous tokens. Declared here rather than left for a dependence check to find.

E3 ابن عرس → ichneumon. Literally son of a weasel; the animal is the Egyptian mongoose, Herpestes ichneumon, and the Arabic name is a set term, not a figure. Mongoose is the plainer word and is what a modern reader knows; ichneumon is what the English fable tradition of this story uses (Knatchbull's chapter title, unopened at the time of this decision, was known only by its programmatic extraction). The choice is for the register of the book, not for the sound, and it is the one lexical decision here that a reviser might simply reverse.

E4 جرجان → Jurjān. Left as a name, as الكرخ was in the neighbouring chapter. Gurgān, in north-eastern Iran; nothing naturalised, nothing footnoted.

E5 G4 — the chapter's one true saj', and English gets it at reduced strength. «أختار له أحسن الأسماء وأحضر له سائر الأدباء»: two exactly matched cola, both verbs on the أفعل pattern, rhyming at their ends on ‑āʾ (al-asmāʾ / al-udabāʾ). English has no rhyme pair that carries names and men of letters. What is put at the place instead is the matched frame kept exact — I shall choose him the choicest of names, and summon to him all the scholars there are — with polyptoton (choose / choicest) in the first colon and alliteration (summon / scholars) in the second. Recorded as answered at reduced strength, in the same terms as T-kalila-nasik-R34-v1 E11. Refused: I shall choose him the best of names and bring him all the men of letters.

E6 G14 — and here English gets a saj' in kind, by paying for it in word order. «ثم إنها انطلقت إلى الحمام، وخلفت زوجها والغلام»: rhyme on ‑ām at the ends of two adjacent cola. English has bath-house / spouse, which rhymes; but the Arabic names the husband first and the boy second, and the rhyme only lands if the boy comes first. The clause is inverted — left behind her boy and her spouse — and nothing else moves. This is the one place in the rendering where a sound figure changed the order of the source's own words, and it is the ordinary price of saj'.

E7 G16 — the jinās, and the one time this project has answered one at full strength. «حتى يعلم حقيقة الحال ويعمل بغير ما يظن»: يَعْلَم and يَعْمَل are the same three consonants with two of them exchanged — a jinás of metathesis, and the sharpest figure in the chapter, because it puts knowing and doing a hair apart in a sentence about a man who did without knowing. English has know / go: a true rhyme, in the matched verb positions, carrying the two senses exactly. Until he should know the truth of the case and go by something other than what he supposed. The neighbouring chapter's jinás (اختلط / تخلع) could only be answered by alliteration and was logged as reduced; this one is not reduced. Refused: until he should learn the truth of the matter and act on something other than his guess.

E8 G1 — the moral's frame. «من لم يكن في أمره متثبتاً لم يزل نادماً»: two matched negative cola, لم repeated in matched position. English repeats never in the same two positions and adds alliteration on firm / free: He who is never firm in his affair is never free of regret. Refused: a man who is not steady in what he does will not stop regretting it.

E9 G7 and G8 — the daydream, and why the English is more repetitive than an editor would like. The hermit's sum is two chains of first-person imperfects, five verbs and four, each verb in the same position in its colon. English can only answer a chain of matched morphology with a chain of matched syntax, so I shall is repeated at every member — nine times in five lines. A hand not under this regime would have coordinated bare verbs (and buy land and seed, hire ploughmen, sow with the oxen). The chains are where the regime costs the most words and the most patience, and they are the strongest single argument in this rendering for the position that answering a figure is not free.

E10 G10 — أدبته وأحسنت تأديبه → school him, and school him well. Root repetition across two cola answered by word repetition. The distinction from the cognate accusative — excluded from the inventory by rule (b) — is that أحسنت تأديبه is a second verb governing the noun, not the intensifying object of the first. The excluded case appears at [17]: «ولم أغدر هذا الغدر», rendered flatly as never done this treachery.

E11 G9 — سريّ نجيب → of birth and of breeding. Two adjectives on the same فَعِيل pattern in matched position: high-born, and of good stock. The English pair is matched in shape and alliterates. Refused: a noble and excellent boy. This locus was the last one admitted to the inventory and it is the one a reviser is most likely to strike from it.

E12 The clause the collation restored, and what it does to the joke. «بكل أربعة أعنز ثوراً أو بقرة» — a bull or a cow for every four goats — is absent from the copy-text and present in the Amīriyya printing (../collation.md §3.1 item 4). It is the exchange rate the whole daydream turns on, and without it the hermit jumps from four hundred goats to a hundred cattle with no arithmetic at all. The English carries it.

E13 من بعض أحجار البيت → out of the stonework of the house. The crux of ../collation.md §3.3: the copy-text reads stones, witness 2's reading is impossible, and the reading the sense wants (أجحار, burrows) is in neither witness this project can reach. Stonework is true under both live readings — a snake coming out from among the stones of a wall is coming out of a hole in it — and the emendation was declined rather than made from memory.

E14 أبشري → Good news. The root بشر returns at [13] in كالمبشر, like one bringing good news. The two are far apart and the recurrence is not a localisable figure, so it is not in the inventory; the English keeps the link anyway because it costs nothing. Refused: Rejoice.

E15 What is not in this rendering. No paragraph divisions exist in the copy-text; the divisions above are mine, cut at speech boundaries. The inner fable is given no section break. Witness 2's marginal notes and its closing catchwords are not rendered.

3. The Arabic's sound figures, enumerated before any English was written

The rule, imported verbatim from T-kalila-saih-R30-v1 §3 so that all three chapters of this work are counted the same way. A locus is admitted if two or more adjacent cola or members show either (i) saj' — rhyme at their ends — or (ii) muwāzana — the same morphological pattern, or the same word or root, repeated in matched position.

Exclusions, published so the enumeration can be tested. (a) Rhyme produced solely by an identical grammatical ending — the enclitic pronoun ‑هُ / ‑ها, or the indefinite accusative ‑اً — is not admitted. This excludes «قوته وحاجته» [9], the whole فضربها … فقتلها … فقطعها sequence [12], «سليماً حياً» [15] and «أرضاً وبذراً» [9]. (b) The cognate accusative (verb plus its own intensifying object) is a separate class and is not counted: «ولم أغدر هذا الغدر» [17], «فأخبرها بالخبر» [18]. (c) Repetition of the narrative formula قال … قال and of the book's frame word مثل is not a chosen figure. (d) A recurrence spread across the chapter — تثبّت at [3], [14] and [19], عجلة at [2], [14], [19] — is a figure of the whole and is not localisable, so it is excluded here.

Nineteen loci admitted, in copy-text order. Sentence numbers are ../copy-text-adopted.txt's.

Density: nineteen loci in 558 words, one per 29.4. «باب الناسك والضيف» ran at one per 28.7 and «باب السائح والصائغ» at one per 71 overall, one per 37 inside its frame discourse. This chapter, like the ichneumon's, is almost entirely frame discourse and admonition, and it lands within 0.7 words of the neighbouring chapter's rate. Three chapters counted by one rule now put the frame-discourse rate at 29–37 words per figure and the narrative rate far below it — which is a fact about this book's prose, not about this project's instrument.

4. The census — did the hand hold the line?

4.1 The nineteen answers

4.2 The three numbers this census exists to print

1. All nineteen figures are answered — eighteen in kind, one at reduced strength. The reduced one is G4, the chapter's one true saj', on a rhyme English cannot carry (E5). The other two hard cases both came out in kind: G14, the second saj', at the price of inverting the order of two nouns (E6), and G16, the jinás, by a rhyme that carries both senses exactly (E7). That is better than the neighbouring chapter managed on the same two classes, and the reason is luck rather than skill: English happened to have house / spouse and know / go, and happened not to have a pair for names / men of letters.

2. The hand put a sound device at a site the inventory does not name: zero times. That is the regime's failure measure and it came out at its floor. It did not come out at its floor by itself — §4.3 lists five devices that were in the draft and were taken out, and the honest statement is that the reflex is live and was caught, not that it is absent.

3. Answering all nineteen cost +35 words, 3.3% of the rendering. The comparison that matters: the ornamentalist next door answered ten figures and invented twelve further sites for +34 words, 6.1%. The two hands spent almost exactly the same number of words, and the restrained one spent all of them on the source's own figures while the ornamentalist spent half of them on its own — 19 answers at +1.84 words each against 22 sites at +1.55. On this pair of chapters ornament is not what costs length; ornament is what the length gets spent on, and the only question is where.

The concentration is worth naming: G7 and G8 alone are +14 of the +35, because a chain of matched morphology can only be answered in English by repeating a syntactic frame (E9). Take the two chains out and the other seventeen answers cost +21 words over 1,075, 2.0%.

4.3 The five devices that were removed — the regime's real content

Under R34 the log recorded what was put in. Under R36 the interesting record is what was taken out, because these are the places the hand reached for sound where the source has none. All five were caught in revision, not avoided.

The last of these is the one worth arguing about. «من غير … ولا …» is a matched negative frame in the Arabic; the inventory declines it because روية and نظر are not the same pattern and do not rhyme. A different enumeration would admit it, and then the removed device would have been a legitimate answer. This is exactly the dependence §Known limitations of R36 warns about: the inventory governs everything, and one locus's admission decides whether the same English words are an answer or an invention.

4.4 What this census is not

It is not blind and it is not evidence about a reader. The same hand enumerated the Arabic, wrote the English, and scored both columns; the enumeration was frozen before any English existed, which is what makes the at-a-figure scoring mechanical, and nothing makes the removal list complete. Whether any of these nineteen answers is audible to anyone — and whether the twelve inventions next door were audible — is the study limb's question and is not asserted here.

5. Contamination

contamination: none. Measured with tools/dependence_check.py after this file and its log were committed at f8647b7 and before the study limb was designed, against Knatchbull 1819's chapter X, The Monk and the Weasel — the only freely reachable published English of this chapter, extracted by chapter boundary from the archive.org scan and stored without being read. Figures in ../dependence.json, both reported per the standing rule:

3 shared 7-grams · 0 twelve-grams · 0 fifteen-grams · longest common run 9 tokens, on 1,217 Knatchbull tokens against 1,077 here. The measurement is a selection gate (CLAUDE.md), not a flag, and 9 tokens sits inside the range this project has measured against published hands (10 against Garnett, 11 against Fell, 24 against a 1923 anthology; note (bhb) puts the lead against itself at up to 37).

The longest run is the clause the collation restored, and it is the best evidence the collation was right: a bull or a cow for every four goats. Knatchbull, translating from de Sacy's Arabic in 1819, has the clause; the Wikisource copy-text does not; the Amīriyya printing does. Nine identical tokens here are convergence on a number, not dependence — there is one way to say an exchange rate in English — and the run's content is exactly what makes it uninformative about influence.

One exposure to declare, because the procedure is what makes the figure worth anything: the comparator was located in the scan by a case-insensitive grep for the animal's names, which printed about thirty single lines of Knatchbull's prose to the console. That happened after this rendering and its log were committed, so no English here can descend from it; but "not opened" is true of the chapter and not of those thirty lines, and the distinction is recorded rather than smoothed over.

6. Lineage and pointers

• Regime: ../../../regimes/R36-answering-hand.md
• Copy-text and collation: ../copy-text-adopted.txt, ../collation.md
• The ornamentalist rendering this is the counterpart of: ../../kalila-nasik/R34-v1/translation.md
• The sound-plain rendering both descend from: ../../kalila-saih/R30-v1/translation.md
• Arm: ../../../../wiki/arms/ARM-invented-ornament.md
• The finding that made the question worth asking: RS-20260816b-invented-figure §11