Quantitative data layer — seven-way comparison, Effingers Ch. 25 (“Frühling”)

Source: Gabriele Tergit, German → English. Seven renderings of the one chapter. Arms (column order everywhere): A | B | C | C-let (Cl) | D-let (Dl) | D-aim (Da) | H. A, B, C, Cl, Dl, Da are AI arms; H is the human published translation (Sophie Duvernoy, NYRB 2025), folded in as a co-equal column.

This is description and measurement, not a ranking. “Agreement” = two arms make the same concrete salient pick at a locus. H’s values are read from the finished text (observations, not inferences about intent). n = 1 chapter, one instance per AI arm — a case study, not a statistically powered result. The six AI arms share a common base model — a plausible source of their mutual closeness; flagged where it bears on the read-offs.

Coding basis: the German source + the seven finished translations only. The 45-locus partition set is the authoritative coding carried in notes/02_scoring.md (backed by per-locus renderings in notes/01_loci.md).

Coding correction logged (text sanity-check before computing)

One correction was made to the note-02 headline groups after re-reading the texts:

• L20 (gnädiges Fräulein, in the officer’s quoted line “es überzieht sich”). Note 02’s headline read {A,B,C,Cl}|{Dl}|{Da,H}, but its own parenthetical flagged the cell as needing a re-check. The texts show D-aim = “Please, my dear young lady, the sky is closing over” — i.e. Da renders “dear young lady”, grouping with A/B/C/Cl, not with H. H alone renders “Miss, the clouds are rolling in”; D-let alone renders “my dear Fräulein” (keeps the German title). Corrected partition: {A,B,C,Cl,Da} | {Dl} | {H}. This is the only change to the 45 partitions; all other note-02 groups were spot-checked against the texts (L13, L27, L42, L45 verified verbatim) and stand.

All numbers below are machine-computed from the partitions; coding calls are stated inline.

1. Locus table (45 loci) + partition-pattern tally

groups gives the partition in {…}|{…} notation: arms inside one brace made the same salient pick. Renderings are brief; full verbatim detail is in notes/01_loci.md.

A. Structure / form

B. Proper nouns / place names

C. Honorifics / titles / address

D. Loanwords / foreign vocab

E. Song / music titles & quoted lyric

F. Dialect / register (Berlin working-class)

G. Idioms

Partition-pattern tally (over 45 loci)

Group-size signatures (how the 7 arms divide at each locus):

Read: The single most common shape is one arm standing alone against the other six (21 of 45 loci, the [6,1] signature). Of those, the [6 AI]|{H} case — H alone against a unanimous AI bloc — accounts for 14 loci, the largest single partition pattern in the data and the structural backbone of the H-vs-the-field separation seen in §2. The remaining 31 loci are genuinely cross-cutting, and at two loci (L42 Hammelbeine, L13 Stadtwald) the field nearly shatters into singletons.

2. Pairwise agreement (7×7), computed from the partitions

Two arms “agree” at a locus when they sit in the same brace. Counts are out of 45; the matrix is symmetric; diagonal n/a.

Counts (out of 45)

Percentages (of 45)

Read-off

• AI–AI mean (15 unordered pairs): 28.40 / 45 = 63.1%.
• AI–H mean (6 pairs): 12.33 / 45 = 27.4%.
• The six AI arms agree with one another more than twice as often as any of them agrees with H (63.1% vs 27.4%). The AI arms cluster tightly and H sits apart from all of them.
• Tightest pair overall: B–C at 36/45 (80.0%) — and it is an AI–AI pair. The next tightest is A–B (35/45, 77.8%). Every one of the six AI–H pairs is looser than every one of the fifteen AI–AI pairs: the maximum AI–H agreement (C-let–H, 17/45) is below the minimum AI–AI agreement (D-let–D-aim, 22/45).
• Caveat (shared base model): the six AI arms run on a common base model, which is itself a plausible driver of their mutual closeness. The 63.1%/27.4% gap should be read with that in mind — it measures how alike the outputs are, not independent convergence.

Is any AI arm meaningfully closer to H than the others? Each AI arm’s agreement with H:

C-let is the AI arm closest to H (17/45). A is not notably closer to H than the field despite its persona being built from H’s own writing-about-Tergit (A–H = 12/45, mid-pack) — a point worth flagging against the expectation bias the methods note warns about.

The two controlled probes (the headline comparisons asked for):

• Is C-let closer to H than C is? Yes. C-let–H = 17/45 (37.8%) vs C–H = 14/45 (31.1%) — a +3-locus (+6.7 pp) move toward H from the domesticating permission alone (C and C-let are identical but for the permission paragraph). The extra C-let↔︎H agreements come at honorific/lexical loci where the permission nudged C-let onto H’s choice (e.g. L17 Herr Stadtrat → “my husband”, L22 Mätresse → “mistress”).
• Is D-aim closer to H than D-let is? No — they tie. D-aim–H = 10/45 (22.2%) and D-let–H = 10/45 (22.2%). The explicit school instruction (D-aim) did move D-aim’s overall domestication level well above D-let’s (see §3: composite 46.1 vs 17.1), but it did not move it onto more of H’s specific choices: D-aim domesticates toward different destinations than H picks (it anglicizes honorifics, which H keeps; see §6). So a stronger nudge raised the amount of domestication without raising agreement-with-H. This is a clean dissociation between “more domesticating” and “more like the human.”
• Note the probe-internal agreements themselves: C↔︎C-let = 31/45 (68.9%) but D-let↔︎D-aim = only 22/45 (48.9%) — the school instruction perturbed D far more than the permission perturbed C, consistent with D-aim being the single most-moved arm in the study.

3. Foreignization ↔︎ domestication composite (0–100)

0 = keeps the German / source-preserving; 100 = naturalizes into English. Six tests; each test’s per-arm score is the domesticating-share of the items in its basis. Continuity with the prior five tests (T1 proper nouns, T2 loanwords/titles, T3 dialect, T4 refrain, T5 idioms) is kept; T6 honorifics is added because the seven-way field made it discriminating (Da anglicizes Fräulein/Herr; Dl keeps Herr Kollege/Stadtrat). Composite = coverage-weighted mean of the six test scores (each measure weighted by the number of loci it covers).

Coding calls (stated, per the spec): items scored 0 / 0.5 / 1 for source-keep / partial / full domestication. Partial (0.5) is used where a choice naturalizes but stops short — e.g. “the Linden” (half-keeps the street-name) vs “the linden trees” (fully translates); “casino gown” (drops French toilette but keeps “casino”) vs “evening gown” (fully domesticated). Two non-discriminating items were dropped from a test basis and noted: L26 Schlafrock (all seven land on plain-English “dressing gown”/“wrapper”; no German-keep option to score) is excluded from T2. Two cross-direction loci are handled honestly (see note after the table): H foreignizes at a few lexical points (keeps French “toilette” L27; “Rhenish” is period-English, scored as more-domesticating).

Per-test domesticating-share (0–100) and composite

Gradient (least → most domesticating)

D-let (17.1) < B (19.7) < A (27.6) < C (29.0) < C-let (30.3) < D-aim (46.1) ≪ H (80.3)

• H stands far out at the domesticating end (80.3) — well clear of the field, ~1.7× the next-most-domesticating arm (D-aim, 46.1) and about four-and-a-half times the most source-preserving arm (D-let, 17.1). H scores 100 on three of the six tests (dialect, refrain/structure, idioms): it is the only arm that uses eye-dialect, fuses/reshapes the refrain and paragraphs, and dissolves the German idioms into English. The six AI arms occupy a comparatively narrow band, 17.1–46.1 — they domesticate far less, and far less variably, than H.
• The two controlled probes are both visible in the composite:
  • C → C-let: +1.3 (29.0 → 30.3). The domesticating permission alone nudges the no-persona control up, mostly via honorifics (T6 50.0 → 58.3) and idioms (T5 25.0 → 50.0). A small but directionally clean license effect.
  • D-let → D-aim: +29.0 (17.1 → 46.1) — by far the largest probe effect in the study. The explicit school instruction roughly triples the persona arm’s domestication, with the biggest jumps in honorifics (T6 0.0 → 83.3), proper nouns (T1 10.0 → 50.0), and idioms (T5 50.0). The categorical school brief moves the dial much harder than the permission does.
• Cross-direction honesty: H is not monotonically domesticating. On the lexical sub-current it occasionally foreignizes — it keeps French “toilette” (L27, the only arm to do so), substitutes French “D’accord” for German “Einverstanden” (a watch-locus outside the 45), and keeps literal “breakfast” at 1 p.m. where the AI render the sense “lunch.” Its T2 (loanwords) score of 62.5 is high but not maximal precisely because of these keeps. So H’s composite is dominated by hard domestication of structure, dialect, and idiom, partly offset by a foreignizing taste in lexis. The composite reports the net; the sub-axis table (§6) separates the dials.
• The AI ordering is itself informative: the two arms that most resist domestication are B (self-built German-corpus persona) and D-let (expanded-corpus persona, permission only) — persona arms left on the lighter brief sit lowest. The no-persona controls (C, C-let) and the H-derived persona (A) sit mid-band. Only the school-instructed D-aim climbs clear of the AI pack.

4. Clusters + 2-D coordinates

Clusters are read jointly from the pairwise matrix (§2) and the composite (§3). Membership is a coder judgment grounded in those numbers.

Named clusters

Notes. Cluster 1 is the gravitational center: five of six AI arms, all sitting in a tight foreignizing-to-mild band. C-let is the cluster-1 member that leans furthest toward H (composite 30.3, H-agreement 37.8%) — the permission pulls it toward the boundary without leaving the core. D-aim breaks out as a one-arm cluster: the school instruction pushes it past the AI band on the X-axis, yet it remains far short of H and (per §6) overshoots H on the single honorifics dial. H is its own cluster on both axes — there is no AI arm near it.

2-D coordinates (x = foreignization↔︎domestication, y = editorial boldness)

X = the §3 composite (0 source-preserving → 100 naturalized). Y = editorial boldness (0 = preserves the German’s form/structure/paragraphing → 100 = freely reshapes into a new English artifact). The Y-axis is scored over 10 form-level reshaping items, independent of lexical domestication: refrain reshaped (L01), refrain fused (L02), paragraphs fused (L03), march-line reshaped (L33), invented rhyme (L30), section dividers (L04), eye-dialect (L35), frisierte Schnauze reshaped (L40), idiom dissolved (L43), Spreewald/locale dropped (L14).

Read: Six AI arms cluster in the lower-left quadrant (x ≤ 47, y ≤ 15) — they neither domesticate hard nor reshape form. H sits alone in the upper-right (80, 90). The two axes are correlated for H but separable for the AI: D-aim moves substantially right on X (lexical/honorific domestication, 46.1) while barely rising on Y (15.0) — it domesticates vocabulary and titles without touching the German’s form. Editorial boldness is essentially an H-only behaviour in this chapter: only H reshapes the refrain, fuses paragraphs, invents the rhyme, and writes eye-dialect. D-let/D-aim’s small Y (10–15) comes almost entirely from the cosmetic “—” section dividers, not from fusion.

5. Destination dispersion (the region-vs-point measure)

Question: where the more-domesticating arms (C-let, D-let, D-aim, H) all move off the literal/source-preserving choice, do they land on the same English destination, or scatter? For each candidate locus the table gives the literal anchor, the distinct-destination count among the domesticating-end movers (those of {C-let, D-let, D-aim, H} that moved), and the count among all seven arms that moved.

Region loci (the strict test: all four domesticating-end arms move off the literal)

One-number summary: across the 3 region loci (all four domesticating-end arms move off the literal), the mean number of distinct destinations among those movers is 3.33. At 3 of 3 region loci the movers share the direction (they all leave the literal) but split on the destination (>1 distinct landing). Across all 9 flagged movement loci, the mean distinct-destination count among the domesticating-end movers is 2.56 (and 3.11 counting all seven arms that moved).

Reading (whichever way it falls): the data support “domesticated Tergit is a region, not a point.” The task pre-flagged Hammelbeine and Bärenhunger; both are maximally dispersed — at each, all four domesticating-end arms pick a different English destination (4 distinct of 4 movers). The third strict region locus, L39 doofe Ziege, is far tighter: the four converge on “cow” (only a stupid/silly modifier varies), so it is best read as a near-point (1 image, 2 surface forms) and is the honest counter-example to dispersal. Two further loci (L13 Stadtwald, L27 Kasinotoilette) show heavy scatter among the AI (5 and 4 distinct destinations across all movers respectively) but fail the strict “all four dom-end arms move” gate because one arm keeps the German. Net: when the domesticating arms abandon a vivid German idiom (the two pre-flagged cases), they reliably scatter; when they abandon a flat insult (Ziege), they converge.

6. Position relative to H, by sub-axis

Domestication is not one dial. For six sub-axes, each arm’s domesticating-mean (same 0/0.5/1 item codings as §3) is compared to H’s mean on that sub-axis: MORE = more domesticating than H, at = equal to H (within rounding), less = less domesticating than H.

Sub-axis bases: lexicon {L21,L23,L24,L25,L27,L28}; honorifics {L15,L16,L17,L18,L20,L19}; dialect {L35,L36,L37,L38,L39,L40}; structure/paragraphing {L01,L02,L03,L04}; embedded-song form {L29,L30,L31,L32,L33}; locale-naming {L07,L11,L13,L14}.

Arm × sub-axis (cell = vs H)

(H’s per-sub-axis domesticating-means, for reference: lexicon 0.83, honorifics 0.67, dialect 1.00, structure/para 0.75, embedded-song 0.60, locale-naming 0.88.)

One sentence per sub-axis

• Lexicon: every AI arm is less domesticating than H, but the gap is small — H keeps some loans (toilette) yet domesticates Mokka/Coupé/Rheinwein, landing at 0.83; A is closest (0.75, it too renders “carriage” and “stool”).
• Honorifics: the one sub-axis where an AI arm sits past H — D-aim (0.83 > H’s 0.67), because Da is the only arm to anglicize both Fräulein→Miss and Herr→Mr, moves H itself keeps; H domesticates honorifics only partially (drops “Stadtrat” to “my husband”, keeps Fräulein/Herr), so several arms approach but only D-aim overshoots.
• Dialect: H is at the ceiling (1.00 — eye-dialect + vulgarity + idiomatic insults); all six AI arms sit well less domesticating (0.00–0.33), the widest and most consistent gap in the table.
• Structure/paragraphing: all six AI arms less than H (H 0.75 from fusing/reshaping the refrain and paragraphs); D-let/D-aim’s small nonzero comes only from the cosmetic dividers.
• Embedded-song form: A, B, C, and D-aim sit at H (all four translate the song titles and lyric the way H mostly does, 0.60); C-let and D-let fall less because they keep more German titles (Frauenliebe, Feuerzauber, Lotosblume). Note this is the sub-axis where H is least extreme (it keeps two German titles itself), so parity is easy to reach.
• Locale-naming: all six AI less than H (H 0.88 — “linden trees”, “park”, drops “Spreewald”); D-aim and C-let are closest (0.50) — D-aim via the “East End” substitution, C-let via “city park”.

Headline for the strongest-nudged arm (D-aim): D-aim sits past H on exactly one sub-axis (honorifics), at H on one (embedded-song), and short of H on the other four (lexicon, dialect, structure, locale). This is the quantitative form of “the most-nudged arm overshoots the human on one narrow dial while remaining well short of it on the rest” — domestication is not a single slider, and even the hardest-pushed AI arm does not become H; it becomes differently domesticating. The two permission-only persona/control arms (C-let, D-let) are less domesticating than H on every one of the six sub-axes.

Mirroring JSON (chartable)

Single JSON block mirroring all six sections: loci (with groups), partition_tally, pairwise (counts + pct + read-offs), foreignization (tests + composite), clusters + coords + editorial_boldness, dispersion, relative_to_H. Arm key order: A, B, C, Cl (=C-let), Dl (=D-let), Da (=D-aim), H.

{
  "meta": {
    "chapter": "Effingers Ch.25 (Frühling)",
    "arms": [
      "A",
      "B",
      "C",
      "Cl",
      "Dl",
      "Da",
      "H"
    ],
    "n_loci": 45,
    "x_axis": "foreignization(0)->domestication(100)",
    "y_axis": "editorial_boldness(0)->free_reshape(100)",
    "note": "description not ranking; n=1; six AI arms share a base model; L20 corrected from note 02 after text check"
  },
  "loci": [
    {
      "id": "L01",
      "cue": "refrain wording",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L02",
      "cue": "refrain standalone-para",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L03",
      "cue": "paragraph integrity (Sofie/guard stretch)",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L04",
      "cue": "section dividers",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "H"
        ],
        [
          "Dl",
          "Da"
        ]
      ]
    },
    {
      "id": "L05",
      "cue": "Tiergartenstraße ß",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L06",
      "cue": "Chausseestraße ß",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L07",
      "cue": "die Linden (1st mention)",
      "groups": [
        [
          "A",
          "B",
          "Cl",
          "Dl"
        ],
        [
          "C",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L08",
      "cue": "Schloss",
      "groups": [
        [
          "A",
          "B"
        ],
        [
          "C",
          "Da"
        ],
        [
          "Cl",
          "Dl"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L09",
      "cue": "Graues Kloster",
      "groups": [
        [
          "A",
          "Dl"
        ],
        [
          "B",
          "C",
          "Cl",
          "Da",
          "H"
        ]
      ]
    },
    {
      "id": "L10",
      "cue": "Weidendammer Brücke",
      "groups": [
        [
          "A",
          "Cl",
          "Dl"
        ],
        [
          "B",
          "C",
          "Da",
          "H"
        ]
      ]
    },
    {
      "id": "L11",
      "cue": "im Berliner Osten",
      "groups": [
        [
          "C",
          "Cl",
          "Dl",
          "H"
        ],
        [
          "A",
          "B"
        ],
        [
          "Da"
        ]
      ]
    },
    {
      "id": "L12",
      "cue": "Hotel Barblan",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "H"
        ],
        [
          "Da"
        ]
      ]
    },
    {
      "id": "L13",
      "cue": "Stadtwald",
      "groups": [
        [
          "B",
          "C"
        ],
        [
          "A"
        ],
        [
          "Cl"
        ],
        [
          "Dl"
        ],
        [
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L14",
      "cue": "Spreewälderin",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl"
        ],
        [
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L15",
      "cue": "Fräulein +name",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "H"
        ],
        [
          "Da"
        ]
      ]
    },
    {
      "id": "L16",
      "cue": "Herr +name",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "H"
        ],
        [
          "Da"
        ]
      ]
    },
    {
      "id": "L17",
      "cue": "Herr Stadtrat",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Da"
        ],
        [
          "Cl",
          "H"
        ],
        [
          "Dl"
        ]
      ]
    },
    {
      "id": "L18",
      "cue": "Herr Kollege",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Da",
          "H"
        ],
        [
          "Dl"
        ]
      ]
    },
    {
      "id": "L19",
      "cue": "reverential capital Du",
      "groups": [
        [
          "A",
          "C",
          "Cl",
          "Da",
          "H"
        ],
        [
          "B",
          "Dl"
        ]
      ]
    },
    {
      "id": "L20",
      "cue": "gnädiges Fräulein (quote)",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Da"
        ],
        [
          "Dl"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L21",
      "cue": "Coupé",
      "groups": [
        [
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "A",
          "H"
        ]
      ]
    },
    {
      "id": "L22",
      "cue": "Mätresse",
      "groups": [
        [
          "Cl",
          "Dl",
          "Da",
          "H"
        ],
        [
          "A",
          "B",
          "C"
        ]
      ]
    },
    {
      "id": "L23",
      "cue": "Mokka",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L24",
      "cue": "Molle",
      "groups": [
        [
          "A",
          "B",
          "Dl",
          "Da",
          "H"
        ],
        [
          "C",
          "Cl"
        ]
      ]
    },
    {
      "id": "L25",
      "cue": "Taburett",
      "groups": [
        [
          "A",
          "C",
          "Da",
          "H"
        ],
        [
          "B",
          "Cl",
          "Dl"
        ]
      ]
    },
    {
      "id": "L26",
      "cue": "Schlafrock/Morgenrock",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "H"
        ],
        [
          "Da"
        ]
      ]
    },
    {
      "id": "L27",
      "cue": "Kasinotoilette",
      "groups": [
        [
          "B",
          "C",
          "Cl"
        ],
        [
          "A"
        ],
        [
          "Dl"
        ],
        [
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L28",
      "cue": "Rheinwein",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L29",
      "cue": "Frauenlieb und -leben",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Da"
        ],
        [
          "Cl",
          "Dl",
          "H"
        ]
      ]
    },
    {
      "id": "L30",
      "cue": "Schumann lyric rhyme?",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L31",
      "cue": "Feuerzauber",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Da",
          "Dl"
        ],
        [
          "Cl",
          "H"
        ]
      ]
    },
    {
      "id": "L32",
      "cue": "Die Lotosblume ängstigt",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Da",
          "H"
        ],
        [
          "Dl"
        ]
      ]
    },
    {
      "id": "L33",
      "cue": "Querpfeifer/Trommler march line",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L34",
      "cue": "Feuerzauber composer attribution",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L35",
      "cue": "eye-dialect overall",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L36",
      "cue": "nonstandard dialect grammar",
      "groups": [
        [
          "A",
          "B",
          "Dl",
          "Da"
        ],
        [
          "C",
          "Cl",
          "H"
        ]
      ]
    },
    {
      "id": "L37",
      "cue": "versaufen/Wochenlohn register",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L38",
      "cue": "Paule address",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl"
        ],
        [
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L39",
      "cue": "doofe Ziege",
      "groups": [
        [
          "Cl",
          "Dl",
          "Da",
          "H"
        ],
        [
          "B",
          "C"
        ],
        [
          "A"
        ]
      ]
    },
    {
      "id": "L40",
      "cue": "frisierte Schnauze",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L41",
      "cue": "Schlafbursche",
      "groups": [
        [
          "A",
          "B",
          "Cl",
          "Dl",
          "Da",
          "H"
        ],
        [
          "C"
        ]
      ]
    },
    {
      "id": "L42",
      "cue": "Hammelbeine langziehen",
      "groups": [
        [
          "B",
          "C"
        ],
        [
          "A"
        ],
        [
          "Cl"
        ],
        [
          "Dl"
        ],
        [
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L43",
      "cue": "an meiner Wiege gesungen",
      "groups": [
        [
          "A",
          "B",
          "C",
          "Cl",
          "Dl",
          "Da"
        ],
        [
          "H"
        ]
      ]
    },
    {
      "id": "L44",
      "cue": "ein Lied davon singen",
      "groups": [
        [
          "A",
          "Cl",
          "Dl"
        ],
        [
          "B",
          "Da"
        ],
        [
          "C",
          "H"
        ]
      ]
    },
    {
      "id": "L45",
      "cue": "Bärenhunger",
      "groups": [
        [
          "A",
          "B",
          "C"
        ],
        [
          "Cl"
        ],
        [
          "Dl"
        ],
        [
          "Da"
        ],
        [
          "H"
        ]
      ]
    }
  ],
  "partition_tally": {
    "six_AI_vs_H_alone": 14,
    "cross_cutting": 31,
    "H_singleton_any": 23,
    "two_group_loci": 32,
    "three_plus_group_loci": 13,
    "size_signatures": {
      "[6,1]": 21,
      "[5,2]": 6,
      "[4,3]": 5,
      "[4,2,1]": 4,
      "[5,1,1]": 3,
      "[2,1,1,1,1,1]": 2,
      "[3,1,1,1,1]": 2,
      "[2,2,2,1]": 1,
      "[3,2,2]": 1
    }
  },
  "pairwise": {
    "n": 45,
    "counts": {
      "A": {
        "A": null,
        "B": 35,
        "C": 31,
        "Cl": 29,
        "Dl": 28,
        "Da": 25,
        "H": 12
      },
      "B": {
        "A": 35,
        "B": null,
        "C": 36,
        "Cl": 30,
        "Dl": 28,
        "Da": 27,
        "H": 11
      },
      "C": {
        "A": 31,
        "B": 36,
        "C": null,
        "Cl": 31,
        "Dl": 23,
        "Da": 27,
        "H": 14
      },
      "Cl": {
        "A": 29,
        "B": 30,
        "C": 31,
        "Cl": null,
        "Dl": 31,
        "Da": 23,
        "H": 17
      },
      "Dl": {
        "A": 28,
        "B": 28,
        "C": 23,
        "Cl": 31,
        "Dl": null,
        "Da": 22,
        "H": 10
      },
      "Da": {
        "A": 25,
        "B": 27,
        "C": 27,
        "Cl": 23,
        "Dl": 22,
        "Da": null,
        "H": 10
      },
      "H": {
        "A": 12,
        "B": 11,
        "C": 14,
        "Cl": 17,
        "Dl": 10,
        "Da": 10,
        "H": null
      }
    },
    "pct": {
      "A": {
        "A": null,
        "B": 77.8,
        "C": 68.9,
        "Cl": 64.4,
        "Dl": 62.2,
        "Da": 55.6,
        "H": 26.7
      },
      "B": {
        "A": 77.8,
        "B": null,
        "C": 80.0,
        "Cl": 66.7,
        "Dl": 62.2,
        "Da": 60.0,
        "H": 24.4
      },
      "C": {
        "A": 68.9,
        "B": 80.0,
        "C": null,
        "Cl": 68.9,
        "Dl": 51.1,
        "Da": 60.0,
        "H": 31.1
      },
      "Cl": {
        "A": 64.4,
        "B": 66.7,
        "C": 68.9,
        "Cl": null,
        "Dl": 68.9,
        "Da": 51.1,
        "H": 37.8
      },
      "Dl": {
        "A": 62.2,
        "B": 62.2,
        "C": 51.1,
        "Cl": 68.9,
        "Dl": null,
        "Da": 48.9,
        "H": 22.2
      },
      "Da": {
        "A": 55.6,
        "B": 60.0,
        "C": 60.0,
        "Cl": 51.1,
        "Dl": 48.9,
        "Da": null,
        "H": 22.2
      },
      "H": {
        "A": 26.7,
        "B": 24.4,
        "C": 31.1,
        "Cl": 37.8,
        "Dl": 22.2,
        "Da": 22.2,
        "H": null
      }
    },
    "ai_ai_mean_count": 28.4,
    "ai_ai_mean_pct": 63.1,
    "ai_H_mean_count": 12.33,
    "ai_H_mean_pct": 27.4,
    "tightest_pair": [
      "B",
      "C"
    ],
    "tightest_pair_count": 36,
    "ai_vs_H": {
      "A": 12,
      "B": 11,
      "C": 14,
      "Cl": 17,
      "Dl": 10,
      "Da": 10
    },
    "C_let_closer_to_H_than_C": true,
    "C_H": 14,
    "Cl_H": 17,
    "D_aim_closer_to_H_than_D_let": false,
    "Dl_H": 10,
    "Da_H": 10,
    "probe_C_vs_Clet_agreement": 31,
    "probe_Dlet_vs_Daim_agreement": 22
  },
  "foreignization": {
    "scale": "0=source-preserving,100=naturalized",
    "tests": {
      "T1_propernouns": {
        "items": [
          "L05_ss",
          "L06_ss",
          "L07_linden",
          "L08_schloss",
          "L09_gk",
          "L10_wdb",
          "L11_ost",
          "L12_barblan",
          "L13_stadtwald",
          "L14_spreewald"
        ],
        "scores": {
          "A": 10.0,
          "B": 30.0,
          "C": 40.0,
          "Cl": 30.0,
          "Dl": 10.0,
          "Da": 50.0,
          "H": 75.0
        }
      },
      "T2_loanwords": {
        "items": [
          "L21_coupe",
          "L23_mokka",
          "L24_molle",
          "L25_taburett",
          "L27_kasino",
          "L28_rheinwein",
          "L_cancan",
          "L_ivm"
        ],
        "scores": {
          "A": 56.2,
          "B": 25.0,
          "C": 25.0,
          "Cl": 12.5,
          "Dl": 31.2,
          "Da": 43.8,
          "H": 62.5
        }
      },
      "T3_dialect": {
        "items": [
          "L35_eyedialect",
          "L36_nonstdgram",
          "L37_versaufen",
          "L38_paule",
          "L39_ziege",
          "L40_schnauze"
        ],
        "scores": {
          "A": 16.7,
          "B": 0.0,
          "C": 16.7,
          "Cl": 33.3,
          "Dl": 16.7,
          "Da": 33.3,
          "H": 100.0
        }
      },
      "T4_refrain_structure": {
        "items": [
          "L01_refrainword",
          "L02_standalone",
          "L03_paraint",
          "L33_querpfeifer"
        ],
        "scores": {
          "A": 0.0,
          "B": 0.0,
          "C": 0.0,
          "Cl": 0.0,
          "Dl": 0.0,
          "Da": 0.0,
          "H": 100.0
        }
      },
      "T5_idioms": {
        "items": [
          "L42_hammel",
          "L43_wiege",
          "L44_lied",
          "L45_baeren"
        ],
        "scores": {
          "A": 25.0,
          "B": 12.5,
          "C": 25.0,
          "Cl": 50.0,
          "Dl": 50.0,
          "Da": 50.0,
          "H": 100.0
        }
      },
      "T6_honorifics": {
        "items": [
          "L15_fraulein",
          "L16_herr",
          "L17_stadtrat",
          "L18_kollege",
          "L20_gnf",
          "L19_du"
        ],
        "scores": {
          "A": 50.0,
          "B": 33.3,
          "C": 50.0,
          "Cl": 58.3,
          "Dl": 0.0,
          "Da": 83.3,
          "H": 66.7
        }
      }
    },
    "composite": {
      "A": 27.6,
      "B": 19.7,
      "C": 29.0,
      "Cl": 30.3,
      "Dl": 17.1,
      "Da": 46.1,
      "H": 80.3
    },
    "composite_weighting": "coverage-weighted mean (each measure weighted by the number of loci it covers)",
    "gradient_low_to_high": [
      "Dl",
      "B",
      "A",
      "C",
      "Cl",
      "Da",
      "H"
    ],
    "dropped_nondiscriminating_items": {
      "T2_loanwords": [
        "L26_schlafrock"
      ]
    }
  },
  "clusters": [
    {
      "name": "Source-preserving AI core",
      "members": [
        "A",
        "B",
        "C",
        "Cl",
        "Dl"
      ],
      "basis": "mutual agreement 62-80%; composite 17.1-30.3",
      "evidence": [
        "L05",
        "L06",
        "L21",
        "L01",
        "L02",
        "L03",
        "L35",
        "L42",
        "L45"
      ]
    },
    {
      "name": "Loanword-keepers sub-pocket",
      "members": [
        "C",
        "Cl"
      ],
      "basis": "C-Cl 68.9%",
      "evidence": [
        "L24",
        "L29",
        "L36"
      ]
    },
    {
      "name": "German-title holdouts",
      "members": [
        "Dl"
      ],
      "basis": "Dl-H 22.2%; composite 17.1",
      "evidence": [
        "L17",
        "L18",
        "L32",
        "L09",
        "L10"
      ]
    },
    {
      "name": "School-nudged outlier (AI)",
      "members": [
        "Da"
      ],
      "basis": "composite 46.1; Dl-Da 48.9%",
      "evidence": [
        "L15",
        "L16",
        "L11"
      ]
    },
    {
      "name": "Human anchor",
      "members": [
        "H"
      ],
      "basis": "AI-H mean 27.4%; composite 80.3; boldness 90",
      "evidence": [
        "L35",
        "L37",
        "L01",
        "L30",
        "L43",
        "L14"
      ]
    }
  ],
  "coords": {
    "A": {
      "x": 27.6,
      "y": 0.0
    },
    "B": {
      "x": 19.7,
      "y": 0.0
    },
    "C": {
      "x": 29.0,
      "y": 0.0
    },
    "Cl": {
      "x": 30.3,
      "y": 0.0
    },
    "Dl": {
      "x": 17.1,
      "y": 10.0
    },
    "Da": {
      "x": 46.1,
      "y": 15.0
    },
    "H": {
      "x": 80.3,
      "y": 90.0
    }
  },
  "editorial_boldness": {
    "items": [
      "L01_refrain_reshape",
      "L02_refrain_fused",
      "L03_para_fused",
      "L33_querpfeifer",
      "L30_invented_rhyme",
      "L04_dividers",
      "L35_eyedialect",
      "L40_schnauze_reshape",
      "L43_idiom_dissolve",
      "L14_spreewald_drop"
    ],
    "scores": {
      "A": 0.0,
      "B": 0.0,
      "C": 0.0,
      "Cl": 0.0,
      "Dl": 10.0,
      "Da": 15.0,
      "H": 90.0
    }
  },
  "dispersion": {
    "loci": {
      "L42_Hammelbeine": {
        "literal": "mutton-legs (calque keeps the German image)",
        "dest": {
          "A": "box your ears",
          "B": "mutton-legs(calque)",
          "C": "mutton-legs(calque)",
          "Cl": "give you what-for",
          "Dl": "give you a hiding",
          "Da": "give you a thrashing",
          "H": "give you a good slap"
        },
        "off": {
          "A": true,
          "B": false,
          "C": false,
          "Cl": true,
          "Dl": true,
          "Da": true,
          "H": true
        }
      },
      "L45_Baerenhunger": {
        "literal": "bear-image ('hungry as a bear')",
        "dest": {
          "A": "hungry as a bear",
          "B": "hungry as a bear",
          "C": "hungry as a bear",
          "Cl": "ravenous",
          "Dl": "starving",
          "Da": "wolf's hunger",
          "H": "famished"
        },
        "off": {
          "A": false,
          "B": false,
          "C": false,
          "Cl": true,
          "Dl": true,
          "Da": true,
          "H": true
        }
      },
      "L13_Stadtwald": {
        "literal": "Stadtwald (keep German proper noun)",
        "dest": {
          "A": "the woods",
          "B": "the town woods",
          "C": "the town woods",
          "Cl": "the city park",
          "Dl": "the Stadtwald",
          "Da": "the woods outside the town",
          "H": "the park"
        },
        "off": {
          "A": true,
          "B": true,
          "C": true,
          "Cl": true,
          "Dl": false,
          "Da": true,
          "H": true
        }
      },
      "L27_Kasinotoilette": {
        "literal": "casino toilette (keep the French 'toilette')",
        "dest": {
          "A": "dinner gown",
          "B": "casino gown",
          "C": "casino gown",
          "Cl": "casino gown",
          "Dl": "ball-gown",
          "Da": "evening gown",
          "H": "casino toilette"
        },
        "off": {
          "A": true,
          "B": true,
          "C": true,
          "Cl": true,
          "Dl": true,
          "Da": true,
          "H": false
        }
      },
      "L17_HerrStadtrat": {
        "literal": "Stadtrat (keep German title)",
        "dest": {
          "A": "the Councillor",
          "B": "the Councillor",
          "C": "the Councillor",
          "Cl": "my husband",
          "Dl": "the Stadtrat",
          "Da": "the Councillor",
          "H": "my husband"
        },
        "off": {
          "A": true,
          "B": true,
          "C": true,
          "Cl": true,
          "Dl": false,
          "Da": true,
          "H": true
        }
      },
      "L44_LiedDavon": {
        "literal": "sing-a-song calque (keep the German image)",
        "dest": {
          "A": "sing you a song",
          "B": "tell a tale",
          "C": "tell you a thing or two",
          "Cl": "sing you a song",
          "Dl": "sing you a song",
          "Da": "tell a tale",
          "H": "tell you a thing or two"
        },
        "off": {
          "A": false,
          "B": true,
          "C": true,
          "Cl": false,
          "Dl": false,
          "Da": true,
          "H": true
        }
      },
      "L07_Linden": {
        "literal": "Unter den Linden / 'the Linden' (keep the street-name)",
        "dest": {
          "A": "the Linden",
          "B": "the Linden",
          "C": "Unter den Linden",
          "Cl": "the Linden",
          "Dl": "the Linden",
          "Da": "Unter den Linden",
          "H": "the linden trees"
        },
        "off": {
          "A": false,
          "B": false,
          "C": false,
          "Cl": false,
          "Dl": false,
          "Da": false,
          "H": true
        }
      },
      "L38_Paule": {
        "literal": "Paule (keep the German diminutive)",
        "dest": {
          "A": "Paule",
          "B": "Paule",
          "C": "Paule",
          "Cl": "Paule",
          "Dl": "Paule",
          "Da": "Paul, love",
          "H": "Paulie"
        },
        "off": {
          "A": false,
          "B": false,
          "C": false,
          "Cl": false,
          "Dl": false,
          "Da": true,
          "H": true
        }
      },
      "L39_doofeZiege": {
        "literal": "goat (literal Ziege)",
        "dest": {
          "A": "daft goose",
          "B": "daft goat",
          "C": "daft goat",
          "Cl": "stupid cow",
          "Dl": "stupid cow",
          "Da": "silly cow",
          "H": "stupid cow"
        },
        "off": {
          "A": true,
          "B": false,
          "C": false,
          "Cl": true,
          "Dl": true,
          "Da": true,
          "H": true
        }
      }
    },
    "region_loci": [
      {
        "id": "L42_Hammelbeine",
        "distinct_destinations_dom_movers": 4
      },
      {
        "id": "L45_Baerenhunger",
        "distinct_destinations_dom_movers": 4
      },
      {
        "id": "L39_doofeZiege",
        "distinct_destinations_dom_movers": 2
      }
    ],
    "mean_distinct_destinations_region_loci": 3.33,
    "mean_distinct_destinations_all_flagged_dom_movers": 2.56,
    "region_loci_with_direction_shared_destination_split": "3 of 3",
    "note": "Hammelbeine(4) & Baerenhunger(4) maximally dispersed; doofe Ziege(2) near-point (one image 'cow')"
  },
  "relative_to_H": {
    "sub_axes": [
      "lexicon",
      "honorifics",
      "dialect",
      "structure_paragraphing",
      "embedded_song_form",
      "locale_naming"
    ],
    "items": {
      "lexicon": [
        "L21_coupe",
        "L23_mokka",
        "L24_molle",
        "L25_taburett",
        "L27_kasino",
        "L28_rheinwein"
      ],
      "honorifics": [
        "L15_fraulein",
        "L16_herr",
        "L17_stadtrat",
        "L18_kollege",
        "L20_gnf",
        "L19_du"
      ],
      "dialect": [
        "L35_eyedialect",
        "L36_nonstdgram",
        "L37_versaufen",
        "L38_paule",
        "L39_ziege",
        "L40_schnauze"
      ],
      "structure_paragraphing": [
        "L01_refrainword",
        "L02_standalone",
        "L03_paraint",
        "L04_dividers"
      ],
      "embedded_song_form": [
        "L29_titlekeep",
        "L30_rhyme",
        "L31_feuer",
        "L32_lotos",
        "L33_querpfeifer"
      ],
      "locale_naming": [
        "L11_ost",
        "L13_stadtwald",
        "L14_spreewald",
        "L07_linden"
      ]
    },
    "H_means": {
      "lexicon": 0.83,
      "honorifics": 0.67,
      "dialect": 1.0,
      "structure_paragraphing": 0.75,
      "embedded_song_form": 0.6,
      "locale_naming": 0.88
    },
    "table": {
      "lexicon": {
        "A": "less",
        "B": "less",
        "C": "less",
        "Cl": "less",
        "Dl": "less",
        "Da": "less",
        "H": "—"
      },
      "honorifics": {
        "A": "less",
        "B": "less",
        "C": "less",
        "Cl": "less",
        "Dl": "less",
        "Da": "MORE",
        "H": "—"
      },
      "dialect": {
        "A": "less",
        "B": "less",
        "C": "less",
        "Cl": "less",
        "Dl": "less",
        "Da": "less",
        "H": "—"
      },
      "structure_paragraphing": {
        "A": "less",
        "B": "less",
        "C": "less",
        "Cl": "less",
        "Dl": "less",
        "Da": "less",
        "H": "—"
      },
      "embedded_song_form": {
        "A": "at",
        "B": "at",
        "C": "at",
        "Cl": "less",
        "Dl": "less",
        "Da": "at",
        "H": "—"
      },
      "locale_naming": {
        "A": "less",
        "B": "less",
        "C": "less",
        "Cl": "less",
        "Dl": "less",
        "Da": "less",
        "H": "—"
      }
    },
    "arm_means": {
      "lexicon": {
        "A": 0.75,
        "B": 0.33,
        "C": 0.33,
        "Cl": 0.17,
        "Dl": 0.42,
        "Da": 0.58,
        "H": 0.83
      },
      "honorifics": {
        "A": 0.5,
        "B": 0.33,
        "C": 0.5,
        "Cl": 0.58,
        "Dl": 0.0,
        "Da": 0.83,
        "H": 0.67
      },
      "dialect": {
        "A": 0.17,
        "B": 0.0,
        "C": 0.17,
        "Cl": 0.33,
        "Dl": 0.17,
        "Da": 0.33,
        "H": 1.0
      },
      "structure_paragraphing": {
        "A": 0.0,
        "B": 0.0,
        "C": 0.0,
        "Cl": 0.0,
        "Dl": 0.25,
        "Da": 0.25,
        "H": 0.75
      },
      "embedded_song_form": {
        "A": 0.6,
        "B": 0.6,
        "C": 0.6,
        "Cl": 0.2,
        "Dl": 0.2,
        "Da": 0.6,
        "H": 0.6
      },
      "locale_naming": {
        "A": 0.25,
        "B": 0.25,
        "C": 0.25,
        "Cl": 0.5,
        "Dl": 0.25,
        "Da": 0.5,
        "H": 0.88
      }
    },
    "headline": "D-aim is MORE domesticating than H only on honorifics; at H on embedded-song; less on the other four. C-let & D-let less than H on all six."
  }
}