Pierre Bezúkhov, Scored by His Plate

Every food, meal, and hunger reference inside Pierre-focused paragraphs of Tolstoy’s War and Peace (Maude translation), set against the emotional temperature of the surrounding text. 2,492 Pierre paragraphs, 542 food references, 346 with enough emotion words to score.

542
Food / meal / hunger refs
346
Scoreable events
+0.17
Mean sentiment near food
14 / 17
Books with a Pierre food scene

Timeline: mean emotional valence per book

Average sentiment of the 220-character windows surrounding each food reference, grouped by book. Range is −1.0 (all negative emotion words) to +1.0 (all positive). Books where Pierre is narratively absent are omitted.

00.51IIVVIIXXIXIIIXVEp.1

Food-scene density

Raw count of food / meal / hunger tokens appearing in Pierre-focused paragraphs per book. The three troughs at II, VII, and Ep. 2 are the three books where Pierre himself barely appears.

050100150IIVVIIXXIIIEp.1Ep.2

Before vs. after each bite

Does eating itself regulate Pierre’s mood? Mean sentiment of the 220 characters immediately preceding the food token compared with the 220 characters immediately following it, per book. If meals soothed him, the green bars would consistently top the blue.

00.51IIVVIIXXIXIIIXVEp.1
Before eating
After eating

Food references by category

The 542 references split nine ways. “Meal” names dominate; genuine “hunger” is rare and almost entirely localized to the retreat from Moscow and the prisoners’ column.

Meal names (din…22%
Setting (table,…17%
Vessels (plate,…15%
Drink (wine, vo…10%
Eating verbs (e…9.6%
Foodstuffs (bre…9.2%
Satiety (full, …8.5%
Drinking verbs …7.9%
Hunger (hungry,…1.3%

Sentiment distribution of scored events

77% of scored events land in one of the two extremes. Pierre almost never eats in a neutral emotional register — meals are either deeply warm or deeply unsettled.

0100200Strongly ne…Mildly nega…Neutral [−0…Mildly posi…Strongly po…

Statistical summary

Global sentiment around food (n = 346)

Mean, combined window

+0.168

Mean, before window

+0.089

Mean, after window

+0.125

Std. dev., combined

0.869

Median (approx. by histogram)

+0.66

Pearson correlations

Before vs. after sentiment, per event (n = 346)

+0.084

Book index vs. mean sentiment (n = 14)

−0.115

Book index vs. event count (n = 17)

−0.316

Event count vs. mean sentiment, per book (n = 14)

+0.048

How to read the correlations
Values are Pearson r in [−1, +1]. |r| below 0.1 is effectively no linear relationship; between 0.1 and 0.3 is weak; 0.3–0.5 is moderate. None of the four correlations rises above the “weak” band — which is itself the finding.

Flagship food scene per book

The food reference in each book with the most extreme sentiment in its surrounding window. Red rows are strongly negative scenes; green rows are strongly positive. These are the emotional pinnacles that the summary statistics collapse together.

BookCategoryKeywordSent.Snippet
I (1805)mealdinner+1.00“Well, what answer am I to take? Will you come to dinner at the Rostóvs’?”
III (1805)drinkwine+1.00…they enjoyed their Rhine wine, sauté, and ices, and however they avoided looking at the young couple…
IV (1806)mealbanquet-1.00Dólokhov’s handsome, insolent, hard, and mocking face as he had seen it at the banquet… pale, quivering, and suffering.
V (1806–07)mealsupper+1.00The old prince came in to supper; this was evidently on Pierre’s account.
VI (1808–10)satietysatisfied+1.00Berg was satisfied and happy. The smile of pleasure never left his face. The party was very successful…
VIII (1811–12)mealdinner-1.00…a buzzing in his head after dinner or supper he chatted or listened… under the influence of wine…
IX (1812)drinkdrank+1.00Pierre still went into society, drank as much and led the same idle and dissipated life…
X (1812)satietyfull-1.00…his lips trembled and he began to cry, in the way full-blooded grown-up men cry, though angry with himself for doing so.
XI (1812)satietyself-satisfied+1.00the officer’s handsome, self-satisfied face, and… the eager enthusiasm with which he spoke of women.
XII (captivity)vesselknife+1.00he took out a knife, cut something, closed the knife, placed it under the head of his bed, and, seating himself comfortably…
XIII (1812)mealtea-1.00…shoes Karatáev had made for him from some leather a French soldier had torn off a tea chest… went up to the sick man…
XIV (1812)hungerhungry-1.00…cold and hungry soldiers should have to stand and guard equally cold and hungry Russians who froze and lagged behind…
XV (1812–13)satietyfull-1.00Pierre listened to her with lips parted and eyes fixed upon her full of tears. As he listened he did not think of Prince Andrew, nor of death…
Epilogue Isatietyfull-1.00…she did these things not under any external impulse as people in the full vigor of life do…

Interpretation: does food stabilize Pierre?

Food is emotionally polarizing, not regulating
47% of scored scenes fall in the strongly-positive bucket and 30% in the strongly-negative bucket — 77% of Pierre’s food moments are emotional extremes, only 13% are neutral. He never eats indifferently: each plate is either an index of warmth (a Rostóv supper, his wedding feast, Karatáev’s baked potato in captivity) or of disintegration (the banquet that precipitates the Dolokhov duel, the prisoners’ march, Andrew’s dying room).
Eating itself does not shift his state
The Pearson correlation between sentiment immediately before and immediately after each food reference is only +0.084 (n = 346). If meals regulated Pierre’s mood, we would expect negative scenes to soften after the act of eating; instead, the emotional temperature of a scene mostly holds, or flips unpredictably, across the bite. Food is setting, not remedy.
The war-and-captivity axis: fewer meals, warmer ones
Food-scene density falls as the novel militarizes (r = −0.32 against book index), yet the two highest mean-sentiment books are XI (Moscow, 1812, +0.51) and XII (captivity, 1812, +0.69) — the only stretch where Pierre is genuinely hungry. Scarcity, not abundance, produces Tolstoy’s warmest food scenes for him. The arc of dissipation → duel → Masonic reform → captivity → integration traced by the plot is mirrored in his food life: the banquet table recedes, the shared rusk of a fellow prisoner advances.
The coda flips it again
Both Book XV and Epilogue I show positive before-windows but negative after-windows (before +0.14 / after −0.05 in XV; +0.16 / −0.33 in the epilogue). Pierre’s domestic post-war table is the one place where food consistently deflates rather than lifts him — marriage, fatherhood, and habituation quietly invert what captivity’s scarcity taught him.

Method notes

2,492 Pierre paragraphs219 food-lexicon tokens87 positive / 133 negative emotion words±220-char sentiment windowsMaude translation, Gutenberg #2600

A paragraph is “Pierre-focused” if it names Pierre or Bezúkhov, or falls within two paragraphs after such a naming with no intervening rival male proper name. Sentiment scores are normalized as (positive − negative) / (positive + negative) over each window. Books II, VII, and the Second Epilogue contribute no events because Pierre is narratively absent from them. Source analysis script: analyze_pierre.py in the workspace root; raw per-event data in pierre_food_analysis.json.