Explorer

Sentence Length

How long is a sentence in each corpus? The distribution you weren’t expecting.

Shakespeare’s median sentence is five words long. The Founders’ median sentences run seventeen to twenty-six words. Washington, alone among them, often writes sentences over forty words. This is the most dramatic single divergence the project finds between the two corpora, and it’s a divergence none of the ten other measures can see, because the units they count (words, collocates, metaphor families, archaic forms) don’t look at sentence structure.

The histograms below show the same data each way: each author’s distribution of sentence lengths (sampled, 5,000 random sentences per author, fixed seed for reproducibility). Bins are 1–5 words, 6–10, 11–15, up to 100+. Read down the column to see how each author distributes their sentence lengths; read across the row to compare any one bin across authors. The Shakespeare row sits at the top so the contrast is immediate.

Shakespeare

mean 9.8 words · median 5 · std dev 13.3

1-5
6-10
11-15
16-20
21-25
26-30
31-40
41-50
51-75
76-100
100+

John Adams

mean 25.1 words · median 17 · std dev 30.0

1-5
6-10
11-15
16-20
21-25
26-30
31-40
41-50
51-75
76-100
100+

Benjamin Franklin

mean 28.6 words · median 20 · std dev 29.4

1-5
6-10
11-15
16-20
21-25
26-30
31-40
41-50
51-75
76-100
100+

Thomas Jefferson

mean 30.1 words · median 17 · std dev 47.9

1-5
6-10
11-15
16-20
21-25
26-30
31-40
41-50
51-75
76-100
100+

James Madison

mean 25.4 words · median 19 · std dev 30.8

1-5
6-10
11-15
16-20
21-25
26-30
31-40
41-50
51-75
76-100
100+

George Washington

mean 38.5 words · median 26 · std dev 42.1

1-5
6-10
11-15
16-20
21-25
26-30
31-40
41-50
51-75
76-100
100+

Alexander Hamilton

mean 28.8 words · median 22 · std dev 30.5

1-5
6-10
11-15
16-20
21-25
26-30
31-40
41-50
51-75
76-100
100+

What the histograms say

Drama is short. Prose is long. There is no overlap.

Almost everything you can ask of the corpus about vocabulary and metaphor and rhetorical figure is a question about resemblance. Sentence length is a question about genre. Shakespeare wrote for the stage. His sentences are interrupted by speaker changes, by stage directions, by the metrical constraints of blank verse. They are short by construction. The Founders wrote letters, essays, official memoranda, and presidential addresses. Their sentences are constructed for the eye reading on the page, the orator addressing a Congress, the diplomat composing for a foreign court.

This is the one measure where the project finds no influence at all. The Founder closest to Shakespeare on this dimension is Adams, whose mean sentence length of 25 words is still nearly three times Shakespeare’s. The most distant is Washington at 38.3: the prose of a soldier-statesman writing with the cadence of military orders. Neither extreme is a literary failing; both are appropriate to the kind of writing being done. But it is worth recording that whatever the Founders absorbed from Shakespeare, they did not absorb his sentences.

The histogram for Shakespeare shows the genre signature most clearly: a sharp spike in the 1–5 and 6–10 word bins (lines of dialogue and short speeches), then a long tail. The Founders’ histograms are essentially Gaussian, shifted right by two or three bins.