Take the 2-minute tour ×
English Language & Usage Stack Exchange is a question and answer site for linguists, etymologists, and serious English language enthusiasts. It's 100% free, no registration required.

I'm talking about percentage of letters, not words, in case it isn't clear. Is there a way to gauge this?

share|improve this question
I'd guess between one and three percent. –  Daniel Sep 28 '11 at 12:43
Hmm... maybe I just don't use proper nouns enough! 2.5% - 5% seems more like it, judging from JSBangs' answer. –  Daniel Sep 28 '11 at 18:43
What prompted this question, out of curiosity? –  onomatomaniak Sep 29 '11 at 15:51
I saw my name coming up in chat stats at our university's DC server as the user who typed most in capitals. It was around 3% and I was wondering was it normal. –  Capt.Nemo Sep 29 '11 at 17:21
@Capt.Nemo Around 3% is normal for books with real sentences, but a lot of IM is either all lowercase or without initial capitals. You must have high standards! –  Hugo Sep 30 '11 at 10:41

2 Answers 2

up vote 14 down vote accepted

Between 2-4%, depending on the text and the genre.

To determine this, I downloaded a variety of texts from Project Gutenberg, then wrote a simple program to count the total number of alphabetic characters and the total number of capitalized characters in each file. Here are the raw numbers:

Title (Author)                  Letters      Caps      Pct Caps
Pride and Prejudice (Austen)    2641527      14177       2.56%
History of ... (Gibbon) [*]     1295410      34893       2.69%
Moby Dick (Melville)             968516      28204       2.91%
Great Expectations (Dickens)     777248      23668       3.05%
Shunned House (Lovecraft)         66779       2223       3.32%
Tom Sawyer (Twain)               312196      10746       3.44%
Somebody... (Doctorow) [!]       495594      17366       3.50%
Bible, King James Version       3343105     117344       3.51%
Ulysses (Joyce)                 1203807      55244       4.58%
Hamlet (Shakespeare)             139132       7812       5.61%

[*] Full title: History of the Decline and Fall of the Roman Empire
[!] Full title: Somebody Comes to Town, Somebody Leaves Town by Cory Doctorow

Hamlet comes in with the highest percentage capitals, probably because it's a script and the repeated character names are always capitalized. Ulysses is also unusually high, because Joyce is weird and uses lots of capitals in unexpected places. The other texts run from about 2.5% to 3.5%.

Edit: Added Melville, Lovecraft, Dickens, Doctorow to fill out the comparison of contemporary, early 20th century, and 19th century authors. I'm not seeing much of a trend here, with the most contemporary authors actually having a somewhat higher percentage of capitals than the earlier models. I suspect that more modern writers have shorter sentences, and therefore more sentence-initial capitalization, and that this effect swamps the effect of freer capitalization in earlier texts.

share|improve this answer
My gut feeling is that total percentage capitalisation would have been continuously declining over the last couple of centuries at least. 2% is probably the upper end of the scale for the last few decades, particularly if you exclude acronyms, of which I'm sure there are far more in common use nowadays. –  FumbleFingers Sep 28 '11 at 14:38
@FumbleFingers, why would you exclude acronyms? Nothing in the question suggests that it would be desirable to do so. –  JSBձոգչ Sep 28 '11 at 15:09
No particular reason. Just making the point that if it is true that we use capitalised acronyms more often today, and if your simple program showed a lower percentage for more recent publications, that would imply the trend towards less capitalisation in "general" writing was even more marked. Although thinking about it I would guess average sentence length has probably fallen over time, which would go the other way. –  FumbleFingers Sep 28 '11 at 15:31
Didn't Joyce also write a book that was all one sentence, or contained no punctuation? –  oosterwal Sep 29 '11 at 16:22
Or maybe because Hamlet is in blank verse, where you have a capital letter at the beginning of every line. –  GEdgar Sep 29 '11 at 18:09

Assuming the Project Gutenberg etext of Herman Melville's Moby-Dick is representative of all English literature:

  • Uppercase letters: 24,559
  • (Lowercase letters: 936,138)
  • All characters: 1,231,937

24,559 / 1,231,937 = 2.00% capitals letters (across all characters).

Or as a percentage of letters (ignoring non-letter characters):

  • Uppercase letters: 24,559
  • (Lowercase letters: 936,138)
  • All letters: 960,737

24,559 / 960,737 = 2.56% capital letters (across all letters).

Edit 2: Taking this a step further, I ran a script on the plain text ebooks from Project Gutenberg's CD and DVDs:

Source                          Caps        Letters   Pct Caps     Characters  Pct Caps
Moby-Dick: 1 ebook            24,559        960,737      2.56%      1,231,937     2.00%
2003 CD: 594 ebooks       11,407,295    319,286,662      3.57%    417,687,793     2.73%
2006 DVD: 16,536 ebooks  179,318,621  4,913,640,039      3.65%  6,380,437,180     2.81%
2010 DVD: 14,792 ebooks  152,637,904  4,102,894,980      3.72%  5,433,866,318     2.81%
Total: 31,923 ebooks     343,388,379  9,336,782,418      3.68%  12,233,223,228    2.81%

The median values are 3.72% and 2.85%, and mode values are 3.12% and 2.29%.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.