The word “the”
Count every word in a whole book and put them in order, commonest first. Whatever the book, whatever the language, the counts fall on the same straight line. Pick a book.
What this is. Seven whole books from Project Gutenberg, in five languages. A word is a run of letters, lower-cased; the Gutenberg licence text at each end is cut off and nothing else is cleaned. The chart shows each word's count against its rank, on scales where every step is ten times the last.
Zipf's law says the word of rank r turns up about 1/r as often as the first. In Moby Dick “the” appears 14,532 times; second is “of” with 6,601, which is 45% as many. The fit is not exact for any one word (in Pride and Prejudice the second word is 92% as common as the first), but across a thousand ranks every one of the seven books has a slope between −0.91 and −1.02. Two consequences hold in all seven: about a hundred different words make up half of the text, and between a third and two thirds of the different words in a book are used exactly once. Nobody has an agreed explanation of why language does this.