Zipf's Law
Fetch the full text of any Wikipedia article — or paste your own — rank every word by frequency, and watch the counts settle onto the same 1/rank curve, smoother the bigger the corpus gets.
Zipf's law says that in natural language a word's frequency is inversely proportional to its rank. The most frequent word shows up about twice as often as the second, three times as often as the third, and so on down a long tail. It holds across languages, authors, and eras with unreasonable consistency.
Paste in any text and it appears almost immediately: the observed counts trace the same 1/rank curve that theory predicts. It is one of those laws that feels like it shouldn't be true — nobody coordinates their word choices — and yet the shape is always there.
Ported from a Python 2 script that pulled a Wikipedia article, tokenized it, and plotted counts with matplotlib. Reimplemented in TypeScript with your own text as the input and a live log-scale chart.
Follow the first link on any Wikipedia article and you end up at Philosophy.