Text Statistics

What exactly is a word

Ask three programs for a word count and you can easily get three answers. The disagreement is not a bug, it is a definition problem. A hyphenated compound like "state-of-the-art" is one word to a typesetter and four to a naive splitter. An apostrophe in "don't" must not split the word, while an apostrophe used as a quotation mark should. Numbers, URLs and email addresses each invite a different rule.

This tool treats a word as a run of letters or digits, optionally continued by an apostrophe and more letters. Because the matching is Unicode-aware, accented Latin, Greek, Cyrillic and CJK characters all count as letters rather than being silently dropped. Punctuation between runs separates them, so "well-known" counts as two and "it's" counts as one. Knowing the rule matters more than which rule is chosen, because it lets you predict the number instead of arguing with it.

Sentences, paragraphs and lines

Sentence detection without a language model is heuristic by nature. Splitting on periods, question marks and exclamation marks works for ordinary prose and fails predictably on abbreviations: "Dr. Smith arrived at 9 a.m." looks like three sentences to a naive splitter. Treat the sentence count as a good estimate for readability work rather than an exact figure.

Paragraphs and lines are more mechanical but still distinct. A line is anything separated by a newline, so a hard-wrapped file has many lines and few paragraphs. A paragraph is a block separated by one or more blank lines, which matches how Markdown and most editors think. Comparing the two numbers tells you quickly whether text arrived hard-wrapped, which matters when pasting into a system that reflows.

Reading time and frequency

The reading-time estimate uses 200 words per minute, a common average for silent reading of ordinary prose. Real speeds vary widely: technical material with code samples runs slower, familiar marketing copy runs faster, and reading on a phone is slower than on a desktop. Use the figure for relative comparison, not as a promise to your readers.

Word frequency is where the tool earns its place in editing. Sorting by count reveals unconscious repetition, the crutch words that appear eleven times in eight hundred, and whether the topic you meant to write about is actually the most frequent noun. The "ignore common words" toggle removes closed-class function words such as articles, prepositions and pronouns, which otherwise dominate every list in every language and tell you nothing.