Across 147,478 entries this dictionary records 207,235 distinct senses — an average of 1.4 per entry. Only 825 entries, 0.6% of the corpus, carry ten meanings or more, and almost all of them are short, common, everyday words: break, cut, run, play, make. Rare and technical words carry one meaning because they were coined to carry exactly one.
The shape of the distribution
207,235 senses across 147,478 entries sounds like plenty to go round. It is not distributed evenly. The great majority of entries carry a single sense, and a very small number carry dozens.
825 entries — 0.6% of the corpus — have ten senses or more. At the extreme, “break” carries 75 in this corpus, “cut” 70, “run” 57, “play” 52 and “make” 51. Every one of those is a one-syllable word in the top commonness band.
At the other end, 99,506 entries sit in the rarest band, and most of them mean one thing. A species name, a mineral, a piece of clinical vocabulary: coined for a referent, still attached to it.
Why the commonest words collect the most meanings
New meanings do not attach to words at random. They attach to words that are already available — short, familiar, cheap to say, already in everyone's mouth. When English needs to talk about something new, it reaches first for what it is already holding.
So “run” acquires a sense for machinery, for software, for a candidate in an election, for a length of stocking, for a score in cricket. None of these required a new word; each borrowed a word that was already doing several jobs and could take another.
The result is circular in a way that is genuinely self-reinforcing: being frequent makes a word available, being available makes it collect senses, and having many senses makes it usable in more contexts, which makes it more frequent. That is why the top of the polysemy list and the top of the frequency list are nearly the same list.
The reverse case: vocabulary coined for one job
17,783 entries here are plant words and 14,320 are animal words — between them a fifth of the corpus. Almost all of them carry exactly one sense.
That is not a shortcoming of the database. A name minted to designate one species designates one species. It has no history of being stretched, because nobody reaches for a technical binomial when they want to say something else.
This is the practical reason a dictionary of 147,478 entries is not a vocabulary of 147,478 words to learn. Much of it is reference material that exists to be looked up once, and the part that rewards study is the small, heavily-loaded core.
What this means when you look a word up
Sense 1 is not “the” meaning, and it is not the oldest or the most basic. Senses here are in WordNet's own order, which is roughly how often each sense was tagged in annotated text — most-attested first. For a word with 40 senses, the one you want may be twelfth.
So checking a spelling and reading an entry are different activities. Checking takes a glance at sense 1. Reading an entry for a common word means scanning the sense list for the one that matches your context, which is why the pages here group by part of speech first — that alone usually removes three-quarters of the list.
It also means a “definition” of a common word is a fiction. There is no single statement of what “set” means. There are dozens of related uses, and the entry is a map of them rather than an answer.
What this means when you write
A common word is ambiguous in direct proportion to how common it is. If a sentence has to be unmistakable, the band-5 word is the risky choice, not the safe one — it is the word carrying the most alternative readings.
The move is not to reach for something obscure. It is to reach one band out: 4,497 entries sit in the Common band and 10,501 in the Familiar band, and those carry far fewer competing senses while remaining words your reader knows.
The exception is speech and plain instruction, where ambiguity is resolved by context faster than a rarer word can be understood. Then the heavily-polysemous word is right, precisely because everyone has all its senses loaded.
A caveat about counting senses at all
“Break has 75 meanings” is a statement about this corpus, not a fact about English. Sense counts depend entirely on how finely a lexicographer splits, and different dictionaries split differently — one may record two senses where another records five.
What survives across every dictionary is the shape, not the number: a small set of frequent words carrying many senses, and a long tail carrying one. Compare figures within one dictionary and they are meaningful. Compare them across two and you are mostly measuring editorial policy.
Questions
- Which English word has the most meanings?
- In this corpus, “break”, with 75 recorded senses, followed by “cut” at 70 and “run” at 57. Other dictionaries name other winners, because the count depends on how finely senses are split rather than on the language.
- How many meanings does the average word have?
- 1.4 — 207,235 senses across 147,478 entries. The average is misleading on its own, because most entries have one and a few have dozens.
- How many English words have ten or more meanings?
- 825 entries here, which is 0.6% of the corpus. They are collected in one list, ordered by sense count.
- Is the first sense listed the main meaning?
- It is the most-attested sense in the annotated corpora WordNet draws on, which is not the same as the oldest, the most basic, or the one you want. For a heavily polysemous word, read past it.