Who Decides What Counts as Bad Language?

Published: October 7, 2026. Updated: October 2026 by Andrew Apell, who builds PureWord and previously built Flookup Data Wrangler

The words we were not supposed to say

Every generation keeps a private museum of forbidden words. Words your grandparents would not have said in front of a vicar are now spoken over breakfast television, while words your grandchildren find unspeakable have not been coined yet. The museum rotates its collection constantly and nobody has ever published the exhibition schedule.

Which raises a question worth asking properly. There is no parliament of vocabulary, no academy with an official blacklist and no international body that meets on Thursdays to demote last year's shocking expression. So who actually decides which words are bad? The short answer is that everybody decides constantly and mostly by accident.

The longer answer is that a word can travel from ordinary to slang to taboo to normalised to outdated while another word quietly takes its place in the queue. Language does not keep a permanent list of villains. It keeps a waiting room.

Slang has always been breaking the rules

Slang is what happens when a community, a profession, a region or a generation decides that standard vocabulary is not doing the job. Surgeons have theirs, sailors had theirs, teenagers have an inexhaustible supply and every workplace develops a vocabulary for the specific misery of its industry. The words are often invented, borrowed or stolen from a neighbouring subculture, and they do their real work as a membership card: use them correctly and you belong.

It helps to picture language as a permanently rebellious teenager. It ignores the rulebook, adopts whatever it heard at school that week, experiments with identity in public and invents three new words before lunch. Occasionally one of those inventions turns out to be genuinely useful and gets invited to stay. Most of the rest quietly leave the house and are never mentioned again.

The point is that slang is not decay. It is language experimenting in public, and most of what we now consider perfectly respectable English arrived exactly that way, wearing the fashions of a community that polite society once disapproved of.

When ordinary words become bad

Words also travel in the other direction. A perfectly ordinary word can acquire a charge because of a historical event, a shift in social attitudes or simply repeated use as an insult, until the original meaning is buried under the reputation. Words do not exist in isolation. They arrive in sentences carrying everything that has ever been done with them.

There is a well known pattern here that linguists call the euphemism treadmill. A polite term for something uncomfortable becomes associated with the thing itself, starts to sound impolite, gets replaced by a newer polite term and then the treadmill starts again. The words for old age, poverty and illness have been swapped out this way many times across the centuries. The discomfort was never in the syllables. It was in the subject, and the subject patiently re-attaches itself to every new word we offer it.

Context does the rest. The same four letters can be a joke between old friends, a dismissal in an office and a disciplinary incident in a classroom. A word is not intrinsically bad in the way arsenic is intrinsically poisonous. It is bad the way a shout in a library is bad, which is to say the situation decides.

The strange life cycle of a bad word

Put the pieces together and most forbidden words follow a recognisable life cycle. They start ordinary, get adopted as slang, pick up offence somewhere along the way, harden into taboo, get softened into euphemism or replaced entirely and are finally forgotten. Today's shocking expression is tomorrow's linguistic fossil, sitting in old novels waiting for a reader to look it up.

History is generous with examples. Whole families of oaths that once made people genuinely flinch, built on religious references that everyone took seriously, now sound so quaint that fantasy novels use them for decoration. Words that could not be printed in respectable newspapers for centuries now appear in family programmes without anyone adjusting the volume. The offence drained out of them one decade at a time, and no committee voted on it.

The cycle rarely announces itself while it is happening. A word does not send a memo when it moves from taboo to old-fashioned. You only notice in hindsight, when a phrase that used to clear a room now sounds like something from a costume drama.

The internet broke the speed limit

For most of history this cycle ran on geological time. A word needed generations to spread beyond its village, and a century or two to become embarrassing. The internet has torn up that schedule completely. A phrase coined in a game chat on Tuesday can be in a television interview by the weekend and annoying by the following Thursday.

Social media, memes, messaging and online communities have compressed the entire life cycle into months. Coin, spread, peak, overuse, cringe and retire, all before a dictionary has finished debating whether the word is real. Lexicographers now track words the way meteorologists track storms, and even they have been known to describe the job as chasing something that is already soaking wet and three towns away.

The result is a peculiar modern experience: watching a word you adopted last month become something people apologise for using. Language has always moved. It has simply never moved this fast, in front of this many witnesses, with receipts.

So what is bad language?

Given all of the above, the idea that a simple blacklist can determine whether language is appropriate starts to look optimistic. Whether a word is inappropriate depends on the context, the audience, the speaker, the intent, the culture, the time period, the relationship between the people talking and what the word is actually doing in the sentence.

That is not a small list of factors, and several of them can change mid-conversation. The same word can be harmless in one sentence and inappropriate in the next, said by the same person to the same person. Anyone who has ever quoted somebody else's swearing in a complaint email has discovered this firsthand: the report of the word is not the word, and everybody somehow knows the difference even though the letters are identical.

So the honest answer to our opening question is this. Nobody decides. Everybody decides. The decision is remade constantly, in every conversation, by the people in the room.

The problem for a language filter

This is where filtering language stops being a vocabulary exercise and starts being a design problem. A basic filter asks one question: is this word on the bad word list? A smarter system has to ask a harder one: how is this word being used? The second question is the reason filtering is deceptively difficult, and the reason a filter that looks trivial on a whiteboard needs real thought in practice.

The obstacles arrive in a steady stream. Slang changes before any list can absorb it. Deliberate misspellings and inserted characters dodge plain text matching, which is why a filter catches far more than it misses but never everything. Euphemisms replace the words that got filtered, sometimes within the same week. Variations in spelling and regional forms multiply the list. Quotations and reported speech raise a philosophical question about whether the filter should censor the mention as well as the use. Reclaimed terminology means a word can be ordinary inside a community and hostile outside it. And the vocabulary itself keeps moving while all of this is being decided.

PureWord takes the practical route. It matches the words on its list as pages load, masks them or replaces them with clean alternatives, and lets you extend the list with custom words for the exact spellings your household or classroom actually meets. It filters page text, not audio, and it does not attempt to judge intent, because intent is not something a word list can see. What it can do is remove the words you decided you do not want on your screen, on every site you visit, which is a much more achievable goal. The step by step setup shows how far that gets you, and the comparison with Advanced Profanity Filter covers where the limits sit.

If language keeps changing, can a filter ever be finished?

Probably not, and that is not a failure. Language changes because people change, and people have an irritating habit of inventing new ways to say things, including several new ways to say the things you would rather not read. A word list is therefore not a monument. It is a garden, and it needs weeding.

The useful expectation is maintenance rather than completion. Set the list to match your own standards, add the spellings you actually encounter, review it now and then and accept that the language will keep supplying new material. The filter does not need to win the argument about which words are bad. It only needs to keep your screen the way you decided you want it, and to be easy to adjust when the language moves again, which it will, probably before you finish reading this sentence.

FAQ

Can a word be rude in one place and normal in another?

Yes. Meaning depends on culture, audience and period, which is why one universal list of forbidden words cannot exist.

Can a language filter catch disguised words?

Only some. Measured against 18 written evasion patterns, PureWord filtered 10. Spaces between letters, repeated letters and reversed text still get through, so add exact spellings as custom words.

Can a language filter ever be finished?

Probably not. Language changes as people change, so a filter needs fresh words and regular review rather than one final list.

Does PureWord decide what counts as rude?

No. It ships with a public word list and you add or remove words yourself. Parental Lock protects those settings behind a password.

Try PureWord free

PureWord masks profanity on every site as it loads. Free and private.

Add to Chrome