Common Irregular Plurals That Trip Up Even Native Speakers

Some irregular plurals are missed almost exclusively by non-native speakers still learning the exceptions. A smaller, more interesting group trips up fluent native speakers specifically, usually because the correct form conflicts with a more common, more intuitive-feeling pattern elsewhere in the language.

Octopus, cactus, and the false-Latin-rule trap

Octopus is probably the single most frequently mishandled plural among native speakers, and the specific error ("octopi") is instructive because it's not a random guess — it's an over-application of a real pattern (Latin "-us" nouns take "-i" plurals, as in cactus/cacti) to a word that doesn't actually follow it, since "octopus" is Greek rather than Latin in origin. The correct standard English plural is "octopuses"; the etymologically "correct" classical form would technically be "octopodes," a form almost nobody actually uses in practice. Fluent speakers who confidently produce "octopi" are, in a real sense, applying grammatical knowledge correctly — just to the wrong word's history.

Formula, index, and the register trap

Formula and index trip people up for a different reason: both have two genuinely correct plurals that are appropriate in different contexts, and using the "wrong" one for a given register reads as an error even though it technically isn't one. Mathematicians and scientists favor "formulae" and "indices"; general writing favors "formulas" and "indexes." A native speaker moving between technical and everyday writing contexts can genuinely second-guess which form belongs where, especially since both are accepted in dictionaries without a hard usage restriction attached.

Person, people, and persons

"Person" has an irregular plural, "people," that's used constantly and without hesitation — but a second, regular plural, "persons," also exists and survives specifically in formal, legal, and official contexts ("no persons shall enter without authorization," "missing persons report"). Native speakers frequently aren't consciously aware that "persons" is a legitimate plural at all, since "people" so thoroughly dominates everyday speech, which is why "persons" can read as an error in casual writing even though it's the historically older and, in certain formal registers, still the preferred plural.

Die and dice: a plural that lost its singular in common speech

The singular of "dice" is technically "die" ("roll the die," referring to one cube), but in everyday spoken usage "dice" has been reanalyzed by many speakers as both the singular and plural form ("roll the dice" used for a single cube as often as for several), effectively erasing the singular/plural distinction in casual speech even though dictionaries still list "die" as the technically correct singular. This is a live, in-progress case of a plural noun's boundary shifting under everyday use pressure — similar in kind to "data" being reanalyzed as an uncountable singular noun, just further along in the process because "die" (meaning the singular gaming cube) has become genuinely rare outside of set phrases like "the die is cast."

Child, children, and a plural marker stacked on top of another

Child/children hides something most native speakers never stop to notice: the word is marked plural twice over, not once. The "-er" ending is an old Germanic plural marker in its own right (the same one that survives in German's "Kinder"), and English then tacked a second plural suffix, "-en," onto that already-plural form — a kind of grammatical belt-and-suspenders that most likely happened because "-en" plurals like ox/oxen had become rare and distinctive enough that speakers unconsciously reached for that ending to mark "children" as unmistakably plural too. Nobody notices the double-marking in practice, because "children" is absorbed whole as a single vocabulary item in early childhood, long before anyone is in a position to pick apart its internal grammar.

Fish versus fishes: a rule native speakers rarely get to exercise

In practically every sentence a native speaker will ever produce, fish stays fish whether you're talking about one or a hundred ("I caught three fish"), so encountering "fishes" in the wild tends to read as a mistake rather than what it actually is: a separate, narrower plural reserved for a specific job — naming several distinct species side by side, not tallying up a catch. A field guide describing "the fishes native to this river system" is using the word correctly to mean the distinct kinds present, not counting individual animals — a genuine, semantically meaningful distinction that almost never comes up outside biology and ichthyology, which is exactly why so few fluent speakers have ever had occasion to use it themselves.

Nouns where two competing plurals signal register, not correctness

Beyond formula and index, several other nouns split along a similar register line without either side being objectively wrong: "antenna" takes "antennae" in biology (referring to insect antennae) but "antennas" in everyday and technical-broadcast contexts (referring to radio or television antennas); "stadium" has largely abandoned its classical Latin plural "stadia" for the fully regular "stadiums" in virtually all contexts except specialized surveying and historical usage, where "stadia" survives as a unit of measurement. Native speakers moving between casual and specialist writing contexts are the ones most likely to genuinely hesitate over which of these paired forms is expected, precisely because both are correct and the choice is entirely about audience and register rather than about one form being an error.

Why native speakers are more confident, not less, and why that backfires

It's worth naming directly why this group of plurals specifically trips up native speakers more than learners in some cases: a fluent speaker's confidence in their own intuition is exactly what causes the error, because the wrong form (octopi, persons-as-error, a mismatched register choice) genuinely sounds correct to an ear that's pattern-matched it against a different, superficially similar rule elsewhere in the language. A learner encountering "octopus" for the first time has no competing intuition pulling them toward "octopi" and is more likely to simply look up or ask about the correct plural rather than confidently generating a plausible-sounding wrong one. This is the same dynamic that produces hypercorrection in other areas of grammar: the errors made by careful, knowledgeable speakers often come from over-applying a real rule to a case it doesn't actually cover, rather than from not knowing any rule at all.