In linguistics or usage, hypercorrection is a non-standard usage that results from the over-application of a perceived rule of grammar or a usage prescription. A speaker or writer who produces a hypercorrection generally believes that the form is correct through misunderstanding of these rules, often combined with a desire to appear formal or educated.
Linguistic hypercorrection occurs when a real or imagined grammatical rule is applied in an inappropriate context, so that an attempt to be "correct" leads to an incorrect result. It does not occur when a speaker follows "a natural speech instinct", according to Otto Jespersen and Robert J. Menner.
Hypercorrection is sometimes found among speakers of less prestigious language varieties who produce forms associated with high-prestige varieties, even in situations where speakers of those varieties would not. Some commentators call such production hyperurbanism.
Studies in sociolinguistics and applied linguistics have noted the over-application of rules of phonology, syntax, or morphology, resulting either from different rules in varieties of the same language or second-language learning. An example of a common hypercorrection based on application of the rules of a second-language is the use of octopi for the plural of octopus in English; this is based on the faulty assumption that octopus is a second declension word of Latin origin when in fact it is third declension and comes from Greek.
Sociolinguists often note hypercorrection in terms of pronunciation (phonology). For example, William Labov noted that all of the English speakers he studied in New York City in the 1960s tended to pronounce words such as hard as rhotic (that is, /hɑːrd/ rather than /hɑːd/) more often when speaking carefully. Furthermore, middle class speakers had more rhotic pronunciation than working class speakers did. However, lower-middle class speakers had more rhotic pronunciation than upper-middle class speakers. Labov suggested that these lower-middle class speakers were attempting to emulate the pronunciation of upper-middle class speakers, but were actually over-producing the very noticeable R-sound.