Liberman & Pullum: Far from the Madding Gerund
Summary: Mark Liberman and Geoffrey Pullum’s collection of Language Log posts — a sustained campaign of descriptive linguistics against prescriptive grammar myths, using corpus data and historical evidence to show that most “grammar rules” are recent inventions unsupported by the practice of great writers.
Sources: Physical book clipnotes/Library_Quotes_Batch_7.md
Source pages: Library Quotes Analysis: Batch 7 (Books 86-100 Selection), Quotes batch 12
Quote pages: Liberman — Far from the Madding Gerund
Last updated: 2026-05-06
Grammar rules as myths
“Most grammar ‘rules’ found in newspapers are myths invented by 18th-century poets.” “Descriptive Integrity: Observing how English is actually used rather than how it ‘should’ be.” “Logic is not the same as grammar; a sentence can be illogical but perfectly grammatical.”
Liberman and Pullum document case after case where the “rules” promulgated by style guides and grammar columns have no historical basis: the prohibition on split infinitives (not found in any pre-19th-century grammar); the injunction against ending sentences with prepositions (famously attributed to 18th-century poet John Dryden, based on a misapplication of Latin grammar); the ban on “hopefully” as a sentence adverb (not applied consistently even by those who enforce it). The rules are prescriptions from authority, not descriptions of how the language works. Compare bryson-troublesome-words for the same argument at a popular level.
The Grammar Snoot as class enforcer
“The ‘Grammar Snoot,’ who uses linguistic pedantry to enforce social class.” “Why do people get so angry about split infinitives when Shakespeare used them?” “Language change is not language decay.”
Pullum (the leading voice here) makes an explicitly political observation: grammatical correctness functions as social gatekeeping. The rules correlate less with clarity than with the speech patterns of educated elites — and enforcing them marks insiders versus outsiders, not careful thinking versus sloppy thinking. Compare nunberg-going-nucular for the political analysis of how pronunciation serves the same function; compare bryson-troublesome-words on clarity as the only rule that matters.
Corpus linguistics
“Corpus Linguistics: Using massive databases of speech to prove that ‘correctness’ is a moving target.” “Does the ending of a sentence with a preposition really harm the clarity of a thought?”
Corpus linguistics — the analysis of very large databases of actual language use — is Liberman and Pullum’s empirical weapon: if a rule is real, it will be followed by actual speakers and writers. When the corpus shows that the best writers routinely violate a rule, the rule is not a rule. This is the quantitative turn in descriptive linguistics; what was previously a theoretical argument becomes an empirical one. Compare crystal-how-language-works on grammaticalization: corpus data is how we track language change over time.
The passive voice myth
“The Passive Voice Myth: Debunking the idea that the passive voice is inherently ‘weak’ or ‘wrong’.”
The passive voice is stylistically overused in bad writing — but so is everything that bad writers reach for. The passive itself is a grammatical construction with legitimate uses (suppressing an unknown or irrelevant agent; topicalizing the affected party; achieving cohesion across sentences). Blanket prohibition reveals ignorance of what the passive actually does. Compare austin-speech-acts: understanding what a construction does (its illocutionary force) is more useful than evaluating it by a content-free style rule.
Quotes
Linguistic Identity
“The way we talk about grammar is almost always a proxy for the way we feel about the people who use it.” (p. 24)
Cognition
“Our brains are hardwired for syntax, but our culture is what determines which rules we feel are ‘correct’.” (p. 56)