A Grammar Of The English Language George Oliver Curme

Working With Curme's Grammar

I ran into a problem a few years ago when I was building a parser for historical English texts. A particular construction appeared in an 18th-century pamphlet that modern grammars simply don't account for, and I needed to figure out whether it was a relic of an older syntactic pattern or just a transcription error. I went to Curme's two-volume set — it's the only reference I trust for the verb system and the morphological categories. The answer ended up sitting in Volume 2 around page 587, buried under a discussion of the subjunctive mood in adverbial clauses. I spent about twenty minutes looking for something that isn't organized topically the way modern textbooks are. That's the thing about A Grammar Of The English Language George Oliver Curme: it doesn't hand you answers efficiently. You have to know where to look. The book came out in 1931, published by Ginn and Company. It runs roughly 1,700 pages across two volumes. Volume 1 covers the noun, pronoun, adjective, verb, and parts of speech. Volume 2 moves into syntax, clause structure, and sentence analysis. It is not a quick reference. It is a survey grammar that leans heavily on descriptive analysis rather than prescriptive rules, which is actually why it survives when so many other texts from the same era are obsolete. What makes it useful today isn't its completeness — every grammar from that period is incomplete by modern standards — but its treatment of Germanic syntactic structures and its detailed breakdown of verbals. Curme was trained in the German tradition of grammar. He treats English as a language with Germanic morphology at its core, and that means his treatment of participles, infinitives, and gerunds is more technically precise than most contemporary handbooks. When I encounter a sentence where the boundary between a present participle clause and a gerundial subject is ambiguous, I open Curme. No other book gets the distinction right with the same consistency.

The way I use it is specific. I don't read it cover to cover. I treat it as a lookup tool organized by grammatical category, and I go to it with a targeted question. Volume 1, sections on the verb — that's where I look first. The subsection on finite versus non-finite verbs, the treatment of modal auxiliaries, the classification of weak and strong verbs. Curme's definitions here are meticulous. He distinguishes between what he calls "pure" and "mixed" infinitives, a categorization that turns out to be genuinely useful when you're analyzing constructions like let him go versus to let him go, and the difference isn't just stylistic. One specific issue I ran into involved a corpus of Middle-to-Early Modern English transition texts where authors used what looked like double infinitives in a periphrastic construction. The expected analysis would classify these as errors or dialectal variants. Curme's treatment of the infinitive in complex predicate structures gave me a framework for reading them as systematic rather than anomalous. He discusses the to-infinitive as a marker of purpose and extent, and the overlap between those functions creates constructions that modern shorthand dismisses as redundant. I cross-referenced his analysis with the actual tokens in the text and found about forty cases across the corpus that fit the pattern. That changed how I tagged the entire subset. There are real limitations you need to account for. Curme's phonology section is thin by modern standards. He was writing before the International Phonetic Alphabet was widely adopted in American linguistics, and the transcription system he uses is inconsistent. If you need to analyze sound change or dialect variation, you will need a secondary source — someone like Lass or McCrum for that particular problem. His syntactic analysis also predates transformational grammar and generative frameworks, so his treatment of embedding, control, and raising constructions is going to feel clunky if you're accustomed to Chomskyan terminology. He describes what happens without formalizing the underlying mechanism. That isn't a flaw in his work, but it is a gap if you're trying to interface his categories with computational models.

Another practical problem: the index is serviceable but not comprehensive. Cross-references are sparse. If you're searching for a topic like antecedent in the context of relative clauses, you might find it under "relative pronouns" instead of where you'd expect. I learned to search by example rather than by keyword, flipping through the relevant chapter until the pattern matched what I was looking for. That takes time, maybe ten to fifteen minutes per search on average, but it's faster than the alternative of hunting through secondary sources that may apply the terminology incorrectly. The book is in the public domain. You can find free digital copies through HathiTrust, Internet Archive, and some university library repositories. The scans are generally legible but vary in quality depending on the source volume. The HathiTrust edition is the most reliable for close reading because the page resolution holds up well at the word level. Some of the older PDF versions from archive.org have OCR artifacts in the footnotes that can make individual entries hard to parse. If you're citing specific passages, verify against the scanned image rather than relying on the full-text search result. I'd recommend it for anyone doing detailed syntactic analysis of historical English or working on a grammar reference that needs to go beyond the surface descriptions found in modern pedagogical texts. It's not a book you read for pleasure. It's a book you open when you need to resolve something that contemporary references either ignore or mislabel. The time investment is real — don't expect to find a quick answer — but the depth of coverage in the verb system and clause structure makes it worth the effort if your work intersects with those areas.

What It Covers That Other Grammars Don't

Curme's discussion of predication is one of the sections I return to most often. He treats predication as a relationship between subject and predicate that isn't reducible to the copula alone, and he lays out a taxonomy of predicative expressions that includes complements, modifiers, and adjuncts in a way that anticipates later functional syntax work. Modern introductions tend to conflate these or sidestep the issue entirely. His distinction between subject complement and object complement is technically precise, and he supports it with a large set of examples drawn from actual usage rather than invented sentences. The section on clause types in Volume 2 covers independent, subordinate, and coordinate structures with a level of detail that most reference grammars omit. He separatesyndetic from asyndetic coordination, which matters when you're analyzing texts where conjunction omission is systematic rather than accidental. I've seen too many people misread asyndetic lists in early modern prose as stylistic flair when they were actually following a syntactic pattern that Curme documents clearly. The pattern holds in legal and religious texts particularly, where the omission of conjunctions tends to cluster in specific structural positions. His treatment of the subjunctive deserves mention because it's the area where the book is most likely to surprise readers who approach it expecting prescriptive guidance. Curme describes the subjunctive as a living feature of English, not a declining one, and he provides evidence from contemporary usage (for his time, the 1920s) that the form was still productive in formal and semi-formal registers. That observation has held up. The decline he anticipated in some contexts did occur, but not across the board. The mandative subjunctive in American English, for example, remains active in a way that his description allows for, even if he couldn't have predicted the dialectal split that would develop later.

If you're working with a text and you encounter a construction that doesn't fit the standard model, try Curme before you assume the text is irregular or the punctuation is wrong. He covers enough edge cases that you'll often find a parallel entry. It takes patience to use the book effectively, and it won't replace a computational tool for large-scale analysis, but for resolving individual grammatical questions, it remains one of the most reliable single-source references available.