This question is specifically about The Cambridge Grammar of the English Language by Huddleston and Pullum.

Here's CGEL's definition of word:

In order to avoid possible misunderstanding we will restrict the term word to the syntactically-oriented sense, so that hard and harder are different words, and likewise are and is. (Page 27)

How CGEL defines noun is not so clear, but from the following distinction between proper noun and proper name, CGEL seems to define noun as a single word:

In their primary use proper names normally refer to the particular entities that they name: in this use they have the syntactic status of NPs.


Proper nouns, by contrast, are word-level units belonging to the category noun. Clinton and Zealand are proper nouns, but New Zealand is not. (Page 516)

In the following excerpt, CGEL seems to define compound noun as a single word (Page 448):

enter image description here

But CGEL also classifies a two-word unit such as full stop as a compound noun (Page 451):

enter image description here


(1) Does CGEL define the term word as a lexical unit having no space within when written or printed?

(2) Does CGEL define the term noun as a single word?

(3) Since CGEL says full stop is a compound noun, does the term compound noun as defined in CGEL include not only a single-word unit but also a multiple-word unit?

(4) If answers to the above questions are all 'yes', how can you say the term noun is a single-word unit and at the same time that the term compound noun can be a multiple-word unit?

  • Compound nouns can be open, closed, or hyphenated: full moon, notebook, six-pack. We don't mash up independent adjectives and nouns as in newcars and London-colleges – Tinfoil Hat Feb 12 '20 at 5:02
  • 4
    I'm voting to close this question as off-topic because this is a factual question about the contents of a particular book. – Colin Fine Feb 12 '20 at 11:16
  • 1
    @ColinFine It's not a factual question about the contents of the book. It's asking about CGEL's treatment of compound nouns, CGEL being the most up-to-date and the most comprehensive grammar of English. You can answer the question with an outside source. – JK2 Feb 12 '20 at 23:18
  • 1
    @ColinFine If you have to close-vote it, then please go ahead to close-vote my other question for the same reason. english.stackexchange.com/q/481884/27275 And many more similar questions specifically about CGEL. – JK2 Feb 12 '20 at 23:21
  • I've put a question querying the propriety of asking questions demanding answers restricted to the views of one particular group of grammarians. In the meanwhile, I'm close-voting here, as in my opinion it is not within the spirit of ELU, which is not 'English Language and Usage as laid down / recommended / explained by Huddleston, Pullum et al'. It's a fine reference, of course, but not unchallengeable. – Edwin Ashworth Mar 14 '20 at 14:16
  1. Does CGEL define the term word as a lexical unit having no space within when written or printed?

    No, that is not a complete definition of the term "word" as used in the CGEL. The CGEL evidently makes a distinction between the concept of an "orthographic word" and some other concept(s) of "word", as shown in your final quotation, and also in other places. From the chapter "Nouns and noun phrases":

    Like the universal personal pronouns, the reciprocals are written as two orthographic words but are single grammatical words.

    One aspect of the CGEL definition of word that bears noting (since it contradicts the definition used by Greg Lee) is that words apparently cannot contain other words in the CGEL terminological system (based on your third quote). I think this is because "words" in the CGEL are conceived as being the smallest units of syntactic constructions.

  2. Does CGEL define the term noun as a single word?

    I'm not certain. Your third quote says that in the compound noun shortbread, bread is a "base" but not a "word", and specifies that it is a base of the type "noun". Since the term "noun" apparently can be applied to the base bread, which is not a word, it seems that being a word in all situations is not a necessary part of the definition of the term "noun". Perhaps "noun" is defined as a construct that can stand as a single word, if not contained in a larger word.

  3. (3) Since CGEL says full stop is a compound noun, does the term compound noun as defined in CGEL include not only a single-word unit but also a multiple-word unit?

    I don't know. It would seem inconsistent to me to refer to "full stop" as a word, but not "New Zealand". I'm not sure why "New Zealand" is described as being more than a single word, as I think it is an atomic unit for the purposes of syntax. As far as I can tell, it doesn't pass any of the syntactic tests given in 14.4 for a composite nominal: we can't modify "New" or "Zealand" or coordinate them with other words ("New Zealand and New York" is not replaceable with "New York and Zealand" or "New Zealand and York"). So "New Zealand" seems to me like it should be categorized as a single word, a compound noun.

  4. If answers to the above questions are all 'yes', how can you say the term noun is a single-word unit and at the same time that the term compound noun can be a multiple-word unit?

    This looks like a rhetorical question addressed at the authors of the CGEL.

Your first quotation seems entirely irrelevant. That part of CGEL isn't concerned with defining the edges of words; it's about establishing the CGEL's distinction in terminology between "word" and "lexeme," and establishing that different inflected forms of a single lexeme are in the CGEL's terminology called different words.

  • In CGEL (and in other grammar), 'compound' includes 'compound noun', 'compound adjective', etc. So, yes, CGEL's saying that full stop is a compound noun. (Also see [28] on page 451.) But I don't know if CGEL's said that it's a "compound word." But in the third quote above, CGEL says ice-cream is "a compound word". Then, I think ice cream is also one in CGEL. – JK2 Feb 12 '20 at 23:46
  • Re the first quote being irrelevant, I think "the syntactically-oriented sense" of the term word sheds light on "how word boundaries can be recognized", don't you? – JK2 Feb 13 '20 at 2:43
  • @JK2: No, I don't. The first quote just says that constructs with different forms and different syntactic behavior (such as "hard" and "harder") will not be defined as the same word, even if they can be considered to be inflected forms of the same lexeme. It doesn't tell you how to determine whether a construct is a word, multiple words, or part of a word. – herisson Feb 13 '20 at 5:38
  • The most important question is no. 4. And it's not a rhetorical question. – JK2 Feb 13 '20 at 8:26

Orthography can be safely ignored x y; x-y and xy are all possible ways of writing a compound and composite nouns.

Far be it from me to suggest that Pullum and Huddleston have relied upon the “I don’t know what they are but I’d recognise one when I see it,” as the difference between a compound noun and a composite noun is clear at the extremes but is fuzzy where the two categories meet and/or overlap.

The precise distinction is thus blurred where one of the elements is capable of having a figurative meaning, e.g. (i) “dog collar” (composite) – a collar for a dog and as opposed to a horse collar or a button-down collar or (ii) a “dog collar” (clerical collar) (compound) worn around the neck of a vicar or priest.

Painting with a broad brush:

A compound noun is a number of words, the sum of whose whole is greater (or more precise, or distinct, etc.) in meaning than the sum of the parts. The nouns or adjectives comprising a compound noun are co-dependent in their formation of meaning.

A compound noun describes a unique object/concept via an imperfect description of the object’s nature. A compound noun resembles an idiom in that the simple reading of the word does not go immediately to the literal understanding/translation. For the student, a compound noun must be learned as if it were a new word. It is, in effect, a label, or a proper noun, or a name for a specific item/concept.

“Shortbread” is neither short nor bread (but see later); “ice-cream” – is not ice, other than it is cold, and is cream mainly by its consistency (have a look at the ingredients on an ice-cream carton.) “Full stop” is not full as there is no such thing as a “partial, incomplete, or empty stop” but it is a stop as used in grammar to indicate the end of a sentence, c.f., the word ‘stop’ as used in telegrams.

We can distinguish between a walking stick, a shooting-stick, and a breadstick in which “stick” is an uncertain generic until defined, albeit imperfectly, by its adjectival. In these cases, the compound noun then becomes a definer of a unique class, item, or concept.

(It is worth noting that a walking stick in the sense of “a stick for (to aid) walking” is a compound noun, but a walking stick as “a stick that also walks” – you will have to imagine a child’s story here – is a composite noun and is basically no different from “a long stick”.

The above is only valid for compound nouns in the current understanding of each element: the words may have an etymology that is quite literal, e.g. shortbread OED: short (adj.) Not tenacious in substance, friable, brittle. Bread: †1. (Only in Old English) Bit, piece, morsel (of food) obsolete.

Composite nouns on the other hand are literal: “new cars” are both new and cars; “London college” is both a college and is in London – the adjective/adjectival is independent of the noun. “Composite nouns” would probably be better called “modified nouns”.

  • The analysis of such strings (defining string as a string of 1 or more orthographic words, counting a hyphenated rendering as an orthographic word) addresses an extremely complex area. There are non-intersective strings such as 'heavy smoker' and anti-intersective strings such as 'fake diamond'. There is a fuzzy gradience from [free combination] (eg 'yellow bicycle') through [weak collocation] (perhaps 'stewed tea') and [strong collocation] ('heavy rain') to [compound noun] ('daisy wheel' etc). (There are, of course other terms and classification schemes.) But ... – Edwin Ashworth Mar 14 '20 at 12:39
  • while I think 'What does CGEL / Strunk & White / Jespersen / Blackadder ... say about this point of grammar?' questions are too narrowly scoped for ELU (and probably guilty of analysis-ism), any answer hereabouts is inadequate if not supported by relevant linked attributed references from recognised authorities (and dictionaries are inadequate here). – Edwin Ashworth Mar 14 '20 at 12:39
  • @Edwin Ashworth – OED's value cannot be overstated. and might have helped you and the answer == > compound 2. Concrete: a. A compound substance; spec. a compounded drug, as opposed to ‘simples’. Chemical compound, a substance composed chemically of two or more elements in definite proportions (as opposed to a mixture). 1887 I. Remsen Elem. Chem. (1897) i. 9 Mechanical Mixtures and Chemical Compounds.— In a mixture the substances are unchanged... In a chemical compound the substances which are in combination are completely changed. ==The parallel I make is good. – Greybeard Mar 14 '20 at 14:44
  • 'OED's value cannot be overstated' is a specious claim when one is looking at rival syntactical analyses. Or trying to find advice on solving a sudoku. And I doubt they mention the choice of terminology CGEL use, and decree unequivocably whether 'peanut butter' is a single noun, a compound noun, one word or two (whether or not their precising definitions would match those of CGEL) (which seem hard to determine anyway; hence the question). – Edwin Ashworth Mar 14 '20 at 15:06
  • @EdwinAshworth "OED's value cannot be overstated' is a specious claim when one is looking at rival syntactical analyses." I don't think so. The OED has many footnotes on usage that are authoritative as the research can be trusted. In the case in point, the OED defines "compound" in the sense that Messrs P&H failed to but intended to: this then reflects on the analysis. – Greybeard Mar 14 '20 at 16:18

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.