I have a question regarding attributive nouns, or noun-noun compounds, and how they are integrated into syntactic rules for NP formation. Typically, the rule given in textbooks for forming a NP is the following:

NP --> (D)(AdjP+) N (PP+)

However, I am trying to figure out how the noun-noun compounds can fit into this. Examples of these include "chocolate chip cookie dough", where "chocolate chip" and "cookie" are both nouns, but here are acting adjective-like in how they modify "dough". More examples include "grandfather clock", "coffee break", and "polka dot sweater". However, it feels wrong to simply call these all adjectives. How do we go about integrating these (what I consider NPs) into the NP rule? Would they act as adjuncts, just like adjectives do when modifying a noun (so NP --> (D)(AdjP+)(NP+) N (PP+))? Any help on this problem would be GREATLY appreciated!

  • You're confusing 'compound' and 'composite nominal'. Your examples are syntactic constructions -- nominals consisting of noun head + noun modifiers(s). A compound is a single word formed by the morphological combining of two or more bases like "greenhouse" or "palm-tree". But "Grandfather clock", for example, consists of "clock" as head and "grandfather" as modifier. It is most certainly not a compound word! Also, an adjective that modifies a noun is not an 'adjunct', but a modifier. Adjuncts are elements in clause structure, not phrase structure.
    – BillJ
    May 23, 2017 at 8:37
  • The salient analysis of your example is this: "chocolate" is modifying "chip" to give the nominal "chocolate chip", which in turn is modifying "cookie" to give the larger nominal "chocolate chip cookie". This in turn is modifying "dough" to give the even larger nominal "chocolate chip cookie dough".
    – BillJ
    May 23, 2017 at 15:29
  • @BillJ Out of curiosity, why don't you consider 'grandfather clock' as a compound? Semantically, they don't seem to have anything to do with grandfathers, so in this regard it seems more like 'black bird' than 'blackbird'. Phonologically, it seems ambiguous: Brits pronounce it as grandfather CLOCK (suggesting it's not a compound) but Americans pronounce it as GRANDfather clock (which seems to suggest it's a compound), according to the CEPD. May 23, 2017 at 17:39
  • Yes sorry, I got ahead of myself in my question, I would consider "polka dot" in "polka dot sweater" and "chocolate chip" in "chocolate chip cookie dough" to be the compound words (not the entire phrase). My question was getting at how you combine these with other nouns in a syntactically-sound way. May 23, 2017 at 18:02
  • I think in your example, "polka-dot" should be treated as a single lexeme, a compound noun (note the hyphen), since it fails all the tests for a composite nominal. But not "chocolate chip" since it is a syntactic construction consisting of a head ("chip") and a modifier "chocolate", not a morphological compound comprised of two bases. Evidence of this is provided by the fact that we can say "chocolate and caramel chip", or "plain chocolate chip". That is the crucial difference
    – BillJ
    May 23, 2017 at 18:45

1 Answer 1


Noun-noun compounds are nouns: N -> N N. The structure of your example is

[N [N [N [N chocolate] [N chip]] [N cookie]] [N dough]]

or possibly

[N [N [N chocolate] [N chip]] [N [N cookie] [N dough]]]

The first means "dough for making chocolate chip cookies"; the second means "cookie dough flavored with chocolate chips" -- a subtle culinary ambiguity.

The interpretation of noun-noun compounds is very free and can be adjectival. In your other examples, there are no adjectives -- the parts are all nouns. You can test that by trying to add modifiers to the parts. If you can add an adverb, it must be an adjective, since adverbs can't modify nouns; if you add an adjective, the resulting N-bar cannot be part of a noun-noun compound, since those are built from nouns, not N-bars. (Does a "small polka dot sweater" have small polka dots?)

  • Thank you so much for shedding some light on this! You've helped to validate my own thoughts on this (which is always nice). I have one other question, if you don't mind me asking. Let's take your 2nd bracket diagram, and work with that. Would there be any internal NPs within this phrase? For example, would it be accurate to rewrite it as follows: [NP [NP [N chocolate] [N chip]] [NP [N cookie] [N dough]]]? I am trying to figure out if it's okay for a NP to take on adjunct noun phrases like this, since this is never mentioned in the rules. May 23, 2017 at 17:03
  • I don't quite understand your question. My proposal for this example is just as I gave it: there are no NPs involved, in my opinion. It's a compound N, made up of Ns some of which are themselves compound nouns. The rule N -> N N is a recursive rule, which generates an infinity of compound Ns which have an infinity of different internal structures. Just one phrase structure rule. It's remarkable. And all those structures occupy just the one N slot in a NP, which has the structure provided by the rule NP -> Det N. You're still not on the right track.
    – Greg Lee
    May 23, 2017 at 19:21
  • Okay, I think I understand what you're saying. My only reason for bringing up the NP is because, in my syntax textbook, the only clue I had towards implementing these elements into tree structures was one that used the example "leather shoe", which it said could be diagrammed as an NP containing an ADJP, but the more agreed way of doing it was as follows: [NP[N leather shoe]. On that note, since this was labeled as an NP (not a N), I was wondering if it was okay to attach more NPs. For instance, "leather shoe maker" --> [NP[[NP[N leather shoe]][N maker]]]. That was my reasoning, at least. May 23, 2017 at 19:48
  • Perhaps I'm just getting mixed up between noun-noun compounds, and nouns that are being modified by another noun (as BillJ above mentioned)? So N-N compounds would follow the N --> N N rule you discussed, whereas the latter would follow a different structure (and this is ultimately what I'm trying to figure out). May 23, 2017 at 19:56
  • "Leather shoe" might be a compound noun, but I'm not sure. The acceptability of "fine leather shoe" is a problem, since ordinarily a noun modified by an adjective cannot be part of a compound noun.
    – Greg Lee
    May 23, 2017 at 22:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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