# What is a group of determiners called?

Verbs, auxiliaries and modals constitute the verb group (Vgp).

Is there an official term used to describe a [group of determiners] (pre-, central-, post-) that pre-modify the noun in the NP? In the following examples, more than one determiner pre-modifies the noun.

[My three] cats [Both the] cats [All the many] cats

We can't call it a determiner phrase because according to the DP-hypothesis, the determiner phrase is an alternative to the NP whereby the determiner heads the whole phrase; hence, DP. This is a whole different topic.

Can it, therefore, be called a determiner group or is there an actual term for it?

• Maybe you could include some examples and counterexamples? Dec 21 '13 at 10:13
• Oh. You're right. Sorry. Dec 21 '13 at 10:14
• Maybe the concept of an extended projection in X'-theoretic syntax is helpful here. Within this system, modals and auxilliaries are part of the verbal extended projection, and determiners, numerals etc. are part of the nominal extended projection. An extended projection consists of a lexical head and the functional structure surrounding the lexical projection. The relevant reference is the work of Jane Grimshaw, see e.g. this paper: docs.google.com/file/d/… Dec 21 '13 at 14:42
• Of course, if you don't accept the DP hypothesis, you could call these determiner phrases. (Or, if you accept Huddleston & Pullum's terms, determinative phrases.)
– user2081
Dec 22 '13 at 10:18
• @snailboat Ah yes...i didn't catch that bit in the question about the DP-hypothesis. To be honest, i don't really see how adopting the DP-hypothesis prevents one from calling these DPs. In fact, i think most people would. Dec 22 '13 at 13:19

A somewhat more useful way is to do what snailboat suggests above,
namely ignore the `DP` Hypothesis, which is after all just a hypothesis that can't be disproved,
and go for something that represents the data rather than the theory.

For instance, McCawley 1998 distinguishes between two types of nouny constituent in English:

1. `Nʹ` (pronounced "N-Bar" -- don't ask), which is a phrasal unit headed by an `N`
and
2. `NP` (pronounced "N-P"), which is the syntactic constituent type
corresponding to the logical type Argument of Predicate
(and thus is outside the X-Bar nodal system).

For instance, it's not clear that the subject of

• That he arrived late is very unfortunate.
(i.e, the complement clause that he arrived late,
which must be a noun phrase since it's the subject),

can be said even to have a head, and if so, it's clearly not headed by a noun.
It's simply an NP, since it corresponds to an argument of the predicate adjective unfortunate.

The constituent containing determiners that precedes prenominal adjectives
is then simply, in X-Bar terminology, a `Dʹ`. In general, an `NP` will contain a `Dʹ`.
For instance, the subject of the following sentence

• Quite a few of the more than ten thousand people took his advice.

would be parsable as [`NP` [`Dʹ` Quite ... thousand `Dʹ`] [N people N] NP].

Whether `Dʹ` is actually the same constituent type as `DP`
depends entirely on one's preferred nomenclatural presuppositions.
It doesn't bear on descriptions of phenomena, except to counsel one
to avoid confusing modes of expression; after all, the idea is to be clear.

• Upvoted for being very interesting. So if i understand correctly, what you call an 'NP' corresponds to semantic argumenthood? Surely that renders the terminology rather confusing, given that an 'NP' is not necessarily a 'Noun Phrase' (in the sense that it isn't necessarily headed by a noun). Why not just call them arguments? With respect to your example: "That he arrived late is very unfortunate", doesn't this just show us that CPs (or clauses, if you like) can be subjects? I don't see any compelling reason to call it an NP. Dec 22 '13 at 19:00
• @jawler one further small question - within McCawley's framework, which i'm not really familiar with, what is the distinction between an X' and an XP? [NP [Dʹ Quite ... thousand Dʹ] [N people N] NP] wouldn't be a valid X'-theoretic object, since D' labels an intermediate projection, and one can't have an intermediate projection without a maximal projection. Dec 22 '13 at 19:04
• `NP` is the only logically-based one, and the only one with a P; it's not in the X-Bar system. The only terms in the X-Bar system are (as I quoted them in Jim's memorial) "`Nʹ`, `Vʹ`, `Pʹ`, `Aʹ`, `Advʹ`, and `0ʹ` (zero-bar – ‘phrasal unit whose head belongs to no part of speech’); and distinguishing `Nʹ` from `NP` (which is not, like `Nʹ`, a phrase headed by an `N`, but rather is a different type, outside the X-Bar system – `NP` is the syntactic constituent type corresponding to the logical type Argument of Predicate)." Dec 22 '13 at 19:26
• Oh, and maximal projections and their ilk don't occur, either, so that's another thing you don't have to worry about. Dec 22 '13 at 19:28
• Thanks for the clarifications. I really should get around to reading McCawley '98. Dec 22 '13 at 21:59

The simple answer to the question is NO, there is no established term such as determiner group used to denote the word combinations in the question, and there is a reason why no term has been established to denote such word combinations. This reason is that the relevant words (e.g. my three, both the, all the many) hardly qualify as a unit of syntax. At most, one can acknowledge that they form a string.

Such word combinations qualify as a constituent in no theory of syntax (that I am aware of), not in phrase structure grammars and not in dependency grammars. Furthermore, such word combinations do not qaulify as chains, that is, they do not qualify as catenae: https://en.wikipedia.org/wiki/Catena_%28linguistics%29. In contrast, verb predicates (e.g. will have fixed, has been fixed, may have been being fixed) do qualify as chains of words, as catenae. The fact that determiner strings qualify neither as constituents nor as catenae means that it is more difficult to denote such combinations with a particular syntactic term. Theoretical syntax tends to focus on word combinations that are recognizable units.

But there is nothing to stop one from establishing a new term, e.g. determiner string or determiner group. In fact, since there is no pre-established term for such combinations, one has more freedom to call them whatever one wants.