Possessives

[Larry Trask]

On Mon, 27 Sep 1999, Patrick C. Ryan wrote:

[LT]

>> OK; some facts.  First, Beekes does not use the term `possessive
>> pronoun' at all in the passage cited by Ryan: he uses only the term
>> `possessive', which no one can object to.  Hence Ryan's rather snide
>> comments are pointless.

> [PR]

> Well, fact: Beekes does not use "determiner";

Indeed, but not relevant.  You were maintaining that Beekes agreed with
your characterization of items like `my' as possessive pronouns.  But he
does not.

> fact: "possessive" is
> defined in AHD as: "of, pertaining to, or *designating* a noun or
> pronoun case that expresses belong or other similar relation".

Like most general-purpose dictionaries of English, the AHD is presenting
a very old-fashioned view of English grammar.  In fact, the AHD is
perhaps slightly better at defining grammatical terms than are some
other desk dictionaries, but its definitions are still, in general, not
acceptable for linguistic purposes.

> [LT]

>> Second, Beekes is talking about PIE, while I was talking about
>> English.

> [PR]

> So what?

"So what?"?  Well, Mr. Ryan, it may come as a surprise to you that the
grammatical facts of all languages are not identical.

In English, possessives like `my' are clearly determiners, and not
pronouns.  But the same is not true of all languages.

In Basque, for example, possessives like <nere> `my' exhibit *none* of
the properties of determiners and cannot be regarded as determiners.
Instead, the Basque items are case-marked NPs.

The grammatical status of possessives varies widely among languages, and
what is true of PIE need not be true of any other language.

> [LT]

>> Whatever may be the case in PIE, or in any other language, the facts of
>> English are clear: words like `my' and `your' are not pronouns, but
>> determiners.  Possessive determiners, of course, but determiners.
>> This is easy to see, using a frame for noun phrases:

>> ___ was nice. (singular); ___ were nice. (plural)

>> Real pronouns can go into these blanks to make good sentences: She was
>> nice; It was nice; They were nice; Something was nice; Nothing was nice;
>> That was nice; and so on.  This is also true for the *real* possessive
>> pronouns in English: Mine was nice; Ours was nice.

>> But it doesn't work with the determiners: *My was nice; *Your was nice;
>> *Our was nice.

> [PR]

> Possessive pronouns (BT = Before Trask)  have two forms: an
> adjectival use: 'her', etc. and nominal use: 'hers'.

Also wrong, though indeed once widely believed.

Items like `my' have no adjectival properties and cannot be classed as
adjectives.  Look at two adjective frames:

	the ___ book
	That book is ___.

These slots accept adjectives, like `big', `red', `new', `expensive',
`dirty', `interesting' and `terrible'.  But they don't accept items like
`my':

	*the my book
	*That book is my.

Instead, `my' goes into slots like this one:

	___ new book

Among the other words that go into this slot are `the', `a', `this' and
`some': in other words, the class of determiners.  QED.

> In casual speech, one might hear: 'My one was nice'; and 'His was
> nice'.

The second is perfectly normal for most of us.  The first is a
regionalism, normal in some varieties but unknown in others (including
my own, where it is impossible).  But neither is a problem for my
analysis.  Let's use `my' and `mine', since `his' does not distinguish
the two forms:

	My one was nice.

Here `my' is again a determiner, just as in `This one was nice'.
And `one' is doing its usual job of being an N-bar.

	Mine was nice.

Here `mine' is a true pronoun, as usual, just as in `That was nice'.

> Of course, *mine* ears have heard a certain amount of overlap
> between forms used for each of these two major employments.

Well, `mine ears' is now obsolete in most varieties of English.  The
facts of modern English are different from the facts of earlier English.

> Now, if we say 'His was nice', the 'his' stands for a possessive N
> like 'John's'. The 'his', or 'her(s)', must have a nominal referent;
> and it stands for ('pro') this nominal referent.

The interpretation is not relevant.  When we are doing grammatical
classification, we must look at the grammatical behavior.  And `my'
behaves like a determiner, not like a pronoun.  It is semantically
*related* to a pronoun, of course, but that's a different matter.
The noun `arrival' is not a verb because it's related to the verb
`arrive'.

> Now I have no great objection to terming "her" a "possessive
> determiner" but using this terminology eliminates the interesting
> connection with pronouns, which I find superfluously
> disadvantageous.

Well, does applying the label `noun' to deverbal nouns like `arrival',
`decision' and `creation' also seem to you "superfluously
disadvantageous"? ;-)

These things are nouns because they behave like nouns, even though they
are related to verbs.  And items like `my' are determiners because they
behave like determiners, even though they are related to pronouns.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk