>Matt Pearson wrote:
>> The English article "the" appears to fit definition (1), but not the
>> others.
>
>I disagree.  Two and three fit "the" as well.  It forms a tight
>morphological bond with the following word.  It's attached to the
>beginning of noun phrase.

The article does form a tight *phonological* bond with the following
word, in the sense that it belongs to the same word-level stress
unit as the following word, but it does NOT form a *morphological*
bond with the following word, in the sense that the article doesn't
care what the category of the following word is (it can be a noun,
adjective, whatever).  Contrast this with object clitic pronouns in
French:  These must occur right in front of the verb, and can only be
separated from it by a small number of particles, suggesting that
they are somehow selectionally dependent on the verb, in the same
way that the past tense suffix "-ed" is dependent on the verb in
English.

Also, technically speaking, articles are not attached to the beginning
of the noun phrase.  They are *internal* to the noun phrase, albeit at
the left edge of the constituent.  Clitics like possessive -'s in English,
by contrast, are *external* to the noun phrase (at least on some
analyses).

Matt.