<< >> TOC

9. Overt definite determiners

If the assumptions on PF- vs. lexical saturation are correct, it may t first glance seem surprising that there are languages in which definite determiners cooccur with unmodified proper names. German and Portuguese are such cases:

(29)

German:

a.

(Die) Maria hat (den Johannes) angerufen.

the-fem Mary has the-masc John called.

‘John called Mary today’.

Portuguese

b.

A Maria telefonou ao João.

the-fem Mary called to-the-masc John.

‘Mary called John.’

In the literature on German, this usage of the definite article is often brushed aside as a merely optional or stylistic surface phenomenon: “a norm without any further communicative basis” (Bisle-Müller 1991: 120). Grimm (1989: 32), however, points out that the definite article stresses the familiarity of the discourse referent. Consider the following:[1]

(30)

a. Ich habe gestern mit der Uschi KARTEN gespielt. *Ich glaube, du kennst sie nicht.

I have yesterday with the Uschi cards played. I believe you know her not.

‘yesterday, I played cards with Uschi. I believe you don’t know her.’

b. Ich habe gestern mit der USCHI Karten gespielt. Ich glaube, du kennst sie nicht.

c. Ich habe gestern mit Uschi KARTEN gespielt. Ich glaube, du kennst sie nicht.

d. Ich habe gestern mit USCHI Karten gespielt. Ich glaube, du kennst sie nicht.

Even though my own intuition can hardly be taken as hard evidence for a pragmatic fact, it seems to me that a proper name can only introduce a discourse referent that is new to the listener if it occurs without the definite article (30c,d) or if the name is stressed (30b).[2] Thus, the definite article is optional insofar as it is not required or prohibited syntactically, but it is not true that there is no difference in meaning between names with and names without article. The meaning supplied by the definite article is exactly the same as posited by Heim (1981) for definite descriptions, i.e. the definite article marks familiarity as usual. To be precise, it marks familiarity for the listener, i.e. it clearly anchors the name in a domain of discourse shared by speaker and listener.

In German, familiarity is impolite. Politeness markers in German reflect a social distance between speaker and discourse referent (or addressee), i.e. the lack of politeness markers or the existence of overt familiarity markers will be interpreted as rude, if it seems that an existing social gap is purposely ignored. The fact that the definite article carries meaning for those speakers who employ it “optionally” becomes very apparent in the following context:

(30-1)

Beim gestrigen Gipfeltreffen trafen sich der Kohl und der Jelzin.

At-the yesterday’s summit met REFL.the Kohl and the Jelzin.

‘Kohl and Jelzin met at yesterday’s summit.’

Even speakers who regularly use the definite article with proper names would cringe if this sentence was uttered by a TV news anchor. The definite article here would exhibit a lack of polite distance towards heads of state that would certainly be regarded as inappropriate. Of course, one may object that this is a matter of style, i.e. that this usage simply does not fit the diction commonly employed by news anchors. I would maintain, however, that the news anchor style is the way it is for a reason, i.e. the stylistic choices of objective reporting are determined by the pragmatic implications of the definite article in this case.

Notice that such pragmatic implications vary from language to language. While Spanish allows the definite article with proper names in informal speech with similar connotations as in German[3], the closely related Portuguese uses the definite article in formal standard language as a politeness marker. This cross-linguistic variance would be hard to capture for theories that assume a separate categorial status for proper names and try to derive the surface syntax from underlying features introduced by the lexical items.

Besides its function as a familiarity marker, the definite article also carries overt case in German, while proper names do not (bare nouns are never marked for case in German). This leads one to expect that the definite article occurs whenever a proper name needs to be overtly marked for case. This is born out by the facts, as the following examples of scrambling structures show:

(31)

a. Ich habe Peter Maria vorgestellt.

     I    have Peter Mary introduced-to.

     ‘I introduced Mary to Peter’.

b. *Ich habe Maria Peter vorgestellt.

     ‘I introduced Mary to Peter’.

c. Ich habe die Maria Peter vorgestellt.

          ...     the-fem-Acc Mary ...

d. Ich habe Maria dem Peter vorgestellt.

... the-masc-Dat Peter...

e. Ich habe die Maria dem Peter vorgestellt.

          ...the-fem-Acc Mary the-masc-Dat Peter...

Sentence (31a) shows the default word order. The scrambled order in (31b) cannot receive the same interpretation, since the nominal arguments cannot be properly associated with the respective q-role without appropriate case marking. As soon as at least one of the arguments is overtly marked for case (31c-e), the sentence is interpretable with the target meaning.

There are at least two ways to interprete these data: the first would be to say that all sentences in (31) are syntactically well-formed, i.e. the ungrammaticality of (31b) is entirely due to the pragmatic limitations to ‘trace’ the scrambled item back to their q-position without overt case. Alternatively, one could argue that the syntactic derivation of (31b) is again well-formed, but that a crash at PF occurs when morphological case fails to be realized on the bare noun adjoined to D. At this point, I cannot make an insightful decision.

Whatever the exact analysis of (31b) turns out to be, it involves that proper names are wrapped in a DP, while the phonological representation of the head D is dependent on a variety of other factors. The assumption of such a wrapper phrases is not new to syntactic theory. Even languages that completely lack determiners are often assumed to have wrapper phrases, such as the case projection KP in Japanese (cf. Takano 1996). In my view, all these wrapper projections are really functionally equivalent. Languages differ only in the exact features that find overt (i.e. phonological) expression within this phrase. In English, only definiteness is overtly marked, in Japanese it is case, and in German, both happen to coincide in one morpheme, namely the determiner.

Crosslinguistically, there is thus plenty of evidence for the syntactic equivalence of proper names and common nouns. There are differences in syntactic realization, but these are due to language-specific pragmatic or syntactic conditions influencing the exact realization of the determiner. The main reason for the widespread belief in the determinerlessness of proper names may be found in the fact that the overwhelming majority of influential semanticists and logicians wrote and thought in (formal standard) German and in English, i.e. in dialects in which the determiner usually remains covert.

<< >> TOC

[1] I happen to come from a region where this usage of the definite article with names is the (informal) standard (Köln-Düsseldorf-Neuss, Northrhine-Westfalia), so I can also speak with the authority of an informant here (but, while I try to stay on the conservative side in my generalizations, a survey on a larger scale is certainly warranted).

[2] If, as is often assumed, phonological stress correlates with syntactic focus, and if a focussed element undergoes LF-movement into a focus position, one may argue that the name in (23b) can be interpreted as if it was not in the domain of the definite article, which makes a novelty interpretation become possible. While the exact analysis of focus is not relevant here, I just wanted to point out that (23b) is not necessarily a counterexample to the assumption that the definite article marks familiarity.

[3] I thank Ana Wells for pointing this fact out to me.