"Pro-drop" with nominal predicates

Hans-Juergen Sasse Hj.Sasse at UNI-KOELN.DE
Sun Dec 6 16:43:38 UTC 1998

This is a delayed reaction to Edit Moravcsik's remarks on "pro-drop" with
nominal predicates in Hungarian. We have been out of communication for a
couple of days, so this is probably obsolete since there was no further
contribution this topic. We nevertheless thought we should re-open the
discussion because as it stands Edit's description seems to be misleading and
calls for further details.

MORAVCSIK Edit wrote:

>       In the wake of David Gil's "Knock-knock" query, the issue that has
> come up is whether in languages where subject
> pronouns can be omitted when the predicate is a verb, subject pronouns
> can or cannot be omitted when the predicate is nominal (or adjectival).
> Hans-Juergen Sasse pointed out that in Hungarian, the subject pronoun can
> be dropped equally well with verbal and nominal predicates; thus, both in
> sentences such as 'He is waiting.' and in sentences such as 'He is a
> teacher.'I think this is not quite so in all cases. Pro-dropped sentences
> such as "Teacher." - to mean 'He is a teacher." - or "My father." - to
> mean 'He is my father.' - can indeed be used but only in special
> situations such as when introducing somebody and the introducer is
> providing prominent, characteristic information about the person that is
> being introduced. A gesture pointing at the person is also required. But
> in normal discourse, a subject would also be used. Thus, if I begin to
> talk about Ja'nos ('John'), I cannot put in a sentence like "Teacher." -
> to mean 'He/Ja'nos is a teacher.'; a subject ("he" or "Ja'nos") would
> be needed. However, a sentence with a verbal predicate - such as "Came
> from New York." - to mean 'He came from New York." - would be OK as
> a continuation of the discourse about Ja'nos. Thus, in normal discourse,
> pro-drop is indeed an option with verbal predicates but with nominal ones
> it isn't.

Some clarification is certainly necessary here. Edit correctly states that in
Hungarian a sentence with a verbal predicate without a pronoun such as „Came
from New York." to mean „He came from New York." would be okay as a
continuation of the discourse about a subject, say Ja’nos, while if I begin
to talk about Ja’nos I cannot put in a sentence like „Teacher.". Since two
different discourse positions are compared here, it is legitimate to ask what
happens with verbal predicates in text-initial positions and with nominal
predicates in the case of the continuation of the discourse about a subject.
The answer is that there is no difference, in principle, between nominal and
verbal predicates when compared in identical text positions. As in English
you cannot normally begin a discourse out of the blue by saying "He came from
New York.", it would be infelicitous to begin a discourse in Hungarian with
"Came from New York." because the listener would immediately ask back: "Who?"
The conclusion is that there are certain contexts which require overt
subjects (such as the introductory context or the deictic context), while in
the anaphoric context the null subject is the normal way of expressing
continuation of the discourse about a subject. This holds equally true for
verbal and nominal predicates. More precisely, the null subject with both
nominal and verbal predicates is the conventionalized formal means of
Hungarian grammar to express an anaphoric relation to a certain established
discourse referent in the third person present, provided that it is not
contrastive or focal (in this case, pronouns would be added, see below).
Consequently, in order to use an expression like "Teacher." or "Came from New
York.", the subject understood must refer back to an entity anchored in
previous discourse. This not necessarily implies previous mention; to a
certain extent it includes cases of associative anaphora and switch reference
(as the examples given below show), i.e. all cases where a discourse referent
has been introduced which may serve as an anchor for an anaphoric
relation.The systems of nominal and verbal predication are entirely parallel
in this respect. The paradigms in both cases provide overt forms for all
persons in the past and for 1st and 2nd persons in the present, while the 3rd
person present is left "unmarked". The addition of pronouns is quite a
different thing; with both verbal and nominal predicates it expresses certain
discourse-semantic components such as deixis, contrast or focus. It does not
serve the purpose of keeping any formal requirement on the structure of
utterances (obligatoriness of overt subjects, etc.) intact. It is therefore
not comparable with the Hebrew and Arabic cases, where the pronouns in
nominal sentences have some sort of copula-like function and at the same time
keep the bipartite structure of nominal sentences - a language-specific
principle of Arabic grammar - intact (cf. Arabic huwa 'ana 'he me' = 'it's
Edith's statement that "pro-dropped sentences such as "Teacher." are
restricted to special situations" is therefore somewhat misleading. The
anaphoric use of bare predicate nouns is so common in Hungarian that you will
find numerous examples even in a short text passage. Below we present two
examples from the novel "Zakaria's" by Kerte'sz A'kos. Both of them provide
examples of subject change. Moreover, in (1), a case of associative anaphora
can be observed (Unicum is a Hungarian herb-flavored liqueur), and (2) shows
that an elliptic interpretation is ruled out because of the negative particle
nem (hogy nem is gyerek 'that not child' = 'that it is not a child'). Note
that in (2) the verbal and nominal predicates bearing the anaphoric relation
to the old man (nem ja'tszik 'he isn't playing', ta'ntorog 'he's tottering',
pocsolyare'szeg 'he's dead-drunk') are exactly parallel: no pronominal affix,
no free pronoun. The relevant points are marked with *...*

(1) Ma'rta beleszagolt a pohara'ba, *Unicum*, a'llapi'totta meg...
Ma'rta smelled into the glass, *(it is) Unicum*, (she) ascertained...

(2) A gyerek ja'te'kosan kacsa'zott, nagy fe'lkoeroeket i'rva szlalomozott a
ja'rda egyik sze'le'to"l a ma'sigik, de
The child playfully acted-like-a-duck, big semicircles writing slalomed the
sidewalk one edge-from the other-to, but
ahogy koezelebb e'rtek, kideruelt, hogy *nem is gyerek* , hanem *egy
alacsony, tagbaszakadt oeregember* u'gy
when closer arrived, turned-out, that *not at-all child*, but *a small,
thickset old-man* so
ha'tvan e's he'tven koezoett, e's nem ja'tszik, hanem ta'ntorog, ugyanis
sixty and seventy between, and not plays, but totters, namely *dead-drunk*.

'The child playfully walked like a duck, describing big semicircles he
slalomed along the sidewalk from one edge to the other, but when they came
closer, it turned out that *he is not a child* but *a small squat old man*
between sixty and seventy or so, and that he is not playing, but tottering,
because *he is dead-drunk*.'

In Hungarian there is no "consecutio temporum"; all predicates of the
that-clause after 'turned out' are in the 3rd person singular present.

Leila Behrens
Hans-Juergen Sasse

More information about the Lingtyp mailing list