LINGUIST List 4.180

Sat 13 Mar 1993

Sum: Pro-drop Languages

Editor for this issue: <>


  1. Mike Maxwell 6369, Pro-drop languages: a summary

Message 1: Pro-drop languages: a summary

Date: 12 Mar 1993 14:50:00 -0600Pro-drop languages: a summary
From: Mike Maxwell 6369 <>
Subject: Pro-drop languages: a summary

This posting summarizes the replies people to have sent to me on
the question
> What languages of the world are NOT null subject languages?
Thanks to all those who replied. I was rather surprised at the
small number of replies; perhaps that should be taken as an
indication that there aren't many such languages.

I've edited in some cases to combine msgs where the responder and
I corresponded back and forth, hopefully without doing any
violence to what they said. Internal comments from me are in []
with my initials (MM).

To give background where I corresponded with someone who had
posted to Linguist, on occasion I'll cite part of a message that
was posted to LINGUIST.

Chris Culy (
...4 of the 5 W. African languages I know something about are not
null subject languages: Bambara (Mande), Fula ((West) Atlantic),
Moore (Gur), Songhai (Nilo-Saharan or isolate). Dogon (isolate or
Gur) is a null subject language. All but Moore have large numbers of
speakers in Mali, while Moore is spoken primarily in Burkina Faso.

Larry G. Hutchinson (
Temne (Sierra Leone) requires a subject pronoun always, even if there
is a nounal subject as well. There was an attempt in 1961 to treat
this subject pronoun as an agreement particle, which required that
all Temne pronouns, which were never observed, be obligatorily
deleted after the verbs were made to agree with them. A ridiculous
solution. How many years did this predate "pro-drop"? [A fascinating
language. It's [-pro-drop] with a vengeance! Are there other
languages like this? -MM]

Matthew Dryer <LINDRYERUBVMS.bitnet>
Gary Gilligan completed a USC dissertation on the topic in 1987 ("A
Cross-Linguistic Approach to the Pro-Drop Parameter"), based on a
sample of 100 languages. Of these 100 languages, only 7 do not allow
null subjects in finite clauses, and these 7 include one
Indo-European language and 2 Indo-European-based Creoles. (The other
four include two Niger-Congo languages, Tagalog, and Guaymi)... Gary
Gilligan is no longer in the field and does not have email. [I
wonder if someone would comment on Tagalog? As I understand, it's
controversial whether Tagalog even HAS subjects, which may confound
the question of whether it allows null subjects. -MM]

George Aaron Broadwell
I've done some fieldwork on Santa Ana del Valle Zapotec, which is not
a pro-drop language. Pronominal subjects cliticize to the verb,
however. Some data:

Gu-naa Juany Maria. 'Juan saw Maria.'

Gu-naa-bi Maria. 'He saw Maria.'

*Gu-naa-bi Juany Maria. (John he saw Maria.)

*Gu-naa Maria (saw Maria.)

I think the 3rd example shows that /-bi/ is a real pronoun, and not
some kind of verb agreement. The 4th example shows that a subject
pronoun can't be omitted.

Maria Vilkuna ( [I'm afraid this gets
long, but it points up the difficulty noticed in recent postings
as to deciding whether a language is pro-drop -MM]
The general wisdom is that Finnish drops (unstressed) 1st and 2nd
person subjects, but not 3rd person ones... So, "is" Finnish a
null-subject language?...

It is quite easy to distinguish between tensed and non-tensed verbs
in Finnish. [so these sentences cannot be analyzed as bare VPs in the
same way as English imperatives can -MM] And 1st and 2nd person
pronouns do drop from subordinated clauses, I assume equally easily
as from root clauses. In the formal varieties of standard Finnish,
where all this is common, 3rd person pronouns would only be omitted
if coreferential with the matrix clause subject; that is, (c) below
is unlikely with the 0 subject.

(a) Kaikki luulevat, ett!a (min!a) osaan englantia.
 all think-3pl that (I) can-1sg English
 'everybody thinks that I can speak English'

(b) (Min!a) luulen, ett!a (min!a) osaan englantia.
 I think-1sg

(c) (Min!a) luulen, ett{ h!an / *0 osaa englantia.
 I think-1sg that s/he /*0 can-3sg English

(d) H!an luulee, ett{ (h!an) osaa englantia.
 s/he think-3sg that s/he can-3sg English

(!a = a with two dots, !o = o with two dots)

[The following is in reply to my question about explitive subjects;
see also recent postings to the effect that a language is [-pro-drop]
iff it has explitive subjects -MM] Expletives! What a fascinating
topic. Officially, they don't exist. Unofficially, in colloquial
speech and dialects, they abound, but they are never obligatory. Want
to hear more..? In fact I wrote some preliminary observations on
place-fillers in my PhD thesis, which exists, as a book, in English.
(Free Word Order in Finnish, 1989)

I'm afraid there is very little information on Finnish pro-drop
available, apart from the standard wisdom concerning 1st and 2nd
versus 3rd person. This is probably to be found in any reference
grammar, of which the most recent ones are: Fred Karlsson 1983:
Finnish Grammar, and Sulkala & Karjalainen 1992: "Finnish", in the
Routledge Descriptive Grammars series. Anne Vainikka's University of
Massachusetts dissertation, Deriving Syntactic Representations in
Finnish (1989), is perhaps the most explicit, recognizing the formal
vs. colloquial distinction, (1) versus (2) in my posting.

Vainikka, by the way, argues against functional explanations of the
3rd/non-3rd asymmetry, although a case could be made on the basis of
the non-3rd inflections being more "specific". The retreat of
pro-drop in colloquial Finnish could be related to the diminishing
status of agreement inflection in most dialects; cf. the present
tense forms of the verb sano- 'say':

 standard Finnish colloquial Finnish (my

1sg sano/n sano/n
2sg sano/t sano/t
3sg sano/o (vowel lengthening) sano/o
1pl sano/mme sano/taan 2pl sano/tte
3pl sano/vat sano/o

In most colloquial variants, the number distinction is usually not
made in the 3rd person; the form used with the 1pl subject is
actually identical to the impersonal passive (like "on" + verb in

That agreement inflection does have some role here is shown by the
common dialectal pattern I mentioned in (3), something like the

 ne kulki polkua pitkin. tulivat sitten yhteen taloon
 they walked(3sg) path along. came-3pl then one-to house-to
 'they walked along the path. They came to a house'

That is, the pronoun is much more likely to drop when the appropriate
inflection appears on the verb, or vice versa. (This MIGHT be a dying
dialect, though.)

One thing that clearly must be mentioned, as it probably affects the
person asymmetry (or: might be able to exist because of the
asymmetry), is the extremely common use of Generic/Arbitrary 3sg null
subjects. That is, patterns like the following:

 Siihen tottuu
 to-it gets-used(3sg)
 'one gets used to it'

 Jos juo liikaa niin lihoo.
 if drinks(3sg) too much so gains-weight(3sg)
 'If you drink too much, then you gain weight'

Although word order facts tend to differ a bit (sorry, this is
getting complicated), things would certainly get confusing if the
same forms would be used for anaphoric/deictic 3rd person reference.

What I am not able to substantiate is my claims about the "stylistic"
uses of pro-drop in what have been taken to be non-pro-drop contexts.
I'd like to do it, but it would take a lot of time; intuition is not
enough here.

Mike Maxwell
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue