LINGUIST List 12.1725

Tue Jul 3 2001

Sum: Typology of Multiple Questions

Editor for this issue: Lydia Grebenyova <>


  1. Ralf Vogel, Typology of Multiple Questions

Message 1: Typology of Multiple Questions

Date: Tue, 3 Jul 2001 14:06:52 +0200
From: Ralf Vogel <>
Subject: Typology of Multiple Questions

Dear colleagues,

A couple of weeks ago I posted a query to LINGUIST. I was afraid not
to receive many answers, and I really didn't. Nevertheless, for those
of you who are interested in the results, here is the summary. 
The original posting was the following:

I'm interested in the typology of the syntax of questions
and would be very happy, if anybody could help me in
answering the following question:

The syntax of interrogative clauses varies in several ways
among languages. First, there are languages that have
wh-movement, like Engish, and languages that don't have it,
like Korean (for ease of reading, I'm only giving glosses):

1 a English: What did John eat _?
 b Korean: John what ate?

Second, many languages can have multiple wh-elements within
one clause, but there are also some that cannot, Italian is
supposed to be such a language. However, Italian is a
wh-movement language. I haven't found yet information about
*wh-in-situ languages that cannot form multiple questions*,

2 a English: What did you do _ where?
 b Italian: *What did you _ where?

3 a Korean: you what where did?
 b ??????: *you what where did?

Does anybody know a language like 3b, i.e., a wh-in-situ
language that has no multiple questions? I would also be
very happy about hints to work dealing with this question. I
will post a summary of the answers to the list, if there are
any interesting results. Thanks in advance.

- ----------

The most interesting answer came from Catherine Rudin:

>From Thu May 24 16:08:48 2001

"Omaha-Ponca (Siouan) may be a language of "type 2b". It has wh in situ (in
fact, normally before the verb, much like your Korean example) but as far
as I can tell it does not allow multiple wh. I am interested in multiple
questions myself, so while doing field work on Omaha-Ponca I tried to
elicit questions like "who hit who?" and "who likes what?", with no
success. My informants cheerfully gave me translations, but they were
always either multi-clausal or else left out the second wh: "Who got hit
and by the way, who hit him?" "Who likes all these things?" -- responses
along those lines. I pushed rather hard and never got a true multiple
 On the other hand, several factors make me reluctant to claim
absolutely that Omaha-Ponca has no multiple questions.
 (1) Omaha-Ponca is spoken only by elderly people, and only in fairly
restricted circumstances. The speakers I worked with were very fluent, but
all used English for most purposes in their daily lives and might not
control the full repertoire of syntactic constructions that earlier
generations of monolingual speakers had.
 (2) Multiple questions are hard to elicit anyway, in any language.
I've had speakers of Bulgarian (and even English) object to multiple
questions in elicitation ("no one would say anything like that") and then
the same speaker will spontaneously use a multiple question in

- ------------------

Omaha-Ponca would be the only language I heard of that displays such a
property. So I hope, Catherine will keep on exploring that language
and inform the public :)

Manideepa Patnaik ( pointed me to the
following pieces of work:

"Srikumar's thesis on Wh-movement in Kannada is a very good source of
information on how Wh-phrases behave in a Dravidian language that is
Wh-in-situ. I am sure you must have looked at Anoop Mahajan's MIT
thsis that was submitted in 1990 on A and A bar movement in HIndi."

I haven't found Srikumar's thesis yet, so I don't know whether it is
relevant here. Hindi is a wh-in-situ labguage, but it has multiple
wh-in-situ. And what I found about Dravidian languages points to
the same direction.

- -------------------------------

Nathan True suggested that French is sort of what I am looking for:

>From Thu May 24 08:06:26 2001

I'm sure you've already received some fine answers to your question, 
but I thought I might contribute anyway. I work with French, which is 
not typically called wh-in-situ, but which does display it:

T(u) es qui?
you are who
"Who are you?"

Tu vas o�?
you go where
"Where did you go?"

T(u) as fais quoi?
you PAST do what
"what did you do?"

Among many other wh-formations, French displays minimal pairs such as:

Tu vas o�?
O� tu vas?

but does not seem to have the same semantic difference as the same 
pair in English would. The first would be an echo question or 
emphatic and would have to have different stress in English but not 
in French.

So, while French does show in-situ wh-elements, it does not allow, as 
far as I know, constructions like the following.

*T'as fais quoi o�?

Nathan True
French and Italian Department
University of Texas at Austin

- ------

Our French informants told us that clauses like the last one mentioned
by Nathan are not that bad. Furthermore, clauses like the following one
are perfect:

qui a donn� quoi a qui?
who has given what to whom

Here, we have three wh-elements, two of which are in-situ, one of them
is fronted. So, I'm afraid that French is not the case we are looking
for. In addition to that, Fabian Heck (
quoted the following example from Boskovic's NELS 30 paper:

(i) il a donne quoi a qui?
 he has given what to whom

This is the perfect counterexample - which is again confirmed by our

- ---------

Hyo Sang Lee informed me that the order of adverb and direct object
should be reversed in the 'glosses' I gave in my query above to yield
good Korean, though the order object-adverb is not ill-formed, but
just restricted to specific contexts - sorry, I should have known
that, or at least guessed, because German is very similar in this:

- ---------
>From Tue May 29 16:36:07 2001

 "As a native speaker of Korean, however, I'd just like to say
that Korean, a WH-in-situ language, too, has some restrictions. Your
example #3a would be appropriate only in some contexts, and sould odd
in other contexts. In general, 'you where what did?' would be most
natural rather than 'you what where did?':

	[You see an acquaintance after a long while]
	Ku-tongan eti-se mwe ha-sy-ess-e-yo?
	* Ku-tongan mwe eti-se ha-sy-ess-e-yo?

You can say: 
	mwusun hakkwi-lul eti -se ha-sy -ess -e-yo?
	what degree -ACC where-LOC do-HONOR-ANT-IE-POL
	'What degree did you get where?'

as well as 
	eti -se mwusun hakkwi-lul ha-sy -ess -e-yo?
	where-LOC what degree -ACC do-HONOR-ANT-IE-POL
	'What degree did you get where?'

when you interview someone for a job.

Of course if you abide the notion of grammaticality in generative
grammar, you could still say your #3a is "grammatical". I am not a
persuant of generative grammar. And I don't believe in their concept of
"grammaticality". I just want to let people know that so-called "free
word order" is not free at all. "

- ------------

I hope this is informative to some of you out there. An interesting
hypothesis behind this apparent typological gap - provided that Omaha
Ponca is no counterexample - would be that the restriction to single
wh-clauses is related to, and perhaps a result of, overt wh-movement
and its properties in the given language.

Best regards,
Ralf Vogel

Ralf Vogel
Institute of Linguistics University of Potsdam
Private - fone: 030 21808465 fax: 0179 336906107 funk: 0179 6906107
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue