LINGUIST List 10.173

Thu Feb 4 1999

Sum: Multiple Wh-Questions

Editor for this issue: Karen Milligan <>


  1. Miura Ikuo, Summary: multiple wh-questions

Message 1: Summary: multiple wh-questions

Date: Thu, 4 Feb 1999 21:44:36 +0900 (JST)
From: Miura Ikuo <>
Subject: Summary: multiple wh-questions

Dear linguists,

 In September, I posted two questions about multiple wh-questions in
English. I'd like to thank everyone who replied to me, and apologize for
taking so long time to post the summary.
 My questions were the following.

1. In the literature, it is observed that while the sentence in (1a) is
grammatical, the corresponding (1b) is not. It is known that even if (1b)
is grammatical, it is only as an echo-question.

(1) a. Who said what?
 b. What did who say?

I want to know whether the following pairs of sentences exhibit the same
contrast as in (1).

(2) a. Whose mother bought what?
 b. What did whose mother buy?
(3) a. People from where bought what?
 b. What did people from where buy?
(4) a. Tell me whose advisor is where.
 b. Tell me Where whose advisor is?

The sentences in (2b) and (3b) are from Stroik (1995). He says that they
are grammatical even if the object wh-phrases have moved across the other
wh-phrases in the same way as (1b). But he doesn't mention about the
grammaticality of (2b) and (3b).

2. In the literature, psych-verbs like 'worry' and 'annoy' which take the
experiencer argument as the object behave differently from verbs like 'say'
with respect to some phenomena like anaphor binding. So I want to know
whether or not multiple wh-questions of psych-verbs like (5) and (6)
exhibit the same contrast as (1).

(5) a. What worries who?
 b. Who does what worry?
(6) a. What annoies who?
 b. Who does what annoy?

 The following are those who answered these questions:

 Dan Faulkner
 Jim Witte
 Bernard Kripkee
 Vincent Jenkins
 And Rosta

As for question 1, all of them agree that (2b) and (3b) have the same
status as (1b). That is, they are ungrammatical, or grammatical in special
contexts. These judgments are surprising, because Stroik takes (2b) and
(3b) to have the same status as (1a) not (1b), i.e., he seems to judge (2b)
and (3b) as grammatical in isolation.
 The contexts which make (1b), (2b), and (3b) acceptable are echo-question
contexts or "the end of a logic puzzle" (see what Vincent Jenkins wrote

Dan Faulkner said:

Imagine a situation where someone enters a room where two people are
discussing a third party's opinion. The conversation goes:

SPEAKER A: "John wasn't happy about the report"


SPEAKER B:"Why? What did he say?"
SPEAKER C:"What did who say?"

If SPEAKER C started listening to the conversation as speaker B asked
his question, then 1(b) is vaid, because speaker C doesn't know that the
person being referred to is John - that is, he doesn't know who the "he"
in SPEAKER B'S question is, OR what was said.

Vincent Jenkins said:

This structure (where "wh" is not fronted but remains in place) is used
when the speaker is asking for information to be repeated because he missed
that bit or cannot believe his ears. In this case the "wh" word also
carries rising or falling pitch. This structure can also be found at the
end of a logic puzzle where the hearer has been prepared by context for an
over-abundance of question words, e.g. "Where did who drive who in whose
what?" An odd point is that I have never seen this mentioned in any text,
but as a native speaker I assure you that unmoved "wh" is a normal part of

 1 b. What did _who_ say? OK

 2 b. What did _whose_ mother buy? OK

 3 b. What did people from _where_ buy? OK

Bernard Kripkee said:

 In isolation, [(1b)] seems ungrammatical. However, I
 can imagine a pragmatic context [something like an echo
 question] and an intonation pattern for it that makes it more
 nearly acceptable:

 A: What was it that Rumplestiltskin said? Something like
 understanding begets faith?

 B: What did who [emphasis] say?

 . . . with the appropriate pragmatic context and
 intonation pattern, [(2b)] can be made acceptable:

 A: Newt Gingrich's mother bought Monica Lewinsky a new

 B: What did whose mother buy?

 In contrast to (1b), (2b) and (3b), judgments of (4b) seem to vary. Two
of the contributors judged it to be simply unacceptable. One takes it to
have the same status (2b) and (3b), i.e., grammatical as an echo-question.
(The other two, who do not mention about the possibility of the
echo-question interpretation of (2b) and (3b), only point out (4b) have the
same status as (1b), (2b), and (3b).) Vincent Jenkins, who said that (4b)
is unacceptable even as a echo-question, wrote that the following minor
change of (4b) makes it acceptable:

 Tell me where the advisor of _who_ is.

 As for the question 2, Most of the contributors seem to agree that (5)
and (6) exhibit the same contrast as (1). That is, (5a) and (6a) are
acceptable on their own, but (5b) and (6b) are ungrammatical, or
grammatical as echo-questions.
 However, And Rosta said that even (5b) and (6b) have the same status as
(1a). This judgment is very interesting for me.

 Finally, I would like to thank again the contributors cited above. But
I need more data about this area of research. Especially, I would like to
know whether there are native speakers who judge the sentences in (2b) and
(3b) to be acceptable on their own. And I would like to contact with more
native speakers who judge (5b) and (6b) to have the same status as (1a).
Anyway, any comment will be welcome.

Many thanks,
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue