I'd like to ask those of you who speak English as a first language whether indirect objects induce CED effects (cf. Huang).


  1. Of which boy did John send [a letter] [to every friend _]?
  2. Of which paper did John persuade [every author _] [PRO to make sure that every citation is properly put]?

I know the sentences above allow subextraction fairly well. But what about the following ones?

  1. Of which boy did John write [every friend _] [a letter]?
  2. Of which paper did John promise [every author _] [PRO to provide a research fund]?

Any significant differences on acceptability compared to the previous ones?

Thank you in advance.

P.S. I just assumed the constituent [every author of which paper] in (4) has a similar status to indirect objects, but be it or not doesn't much matter here.

EDIT: For clarification, Huang (1982) proposed the Condition on Extraction Domain (CED), which states extraction out of a non-complement induces island effects, and subsumed two well-known island constraints under it: namely, the Subject Condition and the Adjunct Condition. Thus, whereas the CED permits (5), it renders (6) and (7) ungrammatical.

  1. Who(m) did Mary meet [a friend of _]?
  2. *Of which car did [the driver _] cause a scandal? (From Chomsky 2008)
  3. *Who did Mary cry [after John hit _]? (From Huang 1982)

I believe it is empirically quite well established that the contrast shown above holds across a wide range of English speakers. What I wanted to know was whether indirect objects in (3) and (4) also show CED effects. If the wording of the sample sentences makes it confusing to judge, I'd rather give you more abstract structure below:

  1. [Subj ...] V [IO ... wh ...] [DO ...]

Question: Is it possible to extract wh from IO in such structure?

Plus, as for grammaticality, I'd like to make sure that it is a different notion from usability, conciseness, etc. It is a common practice in linguistics to deal with borderline cases, which may have no places in everyday usage. For instance, consider:

  1. ?Who(m) do you wonder [whether John met _]?
  2. *Who do you wonder [whether _ met Bill]?

Though both (9) and (10) are not perfectly acceptable, and are not apt for your essays, emails, etc. etc. but still they differ sharply; while the former is just a case of weak violation of the Wh-Island Condition, the latter is a case of both Wh-island and ECP violation.

I hope this may illustrate what I intended by "acceptability," "grammaticality," etc.

  • 3
    Your sample sentences are not grammatical in English so the question cannot be answered.
    – Lambie
    Commented Oct 23, 2023 at 13:46
  • @Lambie Thank you for the comment. I knew that the sample sentences would sound more or less “awkward,” and of not much use, but what I had in mind was a more subtle notion of “grammaticality.” For example, you know, “whom do you wonder whether John met _” and “who do you wonder whether _ met Bill” are both deviant, yet the latter much more severely so. (ECP) I don’t know how awkward the sample sentences are as I’m not a native speaker, but I guess they are not complete gibberish.
    – Tzetachi
    Commented Oct 23, 2023 at 14:14
  • 1
    Problem 1: your question is not clearly stated (are you comparing 1,2 as a group with 3,4, and what if 1,2 or 3,4 are not the same in acceptability. Problem 2: you don't tell us the possible answers (how many degrees of freedom in answers?). Problem 3: this is an opinion poll question which is unsuited for SE. I could post my opinions, jlawler could post his, we could accumulate dozens of opinions – which one would be correct? The concept "correct answer" is fundamental to the operation of SE.
    – user6726
    Commented Oct 23, 2023 at 15:01
  • 1
    @user6726 As for Problem 2 & 3, the notion of degrees of grammaticality is indeed a subtle and vague one and varies from dialect to dialect. But still we can coherently contrast mere marginal deviance to extreme unacceptability as is a common practice. E.g., those classical studies on the subjacency versus ECP violation. And for doing this, we can’t help but relying on the intuition of native speakers. Anyway, it’s true I wasn’t clear enough and assumed too much. I’ll revise my question in the morning. Thank you for the comment.
    – Tzetachi
    Commented Oct 23, 2023 at 17:51
  • 1
    Suffice it to say that this is not a generative-friendly site, so you have to be more explicit and now assume any knowledge of that line of research.
    – user6726
    Commented Oct 23, 2023 at 18:26

1 Answer 1


A self-answer though, I found a paper by Toquero-Pérez (2022), who reports that speakers from Southern California, Northern Virginia and North East Pennsylvania allow wh-extraction of indirect objects, but don't allow subextraction from them as the following contrast shows:

a. Who did MacNulty show t a picture of Baltimore?

b. *Which person did McNulty show [a friend of t] a picture of Baltimore?

(Toquero-Pérez 2022, 23; ex. 43)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.