Arnold Zwicky's Blog

Sen. John McCain on the “Don’t Ask Don’t Tell” survey of the military, in a Morning Edition Saturday story this morning on NPR:

… it’s a little bit like studying the Bible: you can draw most any conclusion from what part of it that you examine ___.

The WH complement clause (serving as the object of the P from) is boldfaced. It has an initial (“fronted”) WH phrase what part of it that functions, within the complement clause, as the object of the verb examine; the fronted phrase is underlined above, and the the position of the “extraction gap” is indicated by underscores.

Ordinarily we’d expect the fronting of the object in the complement clause to yield

what part of it you examine ___

but the complement clause in the McCain quote has a that in it. That is, the complement is doubly marked, with a WH word and a that, in what I call the “WH-that” construction.

There is some literature on WH-that, by (at least) Andrew Radford, Aimo Seppänen and Joe Trotta, Josef Bayer, Beatrice Santorini and Anthony Kroch, and me. Back in 2004, when I reported briefly on it in Language Log, I thought the construction was infrequent. But this turns out to be wrong, as I eventually discovered, thanks to e-mail exchanges with Bayer.

Most of the literature on the construction up to that point used fortuitously found examples, from reading or listening, and these looked sparse indeed — sparse enough that some critics thought that they all were inadvertent errors, blends of complement and relative constructions. I argued against this in my own article on the subject, noting that there were English speakers who used WH-that and judged it to be an acceptable alternative to singly marked complementation with that.

But Bayer began finding examples in great numbers, as did I when I found ways of searching for specific patterns:

{“what sort of * that”}

{“how much * that we”}

{“what kind of * that you would like”} (suggested by Chris Culy)

{“do you know how many times that”}

{“whatever decision that”}

My view now is that the construction is actually very common indeed, in writing as well as speaking, but that it’s hard to detect: we “read through” the examples without seeing the violation of standard English patterns in them.

So my collection of examples has been expanded enormously, with WH-that cases from many linguists and lexicographers, other academics, reporters, Barack Obama (several times), and others — most often in conversation, but sometimes in writing too.

For some speakers, it’s now pretty clearly an alternative to singly marked WH complementation, perhaps serving as a more emphatic variant, and it’s no longer even clear to me that I should be describing it as non-standard, at least for informal spoken English.

In fact, from my files:

Before you write in about the details — especially the extremely strong tendency (close to inviolable) for the WH phrase to be a multi-word expression and apparent parallels in earlier English and in other languages — check out my article linked to above.

This entry was posted on December 4, 2010 at 4:46 pm and is filed under Complementation, Syntax. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.