Slifted allegations

In the letters section of the New York Times on October 25 (in the Week in Review section), readers commented on conflicts between the public’s right to know and the rights of those involved in legal proceeedings. The last letter accused the Times (and other news media) of subverting the presumption of innocence, via the syntax of the sentences the paper uses to report charges (involving a construction known in the syntactic literature as Slifting).

Clark Hoyt, the public editor of the Times, then explained why journalists sometimes chose Slifting, but conceded that the letter-writer had a point.

Here’s the exchange in full:

The real problem is that The Times, like almost all news media, routinely ignores the presumption of innocence. The most common — but, as your column showed, not the only — way in which the presumption is obliterated is by quoting prosecutors who state their allegations as if they were facts.

Even in simple matters like the order of words, an assumption of guilt is frequently conveyed. Any reporter or editor should know that there is a great difference between “Prosecutors allege that Jones shot Smith because of a dispute over money,” and “Jones shot Smith over a dispute about money, prosecutors charge.” Nonetheless, the fact that this is only an allegation is routinely buried at the end of a sentence describing charges.

Brookline, Mass., Oct. 18, 2009

Clark Hoyt: I am leery of claims that The Times “routinely” does this or that, but Mr. Margolis has a point in this case. In fact, headlines, which carry an especially large punch, are often structured the way he describes. Journalistic convention says to give readers the information first, followed by the attribution. But I think Mr. Margolis makes a good case for reversing that when it comes to unproven accusations of criminal conduct.

In the Slifting construction, an object complement clause (without a complementizer) is “promoted” (or “lifted”, hence the label Slifting, for ‘S-lifting’) to main clause status, while the complement-taking verb, along with its subject, is demoted to a status as a sentence-final parenthetical adverbial. That is, there are (at least) two ways of packaging these constituents into a sentence:

Subordinate Object: Subject V (that) S  (They alleged (that) Kim is a spy. They believe (that) Kim is a spy.)

Slifting:  S, Subject V  (Kim is a spy, they alleged. Kim is a spy, they believe.)

[Bibliographic note: the term Slifting is from Haj Ross (a.k.a. John Robert Ross). See his 1973 paper “Slifting” (in Gross, Halle, and Schützenberger, The Formal Analysis of Natural Languages (Mouton)).]

The two constructions differ in the way they function in discourse: Subordinate Object highlights an act of speaking or thinking and the person or persons responsible for this act, while Slifting highlights what is spoken or thought. So Slifting is attractive to journalists, especially in headlines (as Hoyt explained).

The two constructions also differ in the processing burden they put on the reader (or hearer). Slifting can induce a kind of garden-path effect, in which the reader takes the lifted clause to express content that is to be attributed to the writer, but then along comes the parenthetical with a different attribution. That’s the effect that Margolis is complaining about.

I’m not at all sure how troublesome the deferred attribution is for most readers. Certainly, there are plenty of languages with “evidentials”, elements that convey various evidential modalities: I know this from first-hand experience; this is common knowledge, something that “everyone knows”; I heard this from someone else; this is rumored to be so; and so on. These elements can be inflectional affixes, clitics, or particle words, and for a variety of reasons they often end up in sentence-final position. So huge numbers of people all over the world are accustomed to coping with sentences in which indications of evidentiality are deferred to the end of a sentence. Slifting fits into this pattern because the coda in many Slifted sentence expresses evidentiality — plus an attribution of the source of the evidence.

In fact, Slifted sentences are not especially infrequent in modern English, though I have no idea how often they involve verbs — like accuse, allege, assert, charge, claim, maintain, say — that can be used to report allegations in law, nor do I have any idea how often the Times uses Slifting this way, in heads or in the body of stories. (I am of course suspicious of claims that particular sources “routinely” do something or other.) These are questions that could be investigated, but with a fair amount of work.

(There are alternatives to Subordinate Object and Slifting. In heads, there’s the verbless intro of

Prosecutors: Jones shot Smith over money dispute

and there are codas with or so:

Jones shot Smith because of a dispute over money, or so the prosecutors say.

No doubt there are further possibilities.)

2 Responses to “Slifted allegations”

  1. The Ridger Says:

    I said on my blog the truly troubling part of “Jones shot Smith because of a dispute over money, prosecutors allege” no matter where the “allege goes”, is that the “Jones shot Smith” part is rhetorically assumed, and only the “because” part seems to be alleged.

  2. Ben Zimmer Says:

    UK headlines don’t necessarily require Slifting, because they can just use claim quotes instead. (Attribution of the claim in the headline may then be deferred to a subhed or lede graf.)

Leave a Reply

%d bloggers like this: