Shuy: The Language of Defamation Cases (2010)

Roger Shuy is a so-called "forensic linguist," meaning that he appears as an expert witness in court cases that centrally involve questions about language, such as libel suits. He's written a number of books reporting on cases he's been involved in, the most recent being The Language of Defamation Cases.

His discussions are interesting because they show how quickly our unreflected notion of "meaning" starts to come undone when it gets exposed to the extreme hair-splitting that libel suits unavoidably entail. "What a statement means" becomes less rather than more clear when you look at it up close.

Frank Celebrezze v. The Plain Dealer

In 1986, the (Republican) Cleveland newspaper The Plain Dealer ran a series of stories about the (Democratic) Ohio Supreme Court justice Frank Celebrezze. Shortly thereafter, Celebrezze lost the race for re-election for the Supreme Court.

Before the publication of the articles, Celebrezze had taken campaign donations from labour unions, and some of these labour unions had members who were convicted of involvement in organized crime. As a judge, he had also voted against the conviction of people accused of organized crime at least twice.

So, one might conjecture that he was corrupt, and the mafia had bought his vote — and this was exactly what the two journalists Gary Webb and Mary Anne Sharkey claimed in a series of Plain Dealer articles, published immediately before the Supreme Court election.

Or was it? A closer reading of the articles in question revealed that they did not directly state that Celebrezze was corrupt, but one might argue that they suggested as much. Celebrezze consequently sued The Plain Dealer, and both he and the paper hired linguists to back their case.

Shuy represented the paper, and a local English professor was made Frank Celebrezze's case. Both experts wrote a report, but the case was eventually settled outside the court for an undisclosed sum.

The Apple of Discord

A number of quotes were brought to bear on the case. Most of them were headlines like this one:

Chief Justice denies mob role in contributions (p. 101)

But also some excerpts from the articles themselves were used, including this one:

In 1982 Celebrezze cast a tie-breaking vote against convicting Liberatore of arson. State records show that five days later, a Celebrezze campaign fund was given $5,000 … (p. 103)

The question was: Do these two sentences together imply that Celebrezze received the campaign donation as a result of his vote, or does it not?

The Argument Pro

According to Celebrezze's expert, both snippets of text indirectly implied that Celebrezze was guilty. With respect to the headline, the argument was a pragmatic one about presupposition:

If the headline had read Chief Justice Says Accusations Concerning Mob Role are False then the presupposition would have been that someone had made such accusations, not that the mob role was a given fact. (p. 101–102)

In other words, the claim is that the verb deny is veridical, or factive, while accusations concerning is not.

In the case of the quote, the argument refers to discourse coherence relationships:

Juxtaposing the two claims (that the judge had cast a tie-breaking vote and received money five days later) creates the obvious innuendo that there is a cause and effect relationship between the vote cast and the money paid. (p. 103)

So a reader searching for a coherent link between the two sentences will stumble upon cause–effect as the first likely candidate. This can be compared to the following examples, adapted from Jennifer Spenader:

• Bill is worried that John might try to gain access to his safe. He'll have to change the combination. (therefore)
• Bill is worried that John might try to gain access to his safe. He knows the combination. (because)
• Bill is worried that John might try to gain access to his safe. I like spinach. (??)

Although the general claim that people will automatically search for discourse relations is sound enough, it is worth remembering that the choice of discourse relation is notoriously difficult to consistently agree on.

Arguments Contra

Shuy's own argumentation focused on the impotence of linguistics with respect to what a reader will extract from a text, or what the motives of a writer are:

There is no way that a linguist can actually reach inside the mind of the writer to determine what that writer intended. (p. 107)

There is no way that linguistic analysis can prove such attributions of a person's intentions. (p. 109)

The plaintiff's expert's use of expressions such as "must have," "they are affected," and "readers would" attribute results or behaviors [to the reader] that linguistic analysis simply cannot provide. (p. 111)

This is a surprisingly self-defying argument to hear from a person who is himself a forensic linguist. If it were really the case that linguistics has nothing to say about how a text works or what it reveals about the author, what is it doing in the courtroom in the first place?

The argument is also strange given that Shuy in the same chapter explains that he once had a courtroom disagreement with another linguist who "actually agreed with my conclusions but claimed that the field of linguistics could lead neither of us to such findings" (p. 98). To my ears, this sounds surprisingly close to saying that there is "no way that linguistic analysis can prove such attributions of a person's intentions."

The Flexibility of Forensic Linguistics

Further, when Shuy appears on the other side of a court case, he also seems to embrace a much more liberal approach to meaning:

Otherwise benign words and expressions, such as "made plans," "was alone," "secretly left the house," "went unanswered," "left town," and "living with," can convey meanings far beyond their usual dictionary senses, especially in the context of a broadcast about a murder. (p. 61–62)

But oddly, this is not the case with words like deny in the context of newspaper reporting on mob crime.

It seems that the conclusion we can draw is that linguistics either can or cannot say something about meaning, depending on whom its employer is on a given day. In the murder case (chapter 4), journalists are thus clever manipulators setting up snares for their helpless sources and listeners:

Most listeners are not very likely to notice the discourse framing of the broadcasts or that there was little attempt to clarify important ambiguities. (p. 68)

But in the corruption case (chapter 7), they have regained their agency, and the newspaper is liable only for explicit accusations, not innuendo:

… there is no such accusation in the articles, which present the facts about the relationship between the plaintiff and the local unions … The Plain Dealer claimed that the readers could draw whatever conclusions, if any, that these facts suggest to them. (p. 115)

Probably, the oscillation between these two poles of semantic theory is destined to go on as long as linguists continue to see meaning as a feature of words rather than situations.

Jennifer Spenader: "Variation in Demonstrative choice in Swedish" (2004)

This really thoughtful and interesting paper by Jennifer Spenader tries find out when Swedish speakers prefer references of the form det här huset ("this here house") over reference of the form detta hus ("this house").

She sums up her findings in the maxim "avoid marking that which can be determined from the context" (p. 246). More specifically, this means that the compound demonstrative is inappropriate for a recently mentioned referents. It is also less appropriate for animate referents, allegedly because they are more "accessible" and may thus be recovered from context.

Methodology
In order to test what kind of referents her subjects preferred in different situations, she asked them to fill in forms of the following type (cf. p. 249):

Many research studies show an increase in diabetes or "sugar sickness" among women working in the government sector.
Passive work combined with lack of physical activity leads to an unhealthy lifestyle.
The research also shows that the social part of the women's work often involves chatting over an unhealthy snack.
A. ____________ doesn't don't occur among women in the private sector.
C. ____________ate during certain periods great amounts of candy.

The two last sentences are alternatives that suggest either a more abstract (A) or a more concrete (C) referent for the empty slot. The subjects were only presented with one of them.

Results
Using logistic regression, Spenader found that the most predictive factor for the choice of form was sentence recency: The more time that had passed since a referent was mentioned last, the more likely the subject was to use a a compound form (den här ...). The dependency between the discourse distance and choice of form can be seen in this chart:

The second-most predictive factor was animacy/concreteness. This variable could take the three values Animate, Inanimate-abstract, and Inanimate-concrete.

An Alternative (Stochastic) Model
In section 5 of the paper, Spenader describes an alternative model she has devised within the paradigm of stochastic optimality theory. I didn't know what this involved, but I must say that I find it quite odd.

The idea in stochastic optimality theory is not, as one might think, one of putting weights on the different constraints so that the parsing of an input becomes a matter of maximization of a real function. No, the idea is rather to let the order of the different conflicting constraints be stochastic. It thus introduces noise into the application of the rules in order to reproduce rather than explain empirical variations.

In any case, Spenader sets up a system with seven constraints and trains a system to selecting the best ordering. This turns out—again—to select recency as the most decisive factor, followed by animacy. She tests the trained grammar not by a statistical method but by generating a single random observation and comparing that to the empirical data (using chi-square tests).

This seems like a methodologically bad choice. One could instead pick a test statistic and find the probability that a new data set generated from the trained model would be more extreme than the empirically observed data points. If need be, this could be done with Monte Carlo methods, but I'm not sure that it would even be necessary.

Sociolinguistic Explanations
I find it extremely odd that Spenader doesn't even mention the most obviously relevant dimension of explanation, namely that of high vs. low style.

There is something distinctly informal about constructions of the form det här huset, even though slightly less in Swedish than in the (African-American) English this here house. The closest we get to this is the initial discussion of written vs. spoken language (p. 229).

This formal/informal dimension should be seen in conjunction with, for instance, the social aversion against pointing directly at someone, especially if you're not talking directly to them. This tendency may in some cases have a mirror image in the Japanese tendency to suppress certain nominal phrases and find it coarse or impolite to fail to do so.

A striking English example of this tendency is John McCain's infamous reference to Barack Obama as "that one" during a presidential debate in 2008. He also, by the way, pointed at Obama (without looking at him) while making the reference. The results was that he emitted an air of superiority, as if Obama was an inanimate thing.