Hearing voices in the Barr report

I’ve unexpectedly spent a good chunk of my sabbatical semester thinking about heteroglossia. This might sound as if I’ve had just too much time on my hands, but I’ve become convinced that it’s one of the key concerns in academic writing. (Tl;dr version: skip to the bottom for the teaching implications!)

Heteroglossia is a concoction of Greek word parts which means “many voices,” and it’s often attributed to Bakhtin (and sometimes his contemporary Voloshinov). Neither of them used the word heteroglossia because they wrote in Russian, using the word raznorechie, meaning “different-speech-ness” (The Bakhtin Reader, p. 248). For Bakhtin, language exists in a constant tension between a unifying force (the idea that all English speakers understand the meaning of a word in English) and the centripetal (spinning apart) forces of heteroglossia because every word has been used somewhere before and will be used somewhere again. The echoes of those past and future voices stretch the meaning of everything we say. “The living utterance,” Bakhtin writes, “… cannot fail to brush up against the thousands of living dialogic threads.” (The Dialogic Imagination, p. 276).

Martin & White (2005) take this idea of heteroglossia and use it in SFL’s Appraisal theory — more on that later — to mean “whether or not speakers acknowledge alternative positions to their own” (p. 36). While recognizing that all language really exists in a dialogue, Martin and White draw a distinction between monoglossic and heteroglossic utterances. Actually, it would be better to talk about more and less explicitly heteroglossic utterances (cf. White, 2003) as nothing we say is every truly and solely in our own voice:

It must not be forgotten that monoglossia is always in essence relative. After all, one’s own language is never a single language: in it there are always survivals of the past and a potential for other-languagedness. (Bakhtin, “Epic and the Novel,” p. 66)

Take for example, this bit from the opening of Attorney General William Barr’s March 24, 2019, letter to Congress:

On Friday, the Special Counsel submitted to me a “confidential report explaining the prosecution or declination decisions” he has reached, as required by 28 C.F.R. $ 600.8(c). This report is entitled “Report on the Investigation into Russian Interference in the 2016 Presidential Election.”

Look at the word report and think about the echoes it has both within and beyond the text. It resonates with every use of the phrase “the Mueller report,” and indeed is repeated three times in these two sentences, with each incidence reflecting on the others. The same is true of “Russian Interference” or “2016 Presidential Election.” This extract is heteroglossic in another way, the way that Martin & White use the construct: there are two other voices directly implicated: the US federal code (from which the first quote comes) and Mueller himself in his official capacity as Special Counsel in quoting the title of his report. We can also see an element of attribution, since the decisions in the report (but not necessarily the letter) are those Mueller has reached.

(Caveat, here: as noted, I’m still reading and thinking about heteroglossia, and my grasp on both Bakhtin and Appraisal are tenuous at best, so the following should be read as my incomplete attempt at a linguistic analysis.)

I’m using Barr’s letter not out of any political agenda, but because it’s really interesting to consider from the perspective of heteroglossia since the text is a massively important act of re-voicing a text which, at the time of writing, very few people have actually read. At the same time, some segments of the media, social media, the political chattering classes, the administration itself, and the public are treating this letter as transparently and neutrally reporting “the principal conclusions reached by the Special Counsel and the results of his investigation.” I’m not taking a position here on whether I think Barr is acting, as he claims, “in the public interest.” I’m just fascinated by the linguistic hoops he has gone through to balance the Bakhtinian tension between an authoritative, unitary voice and the inevitable “different-speech-ness” of the report which we can only glimpse through the Barr text.

Listen to all the voices in the next paragraph of his letter:

The report explains that the Special Counsel and his staff thoroughly investigated allegations that members of the presidential campaign of Donald J. Trump, and others associated with it, conspired with the Russian government in its efforts to interfere in the 2016 U.S. presidential election, or sought to obstruct the related federal investigations.

There are three levels here:

• Someone (ooh, mystery) alleged conspiracy and obstruction of justice.
• The Special Counsel investigated these allegations.
• Barr tells us what the report explains.

Barr carefully distances himself from both the allegations and the investigation and casts himself as a neutral voice (only three words in the that sentence are in Barr’s voice). So far, so good. The paragraph ends:

The report does not recommend any further indictments, nor did the Special Counsel obtain any sealed indictments that have yet to be made public. Below, I summarize the principal conclusions set out in the Special Counsel’s report.

The recommendation not to indict is attributed to the report, but the second clause is a bit trickier. Here’s where we have to start taking Barr’s word: this clause is what Martin & White term a bare assertion (the Special Counsel did not obtain); in other words, Barr himself takes responsibility for the truth of the statement, unlike the first clause, which puts responsibility for the lack of further indictments on the report. Then, Barr uses “I” for only the second time in the letter, promising to “summarize the principal conclusions“. We are now engaged in a delicate heteroglossic balancing act: every time we read an apparently “monoglossic” statement  — that is, one without any explicit attribution to the report, Mueller, or another voice — we have to decide whether to interpret it as truly single-voiced (Barr‘s) or actually an implicit projection of the report. And since we can’t (yet) read the original report, we can never be certain of the degree of heterolgossia. Again, I’m not saying we can’t trust Barr’s conclusions. I am saying that we can’t be certain which voice(s) are speaking.

Here then comes conclusion #1:

The Special Counsel’s investigation did not find that the Trump campaign or anyone associated with it conspired or coordinated with Russia in its efforts to influence the 2016 U.S. presidential election. As the report states: “[T]he investigation did not establish that members of the Trump Campaign conspired or coordinated with the Russian government in its election interference activities.”

A few things to note here: Russia’s “efforts to influence …” are taken for granted (per Martin & White): they are not at stake in this letter, and both Barr and Mueller, whose voice appears here directly, accept that such efforts took place. (Compare this sentence to “investigated allegations that …” earlier). OK. But these two sentences are … odd. For a start, they appear to say the same thing: at the experiential level, they have more or less the same bits and pieces (participants and processes in SFL): Trump campaign / conspired / coordinated / Russia / election. And the first sentence is Barr’s paraphrase of the sentence he then quotes, one of very few direct quotes in the letter. What’s going on? The first reporting verb (did not find that) is doubly heteroglossic (ooh): the negative conjures up a little voice saying “it did find something!” in order to deny it (per Martin & White), and then the verb projects the indirect speech in the noun clause. Find is very certain: it’s called an endorse move in Appraisal because it tries to align the reader with the finding, implying that there is no doubt as to the Trump campaign’s ignorance innocence. Got it. Except that’s not what the quotation says. As the report says is an attribution, setting up the quote, which itself uses the projection resource (noun clause): “The investigation did not establish that …” Wait, what? If you don’t establish something, it means you can’t prove it, not that you found it did not happen. In other words, there are two contracting moves here: a deny (not) and endorse (establish) move. And while two negatives do not make a positive in linguistics, in this case I think they actually expand the space just a bit for an alternative interpretation. Mueller didn’t have to write the sentence this way, essentially not taking responsibility for the decision and almost blaming the investigation for not “establishing” a conspiracy. In SFL, we interpret language choices by comparison with choices not made. Why not write “According to our investigation, the Trump Campaign did not conspire or coordinate“? That simple negation would be unambiguous. And what is Barr up to? He repeats some version of “the report did not find …” two more times in the next paragraph. If the quotation was intended to legitimate (Hood, 2010) his paraphrase, it doesn’t really work linguistically.

On now to the second part of the letter: is there a case to be made against the President for obstruction of justice? Answer: Meh.

After making a “thorough factual investigation” into these matters, the Special Counsel considered whether to evaluate the conduct under Department standards governing prosecution and declination decisions but ultimately determined not to make a traditional prosecutorial judgment.

Ah, those quotation marks again! Mueller’s voice is once more invoked, presumably to lend this key sentence his authority, but the choice of quote is again strange and even redundant: Barr himself should be in a position to evaluate whether the investigation is thorough and factual. There are once more three voices at stake in this sentence: Barr, Mueller, and now “Department [of Justice] standards” hiding in the background. Linguistically, this section is relatively straightforward: Barr’s paraphrase (did not draw a conclusion one way or the other) closely aligns with the tantalizing snippet from the Mueller report (“while this report does not conclude that the President committed a crime, it also does not exonerate him”). By all accounts, this is indeed not a “traditional prosecutorial judgment”, a phrase in which one might detect a hint of disapproval from Barr. I’m interpreting the last bit as a form of projection using a non-finite to clause because Barr is emphasizing that it is Mueller’s — not Barr’s — decision not to decide.

And in true Bakhtinian spirit, the carnivalesque heteroglossia triggers a delightful intertextual reference to that noted commentary on American politics, Steven Sondheim’s Into the Woods:

Then from out of the blue
And without any guide
You know what your decision is
Which is not to decide
You’ll just leave him a clue
For example: a shoe
And then see what he’ll do
Now it’s he and not you who is stuck with the shoe
In a stew
In the goo

So there we have it. Barr is stuck with Mueller’s gooey shoe on the steps of the White House.

I wrote this post because I wanted to better understand the monogloss/heterogloss tension. What I’ve learned from this exercise is that it’s important to recognize that a lot of apparently single-voiced utterances echo, reflect, or anticipate other voices. That is, we don’t in fact have a binary choice between monogloss and heterogloss (which the Appraisal system network visually suggests) — Susan Hood puts it very nicely:

The engagement network offers a system of semantic choices for constructing discourse as more or less monoglossic or heteroglossic, that is, as more or less single-voiced or multi-voiced. (Hood, 2010, p. 188, my emphasis)

Thus, when Barr picks up Cinderella’s shoe, we get a sentence that is more (but not completely) monoglossic:

Deputy Attorney General Rod Rosenstein and I have concluded that the evidence developed during the Special Counsel’s investigation is not sufficient to establish that the President committed an obstruction-of-justice offense.

Here, he is projecting his own conclusion into the noun clause (in green), co-opting his deputy as additional legitimation (again, linguistically if not necessarily politically). From a technical analysis, this is a heteroglossic but contracting pronouncement move: I am telling you that this is my conclusion, and I am limiting the alternate conclusions you can draw from this sentence. But this is clearly more monoglossic than, say, “the report concludes that …” and more heteroglossic than the bare assertion “the evidence is not sufficient.” Barr makes it clear that since Mueller is still stuck in the goo, he has been left to make the decision (I have concluded that …). In other words, the effect of the heteroglossic resource (the reporting verb) is to project and foreground the use of his own voice. And, as he is well aware, there are going to be other voices that disagree, but none of them will have the authority of the Attorney General and his Deputy.

What does this multivoiced and multicolored analysis mean for teaching academic writing? Here are some preliminary thoughts:

• We have to take much more seriously the teaching of source use at the discourse semantic (or, if you like, rhetorical) level, not (just) as paraphrasing, quoting, and avoiding plagiarism. MANY others have said this (heteroglossia), and they’re all right (monoglossia).
• Students need access to the full range of lexicogrammar that brings other voices into their texts: reporting verbs, attributive phrases, indirect citations, nominalizations (suggestion, proposal, theory, claim, etc.) and more.
• Most undergraduate writing is implicitly heteroglossic: they are re-writing (re-voicing) what they have read in their textbooks and heard in their lectures (e.g. short-answer questions, see Melzer, 2009). We need to know how to help them do this effectively, and not just teach personal “essays” and source-based “research papers.”
• Beyond the old chestnut “(when) can I use I?”, novice student writers need to know when and how they can be authoritative, assertive, and (more) monoglossic. And when they can’t.
• Students need practice identifying and working with the multiple voices in texts — e.g. not just citing an author, but understanding when that author is actually projecting other voices with which they may or may not be aligned (this was an area the Citation Project found that students struggle with).

Comments and corrections of my SFL analysis are most welcome below! Join the dialogue (Bakhtin would approve of blogs, I suspect).

Selected references and other voices:

Bakhtin, M.M. (1935/1981). Discourse in the novel. In The dialogic imagination (trans. M. Holquist & C. Emerson). Austin, TX: University of Texas Press.

Hood, S. (2010). Appraising research: Evaluation in academic writing. London: Palgrave.

Martin, J.R. & White, P.R.R. (2005). The language of evaluation. London: Palgrave

Author: Nigel Caplan

Nigel Caplan, Ph.D., is an associate professor at the University of Delaware English Language Institution, as well as a textbook author, consultant, and speaker. Nigel holds a PhD from the University of Delaware, a master's in TESOL from the University of Pennsylvania, and a bachelor's degree from Cambridge University. He is currently director of Project DELITE, a federal grant providing ESL certification to Delaware teachers. He also brews beer. View all posts by Nigel Caplan