As a person who reads things on the internet, you may or may not be aware that David Foster Wallace wrote a thing about grammar one time. The linguistic aspects of the piece have gotten their share of attention, on account of the fact that people just freak the fuck out about language for whatever reason, but Wallace’s fundamental claim is that language is inherently political; as such, the linguistic argument he makes here is also a political argument. Note in particular that, as published in Consider the Lobster, the title of this essay is “Authority and American Usage.” The important word there is “Authority.”

Structurally, the essay, which is ostensibly a review of a usage guide written by Bryan Garner, begins with an overview of the relevant linguistic issues, transitions into a political analogy for those issues, and then moves back into linguistics to discuss what Wallace feels is the importance of Garner’s book. The political analogy here is important because it means that Wallace’s conclusion about Garner is simultaneously a political argument. The purpose of this post is to explain what this argument is and why it’s wrong.

To start with, there are a number of factual and argumentative errors in Wallace’s essay that need to be discussed insofar as they inform his political analogy. Wallace starts by dividing language-carers up into “Descriptivists” and “Prescriptivists” (he consistently capitalizes both terms to make them seem like real affiliations that he’s not making up, so I’m going to be just as much of a dick and consistently put them in scare quotes). The immediate problem with this is that neither of these groups actually exists. “Descriptivists” are actually “linguists,” that is, people who study language and try to figure out what’s going on with it. The fact that they take a “descriptive” approach is a matter of necessity, not ideology; it’s just kind of how you have to approach things if you want to learn about them.

“Prescriptivists,” meanwhile, don’t exist for the opposite reason: they’re just people who have opinions about language; there isn’t anything specific that unites them. Everyone has opinions on language; a “Prescriptivist” is just anyone with enough privilege to get their opinion in the paper.

Wallace frames “Descriptivism” as a “rejection of conventional usage rules in English,” but this can’t actually be the case, since language obviously had to exist in studiable form before people could start applying explicit rules to it. In fact, the situation is the opposite of what Wallace claims: it’s prescriptive rules that only make sense in the context of an already-existing language; if people weren’t already doing things “wrong,” there wouldn’t be anything to complain about.

And obviously no language is ever created out of “conventional usage rules” in the first place, despite Wallace actually claiming that “language was invented to serve certain very specific purposes.” I’m going to give him the benefit of the doubt here and assume this is a convenient shorthand for “language arose in response to certain necessities” (so, uh, not that much of a shorthand, I guess), but it still indicates a fundamentally flawed way of looking at the issue.

Wallace’s claims here are ultimately based on the idea that the publication of Webster’s Third in 1961 was “the Fort Sumter of the contemporary usage wars,” due to the fact that its introduction set forth the “basic edicts” of “Descriptivism” for the first time. Nothing about this is correct. As this article explains, Wallace’s facts here are wrong in every important way. First, the text in question was not published in Webster’s Third. Second, it was not written by Philip Gove, the editor of Webster’s Third; it was quoted by him from the National Council of Teachers of English. Third, these were not new principles; they were endorsed by the Council precisely because they represented an existing consensus. Thus, it is not true that “Descriptivism” emerged as an “attack” on traditional grammar; to the extent that it qualifies as an ideology at all, it was merely an academic consensus that gradually spread out to influence society in general.

Wallace analogizes “Descriptivists” to political liberals and “Prescriptivists” to political conservatives, which is where the above errors start to become significant. His claim that “Descriptivism” “quickly and thoroughly took over US culture” is, by analogy, equivalent to the familiar conservative argument that liberalism is a recent phenomenon that emerged to attack traditional American values. In fact, Wallace makes this analogy explicit: he claims that “the ideological principles that informed the original Descriptivist revolution” were “the rejections of traditional authority (born of Vietnam) and of traditional inequality (born of the civil rights movement)” (bonus curmudgeon points: he says this in the middle of a complaint about “Political Correctness,” again, capitalized).

The political argument here is wrong in precisely the same way as the analogous linguistic argument. Power accumulation in human societies is a natural phenomenon; it’s only after a particular power structure has been established that it gets framed as “traditional” for the purpose of justification. Furthermore, as long as there’s been arbitrary authority, there’s been resistance. The concept of a “traditional” past that allows the struggle for justice to be framed as a recent aberration rather than a fundamental part of human history is a myth propagated by the ruling class, for obvious reasons.

Having established his framework, Wallace spends quite a lot of time debunking a phantom version of “Descriptivism” that allegedly claims that literally every utterance made by any speaker in any circumstances is 100% correct English, i.e. “there are no rules.” Nobody actually believes this, yet Wallace still feels the need to make arguments like “you can’t actually observe and record every last bit of every last native speaker’s ‘language behavior’.” No shit. Isn’t it pretty obvious that no one is actually trying to do this? General advice: if you feel the need to make an extremely obvious argument like this, you probably don’t understand the position you’re arguing against.

Again, this is analogous to the conservative talking point that anyone against any kind of existing authority is a radical who just wants to destroy society. I should probably clarify here that I’m not arguing that Wallace is himself a political conservative; he’s obviously not, and he implicitly allies himself with liberalism in this essay. The issue is precisely that it’s odd that his argument so closely mirrors the basic mythology of U.S. conservatives. As we’ll see, the problem is not Wallace’s explicit opinions, it’s his approach.

Wallace eventually gets around to the “Descriptivist” argument that’s actually relevant, which is that language consists of a set of actual rules which all fluent speakers instinctively follow (and therefore don’t require explicit elucidation from newspaper columnists), plus a set (actually several sets) of optional rules based on dialect, setting, etc., which may be followed or not without general loss of comprehensibility. Nobody is going to actually misunderstand you if you end a sentence with a preposition. The rest of the essay is a refutation of this argument; specifically, Wallace’s claim is that the “arbitrary” rules that “Prescriptivists” propound are not elitist window-dressing but are actually super important.

I’ve seen some people claiming that Wallace’s argument in this article actually ends up supporting “Descriptivism,” so it’s worth being explicit about his position here. He refers to the first (fake) version of “Descriptivism” as “Methodological Descriptivism” and the second (real) version of “Descriptivism” as “Philosophical Descriptivism,” and then claims that “This argument [“Philosophical Descriptivism”] is not the barrel of drugged trout that Methodological Descriptivism was [ed: yeah, arguments that you make up for the purpose of discrediting your opponents tend to be pretty weak], but it’s still vulnerable to objections,” before moving into “a more serious rejoinder to Philosophical Descriptivism.” The point of all this is that Wallace sets up “Philosophical Descriptivism” as the good kind of “Descriptivism” in order to argue against it. His ultimate conclusion about Garner’s usage guide is that it’s great because it makes a good argument for “Prescriptivism.” For example, when Wallace claims that Garner is “cunning” because he “likes to take bits of Descriptivist rhetoric and use them for very different ends,” he’s making it pretty clear whose side he’s on. Wallace’s explicit purpose in this essay is to defend “Prescriptivism.”

Now, as part of arguing against the good kind of “Descriptivism,” Wallace introduces a very interesting/bizarre argument involving – wait for it – Wittgenstein (if you were hoping that this discussion of David Foster Wallace was not going to end up involving Wittgenstein, I sincerely apologize). This is a multi-stage argument that’s going to take a little bit of work to get through, but it’s important because, as mentioned, the problem with this essay is Wallace’s approach, and his path through this argument makes that problem particularly clear.

The argument actually starts off rather insultingly. Wallace buries it in a footnote, claiming that it’s “lengthy and involved and rather, umm, dense,” and suggesting that “you’d maybe be better off simply granting the truth of the proposition.” As fucking though. This sort of fake homeyness where Wallace tries to act all casual and friendly while also implying that he’s so much smarter than you that you couldn’t possibly follow his argument is by far his most annoying characteristic.

And in fact, the argument isn’t hard to follow at all. It begins with a discussion of the fact that each of us experiences only our own mental states, and that we have no access to anyone else’s. For example, I’m aware of the particular sensation I have upon seeing the color red, and you’re aware of the particular sensation you have in response to the same stimulus, but between us, we have no way to confirm that we’re actually having the same sensation. There is no evidence that can be collected for or against the proposition that the sensation I have upon seeing red is the same sensation that you have upon seeing green. Oddly, Wallace presents this argument as the ridiculous fantasy of a stoned teenager, when in fact it’s just obviously correct. And it’s not merely correct, it’s part of one of the major problems of philosophy: the hard problem of consciousness. Wallace claims that this is a “solipsistic conceit,” but it’s no such thing; it relies on the assumption that other people have mental states comparable to one’s own.

What makes all this sloppiness on Wallace’s part especially odd is that it has nothing to do with his actual argument, which is Wittgenstein’s argument against private language. He claims that the idea of “private colors” is the same as the idea of “private language,” but that’s clearly wrong. The problem of private experience is that we can’t verify each other’s experiences, but we obviously can verify each other’s language choices; that’s what the entire essay is about.

And it’s also what the argument against private language is about. Now, I’ll admit that I haven’t read Wittgenstein, and I’m not claiming to know all the issues and implications here, but I think I can get through the basic argument without butchering it too badly. Basically, the only way we have to verify that our language usage is correct is to test it against other people. For example, if you and I are listening to a piece of music and I start talking about “arpeggios”, and you respond by talking about the same part of the song that I was thinking about, then I know that we’re talking about the same thing. If I were just talking to myself, I could pick an arbitrary aspect of the music and call it an “arpeggio,” but I would have no basis to call this a correct definition. Specifically, if I came up with a set of criteria that defined an arpeggio, and then picked part of a song and tried to decide whether or not it contained arpeggios, it would be impossible for me to be wrong regardless of which way I decided.

The only thing that can “disprove” my definition is another person. For example, if I start talking about the arpeggios in a song with someone else, and they say “what the hell are you talking about, that’s not an arpeggio, it’s a trill,” then maybe I’m using the wrong word. Of course, it’s also possible for the other person to be wrong, which we could find out by verifying it with more people and eventually achieving a common consensus about what the word refers to. Thus, language, as they say, takes a village.

And this is precisely where Wallace goes with this argument: “if the rules can’t be subjective, and they’re not actually ‘out there’ floating around in some kind of metaphysical hyperreality . . . then community consensus is really the only plausible option left.”

Now, this is starting to seem pretty strange, right? Wallace claims that “community consensus” is the only way that correctness in language can be determined, but he presents this as a refutation of “Descriptivism,” despite the fact that the idea that consensus rather than authority is what determines correctness is precisely what “Descriptivism” is. This, presumably, is what has fooled some people into thinking that Wallace is making a “Descriptivist” argument here. But, as mentioned, Wallace is arguing in favor of “Prescriptivism” as embodied by the particular usage guide that he’s favorably reviewing. There’s only one way to reconcile this, and it’s fairly disturbing: Wallace thinks that authority is the best way to establish consensus.

This argument obviously has implications that go far beyond whether splitting infinitives is cool or not, which is where the political aspect of all this comes in. But before getting into that, I want to make it absolutely clear that this is in fact the argument that Wallace is making, since at first blush it seems kind of implausible, not to mention borderline fascist.

Wallace’s ultimate conclusion in this essay is that Garner’s usage guide is awesome because: “Garner structures his judgements very carefully to avoid . . . elitism and anality.” “His personality is oddly effaced, neutralized.” “Garner’s lexical persona kept me from ever asking where the guy was coming from or what particular agendas or ideologies were informing [his judgements].” “Garner, in other words, casts himself as an authority not in an autocratic sense but in a technocratic sense. And the technocrat is not only a thoroughly modern and palatable image of authority but also immune to the charges of elitism/classism that have hobbled traditional Prescriptivism.”

Question: why is any of this a good thing? Why is it good for Garner’s persona to be “effaced” and “palatable?” Why is Wallace happy that a person with an unknown agenda is trying to tell him what to do? Why is it a good thing that Garner is immune to charges of elitism? What if such charges would be correct; what if he actually is an elitist? What if traditional “Prescriptivism” was justly hobbled?

The reason these things are good for Wallace is that they prevent Garner from being undermined as an authority. The question of whether or not Garner deserves to be treated as an authority is, for Wallace, irrelevant. Wallace explicitly celebrates Garner’s book as “basically a rhetorical accomplishment,” meaning that its actual correctness is beside the point. This is the core of Wallace’s support of “Prescriptivism”: the conviction that there must be an authority, no matter what.

(Incidentally, Wallace recommends Garner as a “technocrat” on the basis that’s he’s a lawyer. Is there any dystopia more nightmarish than a technocracy run by lawyers?)

Wallace claims that authority must be earned, but he doesn’t back that shit up. While he does allow that “Prescriptivists” can be wrong, his only actual objection to “Prescriptivism” as an enterprise is that it has a bad public image, hence his celebration of Garner’s purely rhetorical accomplishment. The only thing that “Prescriptivism” ever needed was a image enhancement, and that’s what Garner has supposedly provided.

To be clear, there’s nothing wrong with giving advice on language use. In fact, I’ll give you some right now: when using a pronoun to refer to a person of unspecified gender, you should use “they” in all situations where it’s not completely confusing, and use “she” otherwise. The reason for using “they” is that it already has some traction as a gender-neutral singular pronoun, and since it’s basically impossible to artificially introduce a new word as a basic part of speech like this, “they” is our only shot at a gender-neutral singular pronoun. The reason for defaulting to “she” is that literally everything else in society is male-default, so this is one easy thing you can do to help turn the tide, even if it’s a purely symbolic gesture. You can’t just always use “she,” though; if the gender of the referent is specified, then you need to use the appropriate pronouns. Even if you’re a gender abolitionist, denying the gender identity of actually existing people is actively harmful.

What you’ll notice about the above advice is that I have an agenda. If you disagree with my political goals (e.g. you don’t support absolute gender equality), or you agree with my goals but think my tactics are misguided (e.g. you don’t think gender-neutral language has any practical effect on equality, or you think you have a better way of going about it), then you have reason to disregard my advice. If, on the other hand, I had presented my advice as merely being a “rule” which you are obligated to follow in order to write “correctly,” then your only grounds to object would be “nuh-uh.”

Thus, by supporting this second type of advice, and by supporting Garner’s “invisible” persona, what Wallace is actually supporting is arbitrary authority: correctness without values.

Let’s back up a bit and look at how Wallace gets here. First, he establishes an oversimplified framework for the basic issue using commonplace ideas that everyone’s already comfortable with (“Descriptivists” = “liberals” = “rules are bad” vs. “Prescriptivists” = “conservatives” = “rules are good”). This makes us feel like we know where we’re standing and prevents any deeper aspects of the issues involved from rising to the surface. As one example, the post-war equality movements contained a radical approach to language criticism that went well beyond merely rejecting prescriptive rules. A number of feminist dictionaries were produced that explicitly challenged the patriarchal assumptions embedded in everyday language. This is the sort of real political challenge that demands to be addressed directly, but Wallace is having none of that. He’d much rather survey the general landscape from 10,000 feet in the air.

Furthermore, Wallace actually does recognize that his liberal/conservative split is an oversimplification, but he fails to address the implications of this. Specifically, he claims that “Political Correctness” (there aren’t scare quotes big enough) is an example of “liberal Prescriptivism,” which is entirely correct. As in my example above, “Political Correctness” is “Prescriptivism” with an agenda. But Wallace goes on to merely issue the standard complaints that “Political Correctness” represents “a whole new kind of Language Police” (this is getting to be painful to type), without realizing what this means for his argument. What it means is that “Prescriptivism,” like any other type of statement about what’s “right” and “wrong,” is always based on a value system. Thus, the kind of disinterested “Prescriptivist” that Wallace sees in Garner cannot exist. Everyone works from an ideology, and Garner is no exception. The fact that Garner seems not to have an agenda is precisely the danger: an invisible ideology is one that cannot be challenged.

An important aspect of Wallace’s oversimplification is the way it allows him to pay lip service to inconvenient arguments without actually following through on their implications. Significantly, Wallace admits that “traditional English is conceived and perpetuated by Privileged WASP Males,” but note the cutesy way he phrases this, and they way he puts the admission that this claim “is in fact true” in a footnote, as though this were an extraneous point. In fact, it is the very heart of the issue at hand: since we know for a fact that existing authority is fundamentally oppressive, Wallace needs to do a damn sight better than recommending an authority on a merely rhetorical basis. This realization should have caused Wallace to rethink his entire argument, but presenting it as a side note allows him to glibly glide by.

Second, Wallace busts out his trademark intellectual fireworks by bringing up as many Big Ideas as he can, as fast as he can, which naturally results in all of them being underanalayzed and unconvincingly connected to the actual issue at hand. In the Wittgenstein example above, we saw Wallace segue between unconnected ideas without justification, fail to place those ideas in their proper philosophical context, and finally draw a conclusion entirely at odds with the arguments that were supposedly used to support it. Importantly, Wallace’s conclusion is the result of smuggling in an unstated ideological premise: that authority is required for establishing a consensus. This premise is precisely what Wallace needs to argue for here in order for his conclusion (that an “expert” usage guide is required to resolve the “usage crisis”) to hold, but his fancy footwork allows him to get away without even mentioning it.

Finally, having avoided the implications of the relevant arguments while sneaking in his own unconscious preferences, Wallace is free to drift down onto a comfortable conclusion. Sure, the issues of language may be “complexly political,” but all that’s really needed to resolve them is for a Reasonable Man to write a new usage guide that everyone can agree on. Isn’t it comforting to know that things really are that easy?

The political implications of all this should be pretty clear by now. Wallace’s proposed solution to the “authority crisis” created by the post-war equality movements is, by analogy, the same as his solution to the “usage crisis” created by “Descriptivism”: the reestablishment of a credible authority – any credible authority. This lends a rather dark irony to Wallace’s concluding statement that Garner’s approach is “about as Democratic as you’re going to get these days.”

Wallace is, of course, wrong. The solution to the “usage crisis” is for people like Wallace and Garner to get off their fucking high horses already. Which, recall, does not mean that they aren’t allowed to give advice. It means that they have to stop pretending that their preferences are rules and open themselves up to ideological challenges based on the values they hold that inform those preferences. Analogously, the solution to the “authority crisis” in politics is not to establish a kinder, gentler ruling class. The solution is to finish what’s been started.

(Part 2)