Author: Ron Brooks

In 1970, Ken Macrorie was approached by the journal Research in the Teaching of English. They asked him to review a study. Its purpose “was to find out if computers could analyze and evaluate students’ essays as well as teachers do” (4). To his surprise, Macrorie discovered that on issues of organization, the computer performed as well as any teacher. It also matched the teachers on style and mechanics. When writing about this program, which he named Percival, Macrorie noted that it knew how to recognize a good essay as one with “a variety of sentence structures, frequent long sentences (with dependent clauses and other clearly realized relationships, a title (many papers did not carry titles), frequent paragraphing, few apostrophes, few spelling errors, many connective words, many commas and parentheses marks” (4)

If what Percival can analyze is all we are looking for, Macrorie argued, then we’re in for a really dull time as writing teachers. Macrorie spent the remainder of his career creating classroom environments that allowed students the “freedom to follow the direction of one’s own thought” coupled with “discipline of engaging others’ responses to it.” Macrorie’s classes were workshop-driven, spontaneous, and creative. They were meaningful in both content and genre. By all accounts, they were driven by this two part principle of freedom and discipline.

Macrorie’s optimistic vision had an effect on the field of writing studies and on the teaching of writing. However, on the whole, his nightmare has come true. Whether graded by humans or by machines,  a whole bureaucracy of assessment that evaluates student writing based on organization, style, and mechanics has been imposed on the primary, secondary, and post-secondary writing landscape. Even though the field of writing studies continues to look for ways to emphasize process over product, this bureaucracy still constitutes the majority of what students do when they write “in school.”

Now, 55 years later, we confront the offspring of Macrorie’s nightmare. Large language models like ChatGPT (which I’ve named “Chad”) excel in writing well- structured essays. They demonstrate an impressive degree of sophisticated syntax and grammatical clarity. However, they have little to no mastery of content or personal voice. In other words, the tech industry has given us a machine that can generate exactly the kind of bullshit that automated scorers, both human and nonhuman, value. 

And this value undermines what it means to write and to teach writing. Sure, LLMs can take care of the dirty work of generating emails, proposals, reports and more controversially (at least for us) drafts of rhetorical analysis essays, but this fact only helps a certain kind of user, a person who knows  that what it’s going to produce is a kind of bullshit. “But, hey, if it gives me what I need, what difference does it make? If this is all you want,  an empty response, report, or surface level analysis of a Nike ad using the terms ethos, logos, and pathos, then why shouldn’t I press the button?”

But assuming that we have not already jumped to the conclusion that AI use is inevitable, we are likely still looking for ways to reinscribe a now lost process in the teaching of writing, not necessarily recognizing that students already carried around in their minds preset structures that worked exactly like Chad, structures like the five paragraph theme that can generate drafts without much thought, which is just to say that AI only underscores the reasons that administration fails to understand the teaching of writing.

Administration has always failed to understand the teaching of writing because even when they do understand, someone above them can’t afford to! (I’ve been an administrator. I’ve confronted this impossibility firsthand!) But now we find ourselves, as writing teachers, looking at arguments from above like “good use of AI is critical to the success of our students and to us” and we are asked to teach the good use of AI in addition to the teaching of writing.

The truth is, though, that the use of AI, or even non-use of it, is not the point. The point is giving students the freedom to follow the direction of their own minds and the discipline of being accountable to the other. The first half (freedom) is not all that difficult for administrators, though in the age of AI it raises problems for teachers that I will write about later.

The second half (accountability), can only be brought about when a student shares a draft with a real body (or bodies), and it’s feedback from a real, present body that matters, whether there is AI use or not. But the problem that administration doesn’t want to say out loud is: presence is expensive.

I was a guest on Pods Against Tomorrow. Many thanks to Hugh Manon and Abe Doubleday-Bush for a fun conversation about After Dark, My Sweet.

gen AI is not that kind of problem for writing instruction

A Matter of Words

Why I’m Still Not Convinced that a GenAI Future Is Inevitable

I’ve been presenting about AI lately. Here’s a link to my text-heavy slides. (This is mostly for people who are attending one of my presentations, but if you’ve landed here by other means, you are welcome to take a look! I’ll write up a more typical post about AI once my presentation schedule settles down.)

I recently took a sailing class. It started with two 2-hour zoom sessions (Tuesday/Thursday) and finished with two 8-hour sailing sessions (Saturday/Sunday).

When learning to sail, there is a lot of terminology (declarative knowledge) and there is the actual practice of sailing itself (performative knowledge).

Obviously you can’t sail on Zoom, so what we learned on Tuesday and Thursday was a lot of terminology, as well as the basic procedures of sailing. I did have a little prior experience with sailing, but had hit a wall because I really didn’t have any idea what I was doing. So during the Zoom sessions, I worked on memorizing the terminology (the parts of the boat, the various riggings, etc.) but didn’t do so well. And even though I had been on the water before, I still had some trouble visualizing what the instructor wanted to teach me about manipulating the sail in relation to the wind.

At the same time, though, I was glad to go over a few things before going out into the water. I think I would have been afraid had we just been told to head out without any declarative knowledge whatsoever.

As a teacher, I sometimes have trouble being a student because I’m often thinking about how the class is being taught at the same time that I’m trying to learn as a student. And so while I should have been listening to my teachers online I found myself thinking about how most of my career I’ve tried to find ways of giving students experience writing and terminology at the exact right time.

I often fail.

Why? Because there is no exact formula for when performative and declarative knowledge will meet up in a student. They meet differently for different people at different times. And for writers, no two days are exactly the same, and if my limited experience with sailing is any indication, the same is true for sailing.

So I guess the connection I want to make (and this is mostly for me but I hope you find it interesting too) is that we can’t fully abandon either mode of knowledge about writing. The turn toward threshold concepts is useful for our field, though only as useful as knowing the parts of a boat and sail are to sailing. There is no substitute for getting out on the water and watching what happens when you turn the sail in various directions in the wind.

My WRIT 319 “Technologies of Writing” class and I went back through Gregory Ulmer’s Internet Invention this semester. I decided to write a new widesite along with my students. You can find the navigation to this site in the upper right hand corner of the page.

I enjoyed going back through the process, and I think my students did too.

This was a talk I gave at the Rhetoric Society of America Conference a few years back. An audio version of the talk is here.

In summer 1989, couch-surfing cross country, I picked up this CD, knowing nothing about the composer or the composition, but noticing it was a long piece probably suitable for driving long stretches of road. This accident, and all the variables surrounding it, changed the way I heard music, the way I lived in and saw the world. 

Over the years, few of my room-mates, friends, girlfriends, or casual acquaintances I’d sometimes corner at parties, shared my enthusiasm for his work. Now, my wife doesn’t like it when I play him in the house, saying “I already have a voice inside my head that goes on like that. I don’t need another” and even some of my musicologist friends either dismiss him as ambient music or claim that I’m listening to him wrong: his is the kind of music you leave on while you’re doing other stuff around the house. It’s music to be heard, not listened to. 

(Note: this perception has changed somewhat, recently, after Robert Ashley’s death.)

 So on the release of Rickert’s Ambient Rhetoric and this panel’s suggestion that we might put something together on sound, I thought I would take this opportunity to dig into whatever scholarly work had been done on Perfect Lives. The first piece I found, opened with the following: “of the many things that can be said about _Perfect Lives_, the most important one to consider is that it is a failure.”

So at least for the purpose of this presentation I am putting the music criticism aside and focusing instead on what I see as the generative potential of this opera for work in object oriented rhetoric, by describing a few of its key scenes, not on what they mean but how they make meaning. It is my hope that this description will carve out a space for the opera to be considered as a unique form of epideictic persuasion. 

To enter the world of Robert Ashley is to enter a world of oftentimes commonplace rhythms, spoken observations, vocals that do not attempt to distinguish themselves in vocal registers but simply blend into the musical landscapes surrounding them. Listening intentionally, though, one finds layers of unexpected complexity: musical and verbal cliches coupled with profound insights. 

Perfect Lives is an opera centered around four characters, who hatch a plan to steal all the money from a small town bank in Illinois, carry it over the border into Indiana, drive the money back to IL the following day, put it back in the bank vault surreptitiously, and then let the entire world know that it was once missing. With this high stakes, conceptual art project at its center, the opera weaves together discussions of teenage love, marriage, Tourette’s syndrome, and selfhood with diverse topics like Giordano Bruno, boogie woogie, immigration, language, and the end of human thought–all the while performing invented pop songs, self-help treatises, marriage ceremonies, and philosophical dialogues. 

But what is key is the way that the opera makes its meaning. Traditionally, opera works from a grand central narrative with high drama, elaborate sets, and a libretto meant to emphasize human subjectivity, emotions and decisions. Perfect Lives, by contrast, levels the human with its environment (has the environment become a character in its text). For example, in the episode “The Bar,” the character Buddy, a writer of teach yourself boogie-woogie books, and the character Rodney, a bartender whose daughter just happens to be more or less failing to learn boogie-woogie from Buddy’s books, meet up by accident. While traditional opera would set this as a conflict, this opera presents the encounter in an idiosyncratic way, by leveling human and nonhuman relations. That is to say the performers sing/speak the words of a standard boogie-woogie lick:

CCEE CCGG CCBB CCGG CCEE CCGG CCBB CCGG ALWAYS. 

This boogie woogie lick, which represents for the character Rodney “Industry,” is an example of what Rodney believes separates us from the animals. Rodney’s view is challenged, though, by the composition of the piece, when Ashley juxtaposes the boogie-woogie lick with what he refers to as the “cowlick:”

But this leveling is not obvious to Rodney, who holds on to his subjectivity and his self throughout the narrative: “It does not occur to Rodney that he could lie down with animals. The bar is bad enough.” That is to say that Rodney’s character is haunted by his own lack of subjectivity, that he is always on the border of noticing his lack of it.  

[Following Rickert’s discussions of the relationship between theorists of language and human subjectivity it could be argued that Robert Ashley represents Heidegger’s sense of being in the world, while the character Rodney represents Kenneth Burke. While there is not time in this presentation to unpack this possibility, it is sufficient to say that this opera engages this kind of leveling.] 

There are numerous other examples, notably the discussions and performances of Tourette’s syndrome (a condition that Robert Ashley lived with and which he used to generate an entire opera “Automatic Writing”), but due to the relative weight of that topic, I want to focus on just one more example in the character of Dwayne, who throughout the opera struggles with the dilemma of not being understood. 

Like the imposition of the cowlick in the Bar, in this particular episode the Aria comes at what would normally be a moment of peak conflict, when a preacher asks if there is any man or woman in the congregation who knows for any reason why the couple he is marrying should not be joined. Duane uses this moment to say the following:

Duane’s Aria

Dwayne, my name is three sounds in one word or

Three hills or bumps, a kind of inter-something, where you can’t get  

Them apart. I keep saying to myself: Dquayne Djuayne  

Duwayne Dowayne Dewayne Dhwayne Dwayne. 

I can’t figure out how they got together. 

This attempt at stripping language of its signification through repetition is nothing new, as numerous examples can be found in literature, but what may be new and what is certainly a thread that runs through Ashley’s work is his figuring of this act in the position of aria. 

The open question, then, what does this have to tell us about rhetoric. My focus on these two details has led me to believe that despite what some music critics say to us Ashley should be listened to if for no other reason than to experience the performative effect of human’s being brought to the level of objects. As a genre, opera can perform what happens when characters and objects are leveled in a way that theory alone cannot. Of course, that’s essentially an argument that this opera should be studied. Enjoyment is another matter altogether, and perhaps it’s glib to want to articulate reasons for my own enjoyment of this work. But at least on the level of theory, even (maybe even especially) rhetoric will have to account for enjoyment in order to be complete. I suspect the path to this would be to artificially construct systems which illustrate to us what it means to be part (to use Rickert’s terminology) of a rhetorical lifeworld made up of “elements that lie beyond what humans can do or say” (213).

During the pandemic, Lisa and I moved west, to the Mount Olive, NJ area. Apparently I’m just a few miles from Kenneth Burke’s farm, where he wrote a good bit of his later work, but I haven’t figured out exactly where it is yet, and whether or not any of his papers are nearby. I’ve just mostly been trying to work peacefully with the bear that likes to hang out in my yard.

Twice a week, at Montclair State, I teach a hybrid class, where I teach some students face to face at the same time that I teach others on Zoom. I have a mask and a headset. Sometimes there are a few students in the physical classroom. Other times the classroom is empty and everyone shows up on Zoom.

All of this has me thinking about where “the university” actually lives, as student, teacher, and administrative versions of “the university” seem to have been brought into stark collision during all this.  I guess I want to update Pirsig’s chapter about people’s confusion of the University and its location, as it very much applies to what happened this fall with reopening. Just as the church is different than its building, the University is different than its location. That has always been true. But I wonder if the possibility that university classes can happen online hasn’t complicated that issue in different, more difficult ways. That is to say the forces that have always believed they own the university (trustees, admin, governors, and so on) fought this past summer to make sure that the location IS the university because that’s where the profitable machinery runs.

Faculty, on the other hand, can impart (to some degree) the University from anywhere, but of course we want to get paid for doing so, which binds us to the machinery of location in ways that some of us do not want to acknowledge. (Students have bought in to the University as location, too, because, well, that’s where the fun is at). 

I don’t know what else to say about that right now, but I’m interested in the tension that schools like Harvard are experiencing versus aspirational schools like mine are facing and maybe how that would graph in Lacanian terms. University discourse is unconsciously supported by the master, sure, and the master has been making its long arc toward capitalism for quite a while. But (And?) there are endowments that allow schools to just go online (because after all even the very rich can’t afford to take a gap year from Harvard and be in competition with the next rising class) and there are schools like mine whose endowments would be done in less than a month if they did not offer the promise of university as location.  

In any case, that’s what I’m thinking about this morning as I prepare to go into my campus classroom to stream a class to college students on Zoom. As the weather turns colder, the campus becomes emptier and emptier. Each student seems to decide that the university lives in digital space, and that seems to be a day by day decision.