Welcome to Reasonable Things, an occasional newsletter from Joel Heng Hartse.

Wendell Berry has died.

This makes me very, very sad.

I called Berry “my antidote to grad school” when I was in the thick of my PhD, and I meant it only half-jokingly. Reading Life is a Miracle and Standing by Words while I was knee-deep in navel-gazing theories of language and literacy helped me retain some measure of sanity even as I metaphorically and sometimes literally bashed my head into a spreadsheet with two thousand lines of grammar errors and comments from teachers that I was trying to organize into an argument about what happens when people read writing that they think is bad.

I was especially taken with Standing by Words. Here’s an excerpt of something I wrote about it that I could never quite find a way to publish, maybe in 2013 or so:

In a widening series of circles, Berry describes an individual (in this case, a farmer) rooted in a family; a family, in turn, is rooted in the community. The community is rooted in what Berry refers to as “agriculture” – social life is rooted within the larger human project of cultivating, caring for, and stewarding the land. Agriculture, finally is rooted in nature, or what Berry usually refers to as Creation.

Ultimately, this is all enclosed within what Berry terms “Mystery.” In order for the person to be able to function in this system of systems, and for his words to be meaningful to himself and others, he needs to see his life as a gift, Creation as sacred, and all of it as having ultimate meaning.

In his second diagram illustrating the consequences of industrialization, Berry describes the same farmer in the modern economic model. Instead of a “person” at the center, Berry places the “dairyman,” now an instrumentalized worker rather than an individual. His place is not in the family, but the dairy. His farm, rather than being home, is now a business. This is still rooted in agriculture (humans still need to work with the raw materials of nature) but agriculture, rather than being enclosed within nature or Creation, is now simply a part of and subservient the global industrial economy. Most tellingly, perhaps, “Mystery” is absent from this model. The economy is the final arbiter of meaning and truth.

The linguistic implications of this shift to industrialization are clear to Berry: things are no longer called by their proper names, and words become meaningless as all human action is appropriated by a depersonalized economy. A reasonable view of language becomes strained if a person is not a person, or a family farm becomes argibusiness. The very possibility of coherence in meaning-making is undermined if human lives are understood not as a part of nature and a coherent cosmology, but as cogs in a machine. Berry is concerned that people living in this situation can no longer “stand by words” because words and their meanings are estranged as humans become further estranged from local labor and fruits of that labor.

It is this local life where words ought to derive their meaning, according to Berry, and this is reflected in his critique of the specialization of academic studies of language, such as rhetoric and linguistics, since they encourage a separation of non-specialists from the understanding of their words. Berry argues that the language-using community, rather than scholarly knowledge about language, that must be “promoted” in order for language(s) to flourish. He writes:

“If one wishes to promote the life of language, one must promote the life of the community—a discipline many times more trying, difficult, and long than that of linguistics, but having at least the virtue of hopefulness. It escapes the despair always implicit in specializations: the cultivation of discrete parts without respect or responsibility for the whole.”

Thanks, 2013 Joel; this is 2026 Joel here, and I’m going to talk about AI now.

The great tragedy of Large Language Models is the final undermining of the possibility of standing by words, you might say. While we still live in a world of profound alienation from nature and ourselves, we also now live in a world where the thing that used to only be able to come from us — languaging, or discourse, or seemingly coherent, apparently meaningful sequences of words — can now come into being without a human saying or writing it. The words are real, even their material effects can be real, but we can now have writing without writers, utterances with utterers. It’s pretty f*cked up, honestly.

So here’s what I came to say: the reason I feel like I have “AI psychosis” is not because chatGPT has convinced me I’m a genius or I should kill myself (sadly, both have happened to people, with tragic results), but because I now see AI slop everywhere I look, and I feel like a paranoiac who is unable to trust any writing I see.

Gregory Conti, in “The AI Apocalypse is Already Here,” describes it better than I can:

“This uncertainty about the human provenance of text is an incalculable loss. In both classical and Christian philosophy, the human being was distinguished by his unique relationship to the word. Machines that have command of language are therefore nothing like the mere next step after the microwave and the calculator. The idea that for the rest of my life I will never know for sure if a text was written by the human who claims it or by a machine saddens me greatly, and I suspect this reality will cast a pall over all our lives once it sets in more generally. As things stand, we will face, every day, a task presented to us that no previous human ever could have contemplated: the task of sussing out whether the text we are looking at had a human behind it. This is a huge “transaction cost” that our new technology has imposed on us all, an unwelcome introduction of doubt and wariness and suspicion into the fundamentals of our life together that will only reduce trust and increase atomization.”

This loss of trust is both sad and crazy-making. Many of my own former students, people I know and like and think are smart and interesting, constantly post what I would call AI slop on websites like LinkedIn. Brilliant scholars I know send emails that are run through Grammarly (or maybe wholly generated by Grammarly; who knows?), bulldozed of all personality and original thought. I’ve had my own academic papers “reviewed” by people who were just running them through an LLM; work I allegedly wrote has been “cited” by AI-generated papers confabulating fake papers that do not exist, attributed to me; I’ve had angry, desperate emails from students complaining about injustices large and small that are clearly not written in their “own words”, or perhaps not written by them at all; smart, thoughtful students who know how to write on their own have had their prose flattened into the most facile “it’s not about X; it’s about Y” garbage (aka “the DiGiornio Structure” — “It’s not delivery; it’s DiGiorno!”); less smart and thoughtful students just push the “write me an essay” button and hope for the best — I catch some of these, but not all.

I could go on, and I probably will, for years, hopefully in a less rage-induced, more carefully argued book, someday.

I sometimes use a very good AI detector (against the wishes of my university) called Pangram to check my intutions about whether something has the taint of AI on it, and it almost always confirms my suspicions, but like…then what? I come back to a student and say “hey I think you used Grammarly on this, please don’t,” and they’re like “oh, sorry, I didn’t think Grammarly was AI, yeah I used it to make my grammar better.” I can’t send them a 2,000-word essay about why they’re destroying society by running their writing and ideas through a language-flattening, idea-worsening, trust-undermining machine. (Maybe I should, though?) I have to just tell them I don’t like it, and try to add more specific language to my syllabus next time, and accept that I’m fighting a losing battle.

Berry wrote in 1987 that he would never buy a computer, which seems hopelessly outdated now, but we could not have gotten to the place we have with LLMs without already being accustomed to the disembodiment of our words, with our outsourcing of languaging first to marks on tablets, then to ink on paper, finally to bits and bytes on screens. I’m not saying all of that is bad or evil. Everything I’ve ever written for publication was written with computers. But if I want to live in a world where I can trust that there is someone standing behind words, I’m going to have to start spending a lot less time with them.

♥JHH
Burnaby, BC