Whither the humanities

Liberal arts, the death of universities, the age of AI, and why I teach what I teach

The university sector is in crisis, humanities are under threat, AI is killing our ability to think, tech has ruined us, and the best life is found by unplugging from modern existence.

That is a kind of doomerism, but it is also founded in a whole range of things that are true.

What are the ‘liberal arts’?

There’s two ways to think about liberal arts. They are the following: the range of disciplines that belong to the free (man). That is a definition that stretches back into the classical world, to the Roman and Greek education traditions around artes liberales, the kind of education that a freeborn citizen of the city-state ought to have. Key, then, is that education in the ancient world equips you for public debate, civic engagement, serving in the legal system, participating in the military, and having a fundamental grasp of logic, mathematics, music, etc.; and most of all, as that tradition developed, it was based on classical rhetoric, the study of literature, and philosophy.

The way this tradition evolved in the Western world was essentially tied, as it was in antiquity, to class and wealth. The upper classes, and those that aspired to them, pursues this kind of education.

Schooling, in this view of things, is preparation for life as a citizen, and so that really leads into the second definition. Liberal arts are those that can make us free. They are in fact liberalising.

I think that is a more robust way of thinking about liberal arts, and the humanities in particular, because the humanities are ultimately shaped around pursuing the big questions of life, and engaging in dialogue about those questions, and bringing us into the ongoing conversation of (traditions) human beings about the nature of life, its meaning and purpose, the good, the beautiful, and how we are to live. Those are all things with intrinsic value that cannot, and should not, be reduced to ‘preparing for the labour market’.

The nature of threats to the humanities is multi-dimensional. There has, and always will be, the threat of economic rationalism of one sort of another. That threat sees education primarily as one cog of our economy. Education’s primary goal is to enable people to gain job-worthy skills, so they can go and do labour, and earn money, and so it’s a trade: someone pays money (you or someone else on your behalf) to upskill you, to be more useful to employers. Furthermore, it makes sense then to provide education of the kind that is profitable, both to students, to employers, and to the university (or similar), because the university itself has become a business.

The effect of neo-liberalism on universities has been devastating, but its full effects have yet to work themselves out. But it’s happening. Universities turn into businesses, and the “core” function of a university (whether education or research) is ultimately replaced with the bottom line of capitalism, make money. The way to do that is to hollow out your institution, remove power from academics, increase your managerial body, and turn your teachers into adjuncts (just like the gig-ification of the rest of the economy at the moment).

Generative AI is a new threat to the humanities and to universities, because if you can just plug a prompt into AI, and it will spit out something vaguely looking like an answer, or an essay, or research, what point is there in getting an education? Who needs to read Plato when I can just ask chatGPT or Claude to summarise it for me, make an argument, write an essay, earn my degree?

But AI is also just one piece of a larger and on-going set of troubling developments in modern life. I think this piece on ‘limbic capitalism’ captures it well. We live in a world designed to addict us, and technology, devices, apps, etc., are all highly effective means of one simple thing: capturing our attention and destroying our capacity for attention. I have felt that very acutely in my own life, and I continue to strive to set my own life up in a way that I am not driven to distraction and can focus for sustained lengths of time. I do not always win, and I feel like I am clawing back to where a younger, less-internet-attached, version of myself was. I sincerely pity the younger generations in this.

There are, as I see it, two sets of interrelated questions that require answering. The first is, what’s the point? That is, what positive case can be made for the humanities as a discipline.

The second is how? That is, how can we build a robust way forward to practice the humanities as a set of disciplines, in a way designed to resist these forces.

There are two types of arguments for the value of the humanities. N. Ángel Pinillos outlines them succinctly in his piece here. The first, which he is drawing from Jennifer Frey, is that there is an intrinsic value to the humanities. Pinillos agrees with this but criticises Frey for relying solely on it. I think he’s right in this regard. The intrinsic value of reading and debating Plato is only really valued if you already value it. It may be true, but it’s unconvincing to a broader public. Pinillos deploys a second type of argument, that the humanities are valuable because they teach something that cannot be acquired in another way, and which AI cannot and does not replace, and in fact demands: the ability to read carefully, understand arguments critically, and make cases persuasively. The practice of reading pages and pages of texts and then being required to generate your own pages and pages of text, produces critical thinkers. AI short-cuts that process, but your ability to assess AI outputs, to argue with them, to see through the illusions precisely requires the kind of training that AI usage itself cannot give you.

When I first seriously got ‘into’ teaching classical languages as living languages, it was (and remains) something of an outsider movement. That is, all the primary initiatives around speaking Latin, and the much smaller movement to speak Ancient Greek, happen on the peripheries of academia. They are organisations, associations, small groups, clubs, etc. If it happens in institutional spaces, it happens primarily in schools, almost never colleges or universities. There have been some small inroads into the institutional world of tertiary education, but it remains very much ‘outside’ the gilded cages of academia. Most academics still look on the idea of speaking Ancient Greek or Latin as quaint, silly, a hobby, and not relevant to the serious business of mastering grammar and memorising vocabulary. They are wrong.

There is, however, a bigger question about the place of humanities and its relationship to universities as we know them. The duality and tension inherent in tertiary education, that is whether it is about forming free citizens or whether it’s about creating job-ready employees, stands at the nexus of the crisis around AI, humanities, and the ongoing place of the university. As Jared Henderson writes, if humanities degrees ‘certify’ critical thinking skills, and “if AI lets you cheat your way through college… degrees can no longer serve this signaling function…. In short, having a college degree doesn’t mean you learned anything that will make you more valuable to a future employer.”

RIP college degrees.

But if universities will not, or cannot, sustain the humanities, and if the humanities are valuable in and of themselves, then “we need to find ways for these disciplines to thrive outside of the university.” And I think we are seeing that happen already, to some extent.

Henderson thinks that looks like ‘secular monasteries’, isolated communities dedicated to research and labor. I don’t know if that’s a viable or likely vision. It is worth entertaining the idea though. It’s an idea that somewhat echoes anti-tech calls like Molly Worthen’s ‘Why Universities should be more like Monasteries’ (NYT, paywalled), or more recently and in a Christian vein, Nadya Williams, ‘Instead of Embracing AI, Universities should go Medieval.’ Low or anti tech, high on relationality, small in scale, with a renewed focus on what education is for. That is a vision that ought to thrive, but it’s very hard to do.

I think, however you slice the pie, that the University as we know it is finished. What we don’t know, yet is what that actually means. I suspect universities will continue to exist for some time yet. Some of them changing their role and function, others doubling down the ‘education as business, let’s make money from students’ road, others existing on as zombie-institutions. It’s what comes next that interests me most.

One such enterprise that continues to hold up a model of post-institutional humanities is the Catherine Project, the initiative of Zena Hitz to take the core of the Great Books curriculum of a place like St. John’s College, and translate it into a post-institutional online set of communities. That, and many other endeavours, give me some hope for the humanities beyond the crumbling college walls.

I predominantly teach people languages, Ancient Greek and Latin, as spoken and living languages. I do this because I think it’s the best possible pedagogy for learning those languages, indeed any languages. The reason most people want to learn those languages though is not because they want to spend their time talking in them in particular, like ordering lattes in Ancient Greek and doing their banking in Latin. They want to read and understand texts. It is still a fundamentally text-oriented goal that drives people to learn these goals.

And in facilitating that, I have a very renaissance humanist impulse, ad fontes. For my part, I want them to read texts, and I want to read texts, and I want us to do that not by reading cliff notes or gAI summaries of texts, but by picking up ancient and not-so-ancient texts, in the original languages, and being able to understand them, discuss them, and perhaps even join in the ongoing conversation, the dialogue, about the things that matter in life.

David Bennet writes here about the emergence of a new humanities in the face of AI, the need to re-engage that deeper question of what a human is, what a human is for, and what is owed to a human being. I resonate with this deeply, not least because of the very (explicitly) Augustinian way he gets there. Late stage capitalism, the rise of Muskism, the tyranny of tech bros, and the general enshittification of our digital age, are all forces arrayed towards the dehumanisation of humanity. Humanity’s need is for ongoing renewal, and we will find that through the ongoing practice of the arts that make us free.

When it comes to teaching and the question of AI, I am never an optimist nor a doomer. I have a sceptical and generally critical stance towards AI; I’m mindful of significant ethical questions about AI both in general and in specifics; I recognise that it is significantly and amazingly useful in some regards.¹ A friend of mine, Tom Keeline, gave me the following line which is echoing in my head as I think through these issues:

If you are working in a factory, you would be foolish not to use a forklift to lift things.

If you are working out in a gym, you would be foolish to use a forklift to lift things.

Discerning how best to use AI is figuring out whether the task in front of you is factory work or gym work. To analyse a text of Greek, provide a categorised list of forms, tally up different words, run parsing on every form, and provide all that kind of data in a usable data-set is laborious work, that can be done easily and well by AI (or even non-AI digital tools), and importantly doing it yourself in a more manual way isn’t going to improve your language skills.²

Asking AI to translate a text, or to read it and summarise it, or do anything of these sorts, is asking AI to work out for you. You won’t develop the muscles you need. The work of learning a language is precisely the work of spending time exposing yourself to comprehensible inputs in the target language. Short-cut that, and you won’t learn a language.

One of the greatest dangers for students of languages using AI is this: it’s only in knowing the language that you are able to evaluate the answers AI gives you. I have played around with AI and languages enough to see this for myself. When AI is wrong, it’s always confidently wrong, and yet if you don’t know enough to know whether it’s right or wrong, you can’t evaluate what AI is feeding you. Which goes back to Pinillos argument above. The skill to evaluate AI cannot be gotten from AI itself (despite what it tells you).

Of course, many will ask whether we need to learn any languages at all, if AI is going to be able to serve as a universal translator for us. I think that if you think the answer to that question is yes, you betray the fact that you don’t really understand the value of knowing a language at all. This, echoing the above argument, is somewhat circular – you need to value it in order to value it.

To make the ancients speak, to engage in the ongoing conversation about what a human is for, and how to live a flourishing life, remains the goal beyond the goal in my teaching. I teach languages because only in learning a language can we approach those ancient texts, ad fontes, and ask them cui est homo?

Photo by Andrea De Santis on Unsplash

1. I suppose I should say more on this. I don’t use AI at all in my writing or thinking work. I am not convinced that “AI is plagiarism” is either the correct or a useful framework for ethical evaluation of AI outputs, though I do think that is more true of AI generated art, to which I am opposed. I think there are significant questions about the ethical use of AI in terms of its provenance, and its use of resources, as well as the financial bubble it is creating, and other related issues.

2. It’s at this point that I think it’s worth asking, what’s the different between a digital set of tools that will output this kind of analysis which is better done by digital tools than humans, and getting AI to do it instead? And I don’t entirely know the answer to that. But here’s what I do know: being able to analytically parse a Greek word is making a human function like a machine, and it has nothing much to do with learning a language or understanding an author.