(This blog post is a companion piece for my review article on embodied cognition for Evolutionary Studies in Imaginative Culture. It was originally published here.)
If the personal history behind a research article is sometimes useful to have, its value is less obvious for a review article. Reviewers, after all, should be objective, and even if this can never realistically be achieved, it still seems a little shameless to parade the deficiency. I do so here because I think that the different ways in which I have encountered 4E perspectives illustrate some tensions in the practice of 4E that are usually invisible in the academic literature. These tensions are important because they illustrate how the presentation of 4E as a radical intellectual program by humanities scholars (though not only them) can run ahead of the more prosaic reality of actual 4E research. Should my experiences not chime with those of others who have worked in 4E, I can only remind them that they have foresworn disembodied objectivity on a priori grounds, and thus have no claim against the reviewer who fails to exhibit it.
My first encounter with 4E cognition came when I was a philosophy undergraduate in University College Cork in the late 1990s. At that time, Mark Rowlands (later to be joined by Matthew Ratcliffe) was a junior member of the Philosophy Department. By dint of a peasant cunning that I have since learned to emulate, his lectures usually tracked whatever research project he was working on at the time: back then, that was his influential book, The Body in Mind (1999). I’d like to say that early exposure to what would later come to be called 4E cognition made an instant convert of me, but to my pals and me—bourgeois Kantian idealists to a man—the whole enterprise just seemed hopelessly positivist. Nevertheless, we eventually conceded that retaw was not water, that thinking did not make it so, and that the intuition that we might have bodies was not an entirely discreditable one.
What was important for me about this encounter was that it naturalised 4E as an idea amongst other ideas. Like the categorical imperative, predicate logic, Rawls’s veil of ignorance, and St. Anselm’s proof, 4E had a claim on my attention—but no more of a claim than they did. The result was that, though I learned about 4E quite early in its wider emergence as an intellectual paradigm, I did so without feeling any particular need to go out and evangelise on its behalf.
This nonchalance fed into my second encounter with 4E, which came in the early 2010s at Oxford after I’d changed my intellectual focus to literary studies. By dint of a late-game conversion to truth, rigour, and the empirical way—and aided by the infinite beneficence of the European Union—I took up a cross-training fellowship at Oxford’s Department of Experimental Psychology. My expectation was that this would be a sort of Hallmark romance, where the two cultures would tearfully renounce their differences and retire to a hotel room somewhere to conclude with other business. My expectations, however, were wrong: Robin Dunbar, my mentor at Oxford, had no room for a Bourbon princess on his staff, and started me off as he meant me to go on—which is to say, with a head-first dunking into the sheep pit of statistical methods and experimental science. Having just worked for four years in a research culture where the normality of errors was understood to be an intellectual license rather than a statistical assumption, this new immersion left me feeling more like the tick than the sheep.
Whilst at Oxford, I naturally made efforts to connect with colleagues in the humanities who were also interested in cognitive science. This was (and never ceased to be) a positive experience. What quickly became visible, however, was that theory-heavy 4E perspectives about cognition were very much in the ascendancy amongst humanities scholars. The contrast with my daily ration of “shut up and calculate!” was, I occasionally reflected, striking, but being busy with hard sums I didn’t dwell on it overmuch. When I began to take more notice was after meeting with two colleagues in literary studies for afternoon tea in one of the more salubrious colleges. My colleagues—who were and remain first-rate scholars—both gently gave me to understand that my methods were unsound: the mind, contrary to appearances, was a body, and anyone with an eye on the outcome might profit by acknowledging it. Naturally, because I seemed like a decent chap, they were happy to slip me a few E’s gratis if I needed them.
OK, that’s not how the conversation actually happened. But what did become clear to me was that the take-up of embodied perspectives by scholars in the humanities, though articulated as an effort to reconcile humanistic inquiry with contemporary approaches to cognition, seemed to be captured by only one of the contemporary approaches to cognition. This was particularly evident when it came to computationally focused alternatives like evolutionary psychology. Say what you like about evolutionary psychology—and people do, all the time—the fact remains that, on point of empirical evidence, its conjectures are vastly better supported than nearly any of the hypotheses associated with 4E. This doesn’t mean 4E is wrong (many of its claims are simply hard to evaluate), but it does make one wonder whether the partialism it stimulates has less to do with intellectual rigour than it does with a modish preoccupation with other people’s bodies.
Either way, my next encounter with 4E made it quite clear what empirical research in embodied cognition might look like when it wasn’t channelling its inner Mujahedeen. Unemployable in literary studies for reasons unclear (or perhaps very clear), I took up a position in the Embodied Cognition Lab in Department of Psychology at Lancaster University. There, I worked as a research associate on the Lancaster norms project: an ambitious undertaking that sought to map the sensorimotor associations of 40,000 concepts in English. This was unshowy, detailed work that required online participants to rate large numbers of concepts on what amounted to a production line. Amidst the daily round of herding participants, debugging code, and ensuring that Cronbach’s alphas didn’t quarrel with each other, it occurred to me that this was the kind of work needed to place 4E approaches to language and literature on a credible empirical footing. One can volunteer as many high-level hypotheses about embodiment and literature as one likes, but without coming in at the ground level and establishing how embodiment can be measured (and on what dimensions), these hypotheses can never be anything more than diverting. Fortunately, the Lancaster norms were recently published, so I await with serene confidence their transformative impact on literary scholarship.
Ultimately, I remain of the view that embodied approaches to cognition are of crucial importance in the project of understanding symbolic culture. Nothing in my experience has changed that, and as the mind is very probably embodied, I expect nothing will. All the same, we create tools so that we may solve problems, and when the problems become secondary to the tools, we are decorating the world instead of mending it. I’ve been fortunate enough to work in both the humanities and in cognitive science, and in my experience at least, 4E remains more an ornament than a tool. If there’s a fourth encounter with 4E in my future, I hope it’s one where the excitement of intellectual enterprise is leavened by a more sober appreciation of the value of practical results.
In February of 1978, the sad young man went away. The algorithm that replaced him could pass for human, when it wanted to. It gave occasional interviews and was the subject of a documentary; by all accounts, it had a family and lived in suburban west London. Last year, it even managed to contract cancer and die. But then, the uncanniest valley of all is the one that hides in plain sight.
Scott Walker’s later music unsettles, not because it’s alien, but because it alienates. It’s perversely fitting, then, that the young Walker—let’s call him Sleep Walker—should have been such an exponent of seduction. You can hear it readily enough in those lush Walker Brothers numbers that gave snaggle-tooth England the flutters. “Love Her”, “Make it Easy on Yourself”, “My Ship is Coming in”: each has the neotenous prettiness of pop before Andy Warhol took it behind the bike shed and gave it something to think about. But even in the solo records of the later 1960s, Walker just swapped the high school for the Students’ Union. “Loneliness,” he revealingly crooned, “is a cloak you wear”, and if he went on to thicken up the mix with some Jacques Brel and Albert Camus, they were just props to make the boomer girls feel deep. And why not? Tuition fees had gone forever, and the existential angst had to be sourced somewhere.
It’s to Walker’s credit that he hated it. Most of us, when we have a skill, prescribe it like penicillin: Jeffrey Dahmer, probably, felt serial killing to be a public good. And Walker knew the moves of seduction better than the inside of his own mouth. So, when he chose to become the mirror instead of the reflection, the idol he was killing was himself.
In the strength of the crime
You sing like a stranger
And your failure fulfils
Your most secret defeat.
Sleep Walker’s death wasn’t quick. We can still hear his murderer at work, 28 years later, on The Drift:
Jesse are you listening?
. . .
I lower my head
Press my ear to the prairie.
‘Alive, I'm the only one
Left alive
I'm the only one
Left alive
I'm the only one
Left alive
Alive.’
Thus does Jesse, the unborn brother, the six feet of foetus, answer Elvis, the teen idol, from his grave beneath the American grasslands. Go listen to it, if you haven’t already; it’s the most harrowing vocal in all of music. It’s the sound of truth murdering beauty.
But anyone will tell you that the second murder is easier than the first. Having killed himself, Walker set about killing the rest of us, too. You can’t be convicted of homicide if there’s no such thing as a human.
KURTZ: What did they tell you?
WILLARD: They told me that you had gone totally insane
and that your methods were unsound.
KURTZ: Are my methods unsound?
WILLARD: I don’t see any method at all, sir.
But there was a method, all the same. From 1978’s Nite Flights onwards, Walker diced our culture into a slurry and fed it back to us. As is so often the case, the first step comprehends the whole journey. Nite Flights. Ingest Filth. (Shitting Elf works too, if you’re reading this to your kids.) And, like the prisoners in Salo, Pasolini’s torture-porn, we ate the shit we were served. David Bowie, famously, could never get the taste out of his mouth. Brian Eno maintained the shit was so good that we’ve never had better since. In Bish Bosch, Walker himself allowed that if shit were music, he’d be a brass band.
When AlphaGo beat Lee Sedol in 2016, it made headline news; in 2019, when Gwern taught GPT-2 to write like Shelley, Reuters reckoned that this one could sit on the spike. Fair enough: only nerds dig boardgames and Shelley was mean to his wife. Still, poetry pays in prestige—it’s not like there’s a Dante of checkers—so you’d think the belletrists would at least have broken a sweat when the central bank started printing money:
My heart, why come you here alone?
The wild thing of my heart is grown
To be a thing,
Fairy, and wild, and fair, and whole.
(Yes, I know: there is precisely no one alive today whose inner life is captured by these lines. Do you think that would have been different in 1818?)
In 2007, Walker made his second release on the 4AD label (the first being The Drift). Who Shall go to the Ball? And What Shall go to the Ball? is an instrumental piece commissioned for the Candoco Dance Company, who perform with a mixed cohort of able-bodied and disabled dancers. Unusually, the release was prefaced with a question:
How much of a body does an intelligence need to be potentially socialised in an age of ever-developing AI? This is but one of many questions that informed the approach to the project.
There’s a whiff of intern prose about this; besides, looking for smoking guns is conspiracy-theory try-hard. Still, you have to ask yourself: how much of a body does six feet of foetus need to be potentially socialised in an age of ever-developing AI? Shelley’s wife might have hazarded a guess. It’s such a shame that GPT-2 cloned the monkey instead of the organ-grinder.
Algorithms, we’re told, encode our biases. This is great news. At the time of writing, it’s still illegal to hack the human cortex, which means we’re stuck with the racist assholes. But you can change whatever you like about the distribution of weights in a neural network; it’s brainwashing without tears. Fair warning: this may take you further than you wish to go. Being human is itself a bias, a statistical pattern, a set of expected values in the long-term frequency of genes, behaviours, and symbols. Shuffle these up and the smart money is always on the result being cancer.
When Walker shuffled the deck, something else came up. Sure, it still looks like cancer; see it standing there, its face pressed up against the glass!
As the grossness of spring rose
A tumour balloon to squeak against the window
With the grossness of spring staining into the walls
But this isn’t cancer as a disease of proteins; it’s cancer as a pathology of information. It’s the cancer of a system trying to reproduce itself and gestating something else instead. This cancer is entropy’s answer to narcissism. Walker didn’t need telling that it’s terminal—after all, it’s how he disappeared.
But we do, because we’re still stuck in the nouveau Palaeolithic and it’s getting dark. What Walker began as an aesthetic now manipulates your vote and manages your pension fund (as if you have a pension!). The difference is that Walker starts with the narcissistic image and spirals out into the darkness; the other algorithm starts with the darkness and spirals into the image. The imposture is just about visible, for now. Give it a little and it won’t be; give it a bit more, and you won’t even care. Look how prettiness mutilated Leonard Cohen—and you think you can survive it?
We won’t survive it; no one does. That’s why SETI only ever detects silence. But if we want to at least try, Walker points the way. The arc of compositions from Nite Flights through to Bish Bosch shows us what we look like when the primate isn’t being flattered. If you think it’s ugly or incoherent, then draw the inference before it hangs you. Because the noose is growing tighter and there isn’t much time.
“The Electrician” is a song about a CIA torturer. In the Vietnam era, the innovation was to attach the batteries from field radios to the testicles of informants. These days, Psy Ops finds that approach to be a little roundabout and goes straight for the music instead. The principle, however, is the same:
If I jerk the handle
You'll die in your dreams
If I jerk the handle, jerk the handle
You'll thrill me and thrill me and thrill me
In February of 1978, Scott Walker jerked the handle and we died in our dreams. We’ll stay dead, if we know what’s good for us.