Philip Sugg | Blog

About a week ago, I took a look outside my bedroom window, aimlessly, and saw a hummingbird. It hovered, then zipped, toward the branches of a nearby tree. For a few seconds I wasn’t sure what I was seeing–if I had even seen a bird at all. Then I saw that the bird was still there, and that it was sitting in a nest it had built on a thin descending tree branch, a few feet from the window.

I’ve been checking on it ever since: looking at it through binoculars, photographing it, watching. If the nest were any closer to the building, I wonder if my presence would bother it. Sometimes I think that the bird sees me, is staring back trying to figure out what I am. Mostly–uncharacteristically, for a bird that is always on the move–it sits.

As I watch this creature stay still, waiting for something to happen¹, I keep coming back to life and the acceptance of risk. The wind blows, the tree bends. The branch bounces and sways, tipping the bird and its nest. It could all fall–sometimes these nests do fall–at any time. This bird survives by being fast and nimble in the air, but it has given all of that up. The nest is camouflaged, but it is out in the open. I have seen hawks perch on this tree. Other hummingbirds approach her territory.

The nest could be a small risk, an average risk, compared to the 500 mile migration it will soon undertake, across the Gulf of Mexico.

On the relationship of bots to text and language:

If I am to understand these bots as “thinking things”¹, then their thinking substance is the text, words.

Yes, bots can ingest images, produce video, and speak back to me with a synthetic voice. But, for the moment, they still stake most of their claim to intelligence on their facility with language. More specifically, on the written text, to which these models are indebted for their origins. Large language models would be inconceivable without the collected work of the textual internet.

Text is the interface that I use to converse with bots, to enter into a back and forth with an Other, and to come to know it.² I type things, they answer back. In short order we have an essay–mostly from the bot’s side. They have the capacity for a speed of reply and volume of text that exceeds anything I could have produced on my own. Intelligence–in the “Turing Test” sense of that word–is an effect, a performance that the machine pulls off. One reason that people credit these bots with intelligence may be that they produce so much text, so fast. You could write a history of Western literacy according to its accelerating rate of textual production: from single characters and words etched on tablets by hours of human labor, to millions of fully-formed pseudo-reports slopped out every second by data centers filled with GPUs.

I wonder a lot these days about how the automated production of text is changing everyone’s relationship to it. Even before the bot explosion, it would have been arcane to point out that writing–the written word– was a technology. In a historical moment with mass literacy and information explosion, ordinary human labor had already produced more than enough text to render it unremarkable. Fifty years ago, even, it would have been ludicrous to attach any kind of fascination to the mere appearance of a text, as if writing itself was exceptional. If anything the ontological status of the text had flipped. Maximalist pursuits like advertising spread text across every possible surface: the empty air, the bottom of the ocean. Writing had moved closer to the status of litter, another form of spoilage by human development. Another definition of nature: the absence of text. Text had become just another efficient vehicle for information, competing with other media.

John Guillory, imagining new prospects for humanities PhDs no longer in the profession:

I argued further in my MLA paper that the best way to accomplish this goal is to introduce graduate students to as many alumni of the system as are willing and able to speak to them about their careers after graduate school. Many of these alumni, we know, did not get tenure-track jobs but escaped the trap of adjunct labor; many are now employed in nonacademic professions. Let us invite them to return and tell us what they got from their experience in graduate school. Many of these former students do not regret having spent time working on a doctorate, whatever the benefit of the credential in their later working life. But to the graduate schools they have left behind, it is as though they disappeared from the face of the earth once they entered new professions. This is a waste, the loss of considerable talent and passion to a diaspora.39

Can these former students maintain a relation to literary study without the organization of the profession, without the structure of graduate school? To ask this question is to put the intellectual seriousness of the literary disciplines to the test. Literary study in its disciplinary form obviously cannot be separated from the organizational structures of the university and the departments of which it is composed. But it is surely within the power of these departments to reconnect with former students and bring them into contact with graduate students currently in the system. To do so would be to enlarge, in small increments, the sphere of intellectuality by tapping the intellectual sociability in the corps of former graduate students. There is no reason why intellectual engagement with literature has to exist only in the form of a profession, however gratifying professional life may be, however abundantly scholarship has thrived within the academy. I gesture here to the realm of what Merve Emre calls the “paraliterary,” all those sites where literary study is cultivated outside the purview of graduate education.40 At these sites one might find long-standing projects such as the “medical humanities,” but the more promising locations in this context are less disciplinarily organized. These are sites (for the most part) of intellectual exchange on the internet, new versions of “little magazines,” such as n+1, or of journals such as The Point, as well as the now vast proliferation of blogs on cultural matters, some of which host high-level exchanges.41 Such sites disclose the widespread desire for an engagement with literature and culture that is more serious than the habits of mass consumption and that demands new genres and forms of discourse.

Read more →

I had a dozen or more books that I needed to return to the UChicago library, so I drove over there after putting my son to bed. Late at night, after all the retail in the neighborhood has closed, most people are at home getting ready for bed. That big library building is still lit up. The university is in session for summer classes but there are hardly any students left. Fifteen minutes before closing, I nearly have the place to myself.

This concept of the ’third place’ gets thrown around sometimes in American settings: a place outside of home (the first place) and work or school (second place) that has some significance for communal life, where people can gather and form a culture outside of family or professional functions. Libraries are one type of third place I’ve known for most of my life. In my case I don’t know if they’ve mostly been about community; certainly they have been that at times, but what might be even harder to find than a gathering spot in the US right now is a place that’s devoted to quiet reflection, with a sense of possibility. A place where exploration can be invisible (or not), and needn’t be tracked or justified to anyone. When has the country ever needed a site like that more?

I don’t know what the fate of libraries will be–even in the near future. I doubt they’re getting more funding this year than they had last year. Large-scale storage of paper volumes is getting harder to justify. Private research libraries like UChicago’s are now coasting off the investments made in a different era.

From a short walk last weekend in a mature hardwood forest, occasional light. A delight to see a slow-growing plant like wild ginger covering the whole hillside–like a form of living time laid out in space. The thin but consistent layer of leaf cover makes all the difference in what can grow and thrive on the forest floor.

wild ginger, a slow grower at about 6 inches per year, spreading a hundred or more feet downhill

Huge stands of Ladyfern (Athyrium filix-femina) wherever the tree cover was heavy. In any window where a tree fell and the canopy broke open, the ferns were burned away

Something in the Osmorhiza genus, likely aniseroot. My eye was captured by the segmented, almost fern-like lobes of leaves on this plant.

In the Driftless region of Wisconsin with my family this last weekend, I had a chance to think about the consequences of its geology. Drift-less refers to the relative lack of glacial presence in this region during the last Ice Age, which ended around 12,000 years ago. Glaciers are one of the great earth-moving machines. Some features, like the rock and sand and dirt that make up mountains, are scraped away. They bring new material to the ground they pass over. In Wisconsin this can be hundreds of feet of sand and topsoil¹. The old land is both buried and crushed. But the original lime and sandstone bedrock is still present in the Driftless–usually beneath the ground, sometimes on jagged slabs lifted above the trees.

Spared the resetting effects of the glaciers, the Driftless retained traces of the past billion-plus years of geological history. The present-day soil sits right on top of old rocks. These rocks are a book of changes over time, a geological palimpsest written by the movements of water. Water nibbles at the bedrock, creating craggy, spindly passageways for itself. _Dendritic drainage, like the roots of a tree. Smaller channels open up, then larger, like an organic network of underground pipes. Water spouts up from these depths, changing the course of streams each year and the contours of the land over decades, then disappears back into the earth, rejoining the subterranean rivers and lakes. The surface is changeable. What has endured is the foundation–rock–on which the land sits. Water is the messenger between surface and depths.

On Wildcat Mountain, in Wisconsin’s Driftless region. This high observation point–from where the photo was taken–is around 1200 feet elevation, no more than a few hundred feet above the height of the surrounding plateau. But I was high enough to gain an awareness of the overall shape of the terrain (e.g., the knob to the left of the photo), to see the pattern of the topography (woods interrupted by open fields) and to see over all impediments, to the straight line of the ultimate horizon. I’ve stood on many mountains higher than this, and here I had the sensation that I was on a peak thousands of feet higher than it was–at great height. What is the term for when one landscape evokes another? The imagination allows itself to be tricked, perhaps out of nostalgia for a more familiar scene or a homeland. Heimatgefühl, the Germans would say.

The trick here is the elongation of relatively modest slice of vertical distance. In a few hundred feet this “mountain” (apparently the locals called it a hill before it was turned into a state park) gives the viewer a sense of detachment from the flatlands. There is a productive loss of connection between the high ground one stands on and the plateau below–a little bit like you’re flying. When I saw this view in person it made me feel like I was seeing the jigsaw puzzle of my surroundings. One doesn’t usually get views this comprehensive from Wisconsin. Looking at the photo now, I am reminded of the title to Rebecca Solnit’s 2013 book The Faraway Nearby. At this paradoxical point in space, we get close to something far away, and become far away from something near. Height creates both sensations at the same time. This is one source of its sublimity. Isn’t this feeling a privilege of rest, of getting away from the everyday, of being in a position to contemplate? Let’s hope that it can happen from more places, from even more unassuming elevations.

Image credit Internaz

I was saddened to see that William Langewiesche, the writer and long-form journalist, has died at 70. His was one of those bylines that I’d learned to get excited about a few times a year over the last decades, especially on topics related to aviation. He began his career as a pilot, and used that credibility to establish himself as the unofficial interpreter of flight–and flight disasters–for educated, literary audiences in the English-speaking world.

Judging by his recent work, he was very much still in the game. But Langewiesche was far more than a writer about airplanes. And what was that more? Not a writer who stuck to a competency earned through first-hand experience, but a writer whose technical skills revealed a more general comfort with complexity. In stepping out of his lane, he performed a demystification of expertise. Langewiesche’s early life as a pilot required an instinctive comfort with detail. But what set him apart was not his command of the details but his drive to understand them, to integrate them into an ever-widening but credible synthesis. All of these traits made him the ideal writer for an increasingly technological moment.

Langewiesche was a nonfiction writer who showed surprising range, but he was not a generalist on the level of, say, John McPhee.¹ In a sense he wrote many versions of the same story. Complexity is one of the defining conditions of the modern age. What happens when it is ignored, denied or misunderstood? This problem has its mythic qualites. It is often a story about what happens when human actors–mortals–lose sight of their limits, when they are reminded of their capacity for error and the ever-present possibility of the rare event.

Text-generating AIs get a lot of attention for their (apparent) capacity to outperform a human subject at any number of useful analytical tasks. An equally demanding challenge might be to persuade an audience that AI-generated ‘artwork’ was produced by a human. All the image generators I have seen might be skilled in some realistic or technical sense. Still, by now the visual tip-offs that an AI was involved are increasingly well-known. The specifics differ somewhat by the model, but the overall AI ‘house style’ for visuals is–to my eye–quite recognizable and consistent at this moment. Harder to explain is this: how does one recognize a piece of art as the product of a human? The human is what is left when all known AI styles are subtracted from the world.