Reading the Unreadable Book: Aliocha Coll's Attila

It’s normal for reviewers on Amazon and Goodreads to be confused, annoyed, and sometimes hypnotized by books they can’t grasp:

I cannot rate this book. It is too perplexing and far-reaching and hallucinatory to begin to comprehend. It is its own planet with a gravitational pull. I’m pulled in from faraway space, unable to control my own direction. I will think about Attila for decades, and yet, maybe never truly make sense of it…

Another review of Aliocha Coll’s Attila, by Damian Murphy, includes these lines:

I regard this book as a towering masterpiece. While I don't know if it’s possible to completely comprehend it in the conventional sense of the term, the text is woven with threads of implicit meaning which… are felt more than understood. It has an internal consistency which is multiply mirrored to form strange and magnificent symmetries. The prose is so lyrical that, even in the densest sections, I found myself eager to push through.

It’s less common for the publisher and translator to also be at loose ends. Katie Whittemore, the translator, admits her “total confusion as to the nature of the book.” I don’t remember any other translator who goes so far out of her way to say that she doesn’t understand the book she’s translated. “In the first place,” she writes in her introduction, titled “On Translating Coll,” “I barely understand what I’m reading.” She senses “it is not stream of consciousness,” but the “metaphors and images” are “obscure.” Her ordinary way of working is “constantly broken, violently so, by a sense of bewilderment and frustration.” She encounters “beautful images, startling turns of phrase, hints of philosophical and even metaphysical reflections, explosions and purrs.” But she fears she doesn’t “really get it.” In the end she hopes future readers will be “more attuned” to the “work’s complexities” and “appreciate it in ways that elude us.”

I’m not criticizing Whittemore here. (I’m not competent to do so.) A good translation might well result from an intuitive feel as much as from attentiveness to detail. Rather I’m registering how strange it is that the unspecific adulation and awe that characterizes so much of the book’s reception is present even in the book’s English translation.

1. Javier Serena’s book, also called Attila

Serena’s Attila, which shares its title with the original, has been published as a companion volume by Open Letter. It is an attempt to imagine what a life like Coll’s must have been like. (Coll was unrecognized, and died by suicide after completing Attila.) In an author’s note at the end of the book Serena says he was inspired by the “extreme complexity” of Coll’s works, his “complete audacity and honesty,” and his fidelity to “his creative convictions.” Serena’s Attila conjures the writer he calls Alioscha¹ using a range of ideas about the genius who lives on the edge of self-destruction and whose life is entirely devoted to art (and love affairs, and voyeurism). Alioscha is a “naïve and imaginative man prone to intense and lasting excitement”; he has “fantastical inclinations,” and plunges into “mythical reveries… abruptly, for no apparent reason”; he is emotionally “fragile,” and “awkward” in society. The descriptions are full of pathos and are sometimes maudlin, as in a set piece about Coll taking out his building’s trash while kids in the neigborhood make fun of him. These are all clichés of romantic literature that can be found in Poe, Balzac, Hoffmann, Nerval, Gautier, Baudelaire, and Stifter. There is nothing modern about the way Serena pictures his tortured genius.

It’s appropriate, then, that Alioscha’s writing is absent from Serena’s book. The narrator is “impressed” by Alioscha’s “erudite quotes” but doesn’t understand his writing at all, and says as much. When Alioscha reads passages out loud, the narrator remarks, “I let him recite as many pages as he pleased, unable to comprehend them.” The only thing we hear about the writing itself is that it was “structured around the patterns in specific Greek myths.”

2. The reviews²

Even though Serena’s Attila is nominally fiction, it conforms the tendency of reviewers to romanticize the author and evade the text. Marc Nash’s review on Youtube asks “What is this book about? I couldn’t tell you… You have to hold up your hands when you dive into this book, and say, ‘I’m not going to look for meaning, I’m not going to look for coherence.”

Most reviewers praise the book effusively and then either throw up their hands at the possibility of understanding it, or else assume it is best understood as poetry. John Repetti prefers to consider the book as a whole, for what it signifies about the future of literature. In all these ways the book goes largely unread.

A few reviewers, like Noel Thorne here on Substack, reject Attila as “art-fart bullshit”:

It’s the kind of writing where you read a sentence, a paragraph, a page, and have absolutely no idea what the author’s banging on about. Think Finnegans Wake. Ugh.

Coll himself apparently said that Finnegans Wake is a challenge to all writers. I don’t disagree, but as in the case of other writers who have said the same thing, it is difficult to see how Coll thought he was responding.

The most substantive review in English is in The Untranslated, and I find it’s often cited by other reviewers who don’t want to guess about what the book means. The review offers two structural metaphors: Mudéjar architecture and the Sierpinski triangle. As the reviewer no doubt knows, the Sierpinski triangle was suggested by David Foster Wallace to Michael Siverblatt as the model for Infinite Jest. But it’s no more useful in this context: it’s only a way of pointing to an effectively infinite fractal complexity—which is a way of saying that the principles of form in Attila remain unknown.

3. Subtracting the value accorded to erudition

A first step to assessing the book is to set aside praise for the author’s erudition. Whittemore is impressed, for example, by Coll’s vocabulary. “I’m forced to work very slowly,” she says, “looking up more words than ever before as a long-time reader of works in Spanish.”

Erudition in a writer is, I think, a distraction, because it is relative. Readers have been impressed by Richard Powers’s research into science, Cormac MacCarthy’s understanding of physics, Geoffrey Eugenides’s discussions of molecular biology, Robert Musil’s study of mathematics and mechanics, and the wide reading of authors like Leopardi, Canetti, Eco, and Schmidt. But what appears as mysterious mastery to one reader is the day job of another. Learning and erudition are relative. Some writers try to impress readers with their knowledge, like Énard or Nooteboom, but there will always be readers who know the material they showcase, and for them, the gesture won’t work.

Coll uses a wild variety of words from Greek and Latin and from modern science, some of them neologisms and others rare or applied in unusual ways outside their normal uses. Readers have also been awestruck by the range of cultural allusions, in particular to myths like Laocoön.

A helpful way to proceed is to look at the origins and range of uses of each unfamiliar theme or word, see where the author got it, and what breadth or depth of reading might be inferred from the usage. Read about Laocoön, not to see the story, but to see what sources Coll used and where he invented or changed the story. There’s a difference between an unapproachable mastery of mythology—as in Robert Graves, Arnaldo Momigliano, or Theodor Mommsen—and an author’s selective reading. Find the books read by the author, and you find the author.

Coll was an opportunist: he collected narratives and words from classical texts and terms from modern science and mathematics, and used them to create a new and unfamiliar language that would conjure an undefinable time and place—at least that is what he wanted. The result, for me, is a fanciful collection of themes and words from different contexts, jumbled together. Knowing the historical uses of the words makes it impossible to read them as elements of some timeless, ancient, or mythical language: they read as the collection of a twentieth-century writer interested in classics and popular science.

4. Books Coll read

One of the principal influences on Attila is surrealism. It can be helpful to read the book alongside, or after, writers like Éluard, Aragon, Breton, Savinio, and Roussel, because their ways of turning conventional sentences into contradictions or paradoxes, in the name of generating new meaning, illuminates large portions of Coll’s text.

Another modernist influence is painting, in particular cubism and abstraction. Coll is quoted as saying there hadn’t been a Mondrian in literature, so he aspired to fill the role. The way that model shows itself, for me, is in the unexpected geometrization of landscapes and people. “Slowly her incredulous mouth abandoned the square for the circle” is a typical example.

Perhaps the principal stylistic ambition in Attila is a persistent classicizing archaism, which shows as a predilection for awkward, formulaic locutions.

While the two laughed, the goat, as unperturbable as a fig tree, resumed its suppression of the shrub, not deviating one inch from its path.

The simile sounds like classic Greek or Latin pastoral poetry, as in Vergil’s Eclogues or Georgics, or Theocritus. The last clause is a Homeric simile, as in the Iliad, where people do not deviate from their paths, or are unbending, or press directly onward.

The heroic is a tricky voice to sustain. We’ve all been subjected to antiquated translations of the classics, where the translators’ attempts at noble or serious tone end up sounding awkward or pompous—and in popular culture, we’ve all been bemused by mid-century costume dramas (Victor Mature as Demetrius of Syria, as Hannibal, as Samson, and as Ulysses). Occasionally Coll seems to see the humor:

The emissary left the atrium with the same speedy diligence as a piece leaves a checkerboard.

But mostly Attila is as serious as Samson and Delilah. It is full of echoes of classical and other ancient texts. I suspect the battle scenes come from Lucan or Livy, and possibly Statius. Here’s a pastiche of Ecclesiastes:

From head to toe, the sun grows and shrinks, and from side to side the moon waxes and wanes… So that the sun always returns but the moon comes and goes.

Compare Ecclesiastes 1:5-6:

The sun rises and the sun sets, / and hurries back to where it rises. / The wind blows to the south / and turns to the north; / round and round it goes, / ever returning on its course.

I call Coll’s lines a pastiche because they don’t add anything to the original. He dilutes Ecclesiastes by elaborating (“head to toe,” “grows and shrinks”). The idea that the sun returns but the moon comes and goes, is nonsensical. The same unnecessary and distracting ornamentation is present in this passage:

This time Ipsibidimidiata’s tears spilled, not as discrete tears but in a curtain. She cried in silence, her expressive face writing in Quixote’s eyes that she did it out of anguish, reassurance, and lyricism.

“Writing in Quixote’s eyes” is part of Coll’s faux heroicism; the trope of a curtain of tears is in Homer and Callimachus, and I assume many others. The three qualifiers at the end are characteristic steps too far: anguish would be expected in context, and reassurance is plausible and can be imagined, but lyricism wouldn’t be used this way in a classical text, and a grammarian would say it’s a broken parallelism. Did Ipsibidimidiata feel “lyricism” the way she felt anguish? How exactly did she communicate “lyricism”? As it often happens, Coll falls between the stools of classical rhetoric and surrealism.

To explore these issues it’s necessary to do some close readings. The review in The Untranslated has the only close reading I’ve encountered:

The following sentence appears in a passage describing the steppe shrouded in the clouds of dust raised by the horses of the Huns: “El polvo no pudo esconderse en el lubricán ni en el polvo el lobo.” (The dust could not hide in the twilight nor could the wolf in the dust.) Even in such a short and seemingly simple sentence, there is evidence of serious work with the language. Aliocha Coll chooses the rare word lubricán for “twilight”, not only because it has the l and n sounds necessary for the alliterative pattern but also because there is “wolf” hiding in the word, which is derived from the Latin words lupus and canis, i. e. wolf and dog.

This sounds right to me; I can’t judge this judgment of Spanish. Whittemore’s rendering is “The dust could not hide in the gloaming nor the wolf in the dust.” But assuming the reading here is right, there’s a further issue. What does “The dust could not hide in the twilight nor could the wolf in the dust” mean? Saying that Attila is “just poetry” or is made available for “feeling” and not for sense avoids the hard work of reading it as poetry, or articulating the feelings. In this case the principal purpose seems to be to put alliteration to work in a classical chiasmus, in which case the words themselves are partly for sound and partly for their violent feeling. For me the alliteration and chiasmus are unrewarding, especially because they have no echoes in the lines before or after. Coll wrote one trope, one rhetorical device, one adage or philosopheme, at a time, twisting until he got a high enough quotient of archaism and surrealist surprise.

Pastiche isn’t necessarily bad, and there are passages in which these concretions of different sources, vocabularies, and styles create something new. Here’s part of a report about what scouts find on a battlefield:

Many more scouts who had tracked, seen around all Nature, not legitimized by its concretion but abstracted by its laws, combustible fuel of combustible man and cumbustor of comburent man, implicated in that desouling, in those deposited pieces of the living only chimerically recomposable, in that pyramid of members flat at the base and hyperbolic on the face, with the luminous vertex inside and illuminated from outside, luminous like the wing of a crow that comes gleaming, carbonically illuminated between an obelisk erected and infanticidal and an obelisk inverted and matricidal, one flooded and the other searing and chamberlains of that tight and trans-darkened epigeous of broken weapons and flesh.

It’s a memorable anacoluthon, in the tradition of postwar French surrealist poetry. There’s an admixture of fantasy as in Lovecraft (Mountains of Madness) or Verne, a hint of Poe (“luminous like the wing of a crow”), lots of weird geometry, the usual arcane vocabulary culled from disparate sources (“comburent,” from chemistry; “epigeous,” from botany), and faux philosophy (“not legitimized by its concretion but abstracted by its laws”). Most passages in Attila don’t manage such a mixture.

5. Substance of the book

Then comes the question of the book’s overall position in relation to its subject matter. Attila is a “meditation on history, civilization, and art” (in Whittemore’s words), but what exactly is the content of that meditation?

One of the central concerns of Attila is the contrast between pagan and Roman culture. This is itself a trope in literature. Flaubert was the first to note the interest of the moment between Roman religion and Christianity, where people could, in his view, think freely. Marguerite Yourcenar’s wonderful Memoirs of Hadrian develops Flaubert’s idea. Coll isn’t interested in Stoicism or Epicureanism, but he clearly inherits the idea of the turning point. There’s a set piece in which Quixote (one of the principal characters, quixotically named in order to dislocate him in time) sees a procession in Rome with “the cross looming overhead.” But unlike Ammianus’s Res Gestae, he doesn’t see Christianity as a cult or a threat. Instead Quixote’s response is tangled in surrealist paradox. He thinks,

surprising himself, “what a fundamental symbol of evil, since it does not cardinally associate the empire with pain? Yes, cardinally, possession and suffering?”

This would make more sense if the thought was, “Christianity does not recognize that the Roman Empire controls real suffering,” but the message is distorted by Quixote’s use of the word “evil,” which is not a Roman concept, and by his stress on cardinality, which doesn’t add logical content and may be nothing more than a riff on the slant rhyme with Cardinals.

There are hundreds of passages in Attila that follow the same strategy: begin with a familiar observation and substitute or add words until it becomes partly illegible and therefore putatively more meaningful. This may sound overly critical: but unexpected substitutions and juxtapositions have been cardinal surrealist strategies ever since Lautréamont.

In a crucial scene, Attila describes his intentions in regard to Rome.

We will not occupy this Empire.. What we will do… is to harvest the classical legacy… [and] the Christian promise… To instill in our culture and in Nature those ingredients… We will withdraw to the meadow with… a culture without doctrines… [and abandon] the sumptuous carrion of the Empire.

This isn’t an interesting program. Syntheses of Roman law and culture and Christian belief have been proposed by writers from Origen and Boethius onward, and Attila’s version has no content except for the illogical notion that the result would be “a culture without doctrines.” The same received ideas, schematically and paradoxically distorted, appear in a later section of the book about China. This lack of content is why I wonder whether a plot synopsis, such as the one provided by The Untranslated, is helpful in the end: plot would matter more if this were a philosophical novel, with ideas to contemplate regarding civilizations and cultures. Instead, most of the plot is scaffolding for the many inversions of sense and meaning that Coll seems to have taken to be the faithful response to the challenge of modernist writing in general and Finnegans Wake in particular.

It may be that Attila will “will keep generations of critics busy” as The Untranslated says, but I doubt it.

6. Literary criticism in 2025

I’ve written a lot about criticism in other media (mainly visual art), and I’m a long-time observer of the state of art criticism. I wondered, writing this piece, why it seemed necessary for me to try to say these things about the reception of Attila. Why take the trouble to argue against readers who like the book, for whatever reason? I have two thoughts along those lines.

To argue in the way I have tried to do, it is necessary to include history in the conversation. Even though I don’t have evidence of what books Coll read or praised, I can guess at them based on his allusions and choices of terms and proper names. Once he is placed in the history of literature, he becomes interesting in a completely different way that does not require mystification. It would be great to have more history here on Substack, as I’ve suggested a couple of times. (Here and a bit here.)

Second, “close reading” has a bad name that it doesn’t deserve. There’s plenty of close reading here on Substack, but it’s mainly among our “neo-romantics” and classicists. In literary criticism close reading is associated with a certain period in the development of literary studies in universities. (As in John Guillory’s book.) It came from an emerging sense of professionalization, and it has been partly left behind by practices like “surface reading” and “distant reading.” But what if “close reading” just means pondering choices of words? And therefore asking questions that can lead into any politics? A more forgiving notion of close reading might help prevent the confusions of critics who think books like Attila are either incomprehensibly brilliant or irredeemably awful.

1. I am following the spelling change here; Aliocha was a pen name adopted to evoke Alexei Fyodorovich Karamazov in The Brothers Karamazov.

2. Here I’m only looking at the English-language reception. I imagine, since Enrique Vila-Matas endorsed Serena’s Attila, that the Spanish-language reception might be different. (Why did an author as involved in metafiction and postmodernism endorse Serena’s book?) In English, the novel has been rapturously received by almost uniformly confused readers.