Anderson on Powell and Proust.

Perry Anderson writes long, long essays for the London Review of Books; they are always interesting, but I confess when their topics are not central to my life, I tend to skim long chunks of them. However, I read all of his very, very, very long piece in the latest issue; ostensibly a review of Hilary Spurling’s new biography of Anthony Powell, it is in fact a detailed comparison of Powell and Proust, frequently to the detriment of the latter. As one of the (probably lamentably few) readers who have made their way through both A la recherche du temps perdu and A Dance to the Music of Time (“1,240,000 words in Proust, 1,130,000 in Powell”), I gobbled it all up, and those who have read one or the other massive set of novels (or, if you prefer, massive novels divided into chunks for publication) may well want to read the relevant portions of the review. I confess I was originally going to post just the following bit, for the sake of the pun (pompe means both ‘pomp’ and ‘pump’):

[Proust’s] father was a friend of Félix Faure, president of the Third Republic in the last years of the 19th century, who expired in one of its most famous scandals, in the course – his pompe funèbre, as it was widely dubbed – of fellation by his mistress.

But as I read on I found he had such interesting things to say that I thought I’d reproduce a few paragraphs here:

All literary magnitudes are finite. In the case of great writers, it is understandable that their vices should attract so much less attention, not infrequently to vanishing point, than their virtues; but not defensible. For all its gifts, Proust’s work has conspicuous shortcomings, regularly ignored in the cult of it. Time – the grandiloquent leitmotif of the novel – is handled erratically at the episodic level of the narrative, in ways that cannot be explained by the vagaries of the narrator’s memory. Proust, who rarely dated his letters, blurred temporality by avoidance of any specifications of age, and frequent resort to the iterative form of the past imperfect (‘Aunt Léonie would … ’), but lapses of control are obvious enough. Odette’s visit as a courtesan to Uncle Adolphe when the narrator is plainly older than when she figures for him as Swann’s wife; references to historical events impossible at the time of fictional episodes – Edward VII’s state visit to Paris (1903), the Russo-Japanese War (1904-5), the death of the Swedish king Oscar II (1907), all at an early stage of talk about the Dreyfus Affair (1898) etc: the external and internal chronologies of the novel do not fit. More significant is the stasis of the society depicted in the arc of the narrative as a whole. Between the 1870s, when the story of Swann hypothetically begins, and the aftermath of the First World War, when the novel ends, virtually the only area of change is technological. Electrical lights replace gas lamps, motor cars replace coaches, telephones and aeroplanes appear: otherwise, all remains essentially as it was. Even the war alters only fashions in dress, and the guests at the same receptions. History, for Proust, was romantic imaginings of a medieval or Renaissance past – ‘the distant gules of kingly France’. It had little or no meaning for him in the present. The passage of time is purely existential, the slide towards decrepitude and death. The mystique of involuntary memory he erected around it was essentially defensive. In his remarkable Proust among the Stars, the one study of A la recherche equal to its splendours, Malcolm Bowie observed with justice that while Time in this mode is ‘a “big” controlling theme’, calling forth ‘an impressive philosophical diction’, it ‘levitates too obligingly above the restless detail of Proust’s writing’ in ways that can be ‘intimidating and coercive’, the strengths in his handling of it lying elsewhere. ‘It is down among Proust’s intricate propositional structures with their outrageous embeddings, suspensions and redundancies that his boldest pieces of temporal architecture are to be found.’

Like capitalised time, characters are another overdone feature in standard encomia of the novel. For the most part, these have plenty of life, but lack any depth. Where they are most vivid, Proust’s portraits belong to the art of caricature. They depend on repetition and exaggeration, resembling the garish dummies that are a feature of Dickens more than anything in prior French fiction. There is a range here: at the lower end, the dullards of the Verdurin salon, Cottard, Brichot, Saniette and the rest declaiming their tag-lines, the poseur Bloch with his ludicrous affectations, the hypocritical snob Legrandin: grotesques pure and simple. A notch up, the sententious diplomat Norpois and his one-time mistress Madame de Villeparisis. Beyond these, the bluff Duc de Guermantes and the ineffable Verdurin couple themselves. At the heights, his most famous creation, where pathos transfigures a comic monster into something closer to, if never altogether becoming, a credible human being, the Baron de Charlus. This gallery does not, of course, exhaust the population of the novel. But characters who escape such treatment pay a penalty: they remain curiously blank. Famously so in the vacant mystery of Albertine, but in large measure true also of Swann and Odette, figures memorable more for their repertoires and accessories, and the stylised roles they play, than for their indistinct persons; of Gilberte, given little more than the red hair and freckles of her childhood; or even, after his initial scenes, monocle dancing in front of him like a butterfly, Saint-Loup. These are not caricatured, but they too lack complexity. […]

Flooding the narrative, total immersion in the self yielded prodigious riches of interior perception and reverie, delivered with continual energy and subtlety. At such a pitch of creative intensity, however, the egoism of Proust’s imagination permitted no more than fitful and superficial attention to others. That might have been just a limitation, of the kind to be found in any writer. Proust, however, erected it into a doctrine: the impossibility – not just in his, but in every case – of anyone ever knowing anything of others. Hence the ease with which his characters become their opposites. He once wrote that ‘Something repeated ten times is the opposite of art,’ and in the novel has his narrator disdainfully declare: ‘A work in which there are theories is like an object which still has its price-tag on it.’ Yet, of course, A la recherche is laden with theories, none more tirelessly repeated than a dogmatics of nescience, the natural epistemology of an artless narcissism. Persons are but ‘a shadow which we can never succeed in penetrating, of which there can be no such thing as direct knowledge’; ‘man is a creature that cannot emerge from himself, that knows his fellows only in himself, and if he says the contrary, lies’; ‘it is the tragedy of other people that they are to us merely showcases for the perishable collections of our mind’: ad libitum.​ […]

[On Powell:] Such precision of long-range balletics is very rare, perhaps unique. It unfolds within a structure in which each of the 12 volumes has its own self-contained plot, built of five or so episodes separated by intervals, within the overall narrative. Openings are typically meditative, passages of controlled, often ironic reverie; endings alternating between deadpan or wry deflation of previous intensities of emotion or action, and bleak registration or stoical foreboding of events and the passage of the time. In between, the dramas dominating the narrative are formed by social juxtapositions of one kind or another: visits, receptions, parties, meals, weddings, drinks, soirées, exhibitions – a much wider variety than in Proust, where dinners and receptions occupy a third of the novel, but with the same set pieces in large gatherings. Parties in A la recherche roll on for well over a hundred pages apiece. More concise, Powell’s certainly owe something to Proust’s example, but are closer to the art of Dostoevsky, whose Possessed he regarded as perhaps the greatest of all novels. […]

Out of this frequentation with forms of writing prior to Augustan normalisation, and a much wider range of wartime and postwar reading, not just in English literature, Powell developed a style as unlike his prewar fiction as the characters of The Acceptance World are from the matchstick figures of Afternoon Men: combining formal elegance with colloquial energy, brilliant descriptive passages with abrupt analytic reflections, elastic clauses and sparing connectives, a prose uniquely at once alembicated and taut. Powell’s sentences are nowhere near the length or complication of Proust’s, which average 35 words apiece, and his imagery, vivid enough in its own right, does not compare with the poetic and scientific range or density of Proust’s. His art is more laconic. Counterintuitively, English has a more Latinate syntax than French, a single verb able to control a sequence of clauses where French requires cumbersome repetitions; it also allows participial constructions barred or reproved in French. Powell made extensive use of both possibilities, especially the latter, making an English equivalent of the Latin ablative absolute one of the trademarks of his style, with sovereign indifference to schoolroom objections to the pendant participle. English also, of course, has twice the vocabulary of French, because of its double Latin and Teutonic roots. Powell made use of that too: contrary to expectation, words that would normally be regarded as recondite, mostly of Latin origin, are more frequent in A Dance than in A la recherche.

For my own complaints about Proust, see my decade-old summing-up post. (Even though my wife and I spent many months reading Powell, I don’t seem to have posted about the books themselves, as opposed to linguistic tidbits that arose, like erk.)


  1. In scale and design, the architecture of A Dance to the Music of Time is unique in Western literature.

    Not. Artamène ou le Grand Cyrus (Madeleine de Scudery alias George de Scudery, the 1650s) is considerably longer than either Powell or Proust: 13,000 pages, ten volumes, almost 2 million words. Similarly, Het Bureau (J. J. Voskuil, 1996-2000) is 5,000 pages, seven volumes, 1.6 million words. Then there is the Croatian work Gordana (Marija Jurić Zagorka, 2007), with 8,800 pages, twelve volumes, 1.4 million words. All of them have similar scope of internal time. Only after these three (and several non-Western works) come Proust and Powell.

  2. Trond Engen says

    Karl Ove Knausgård’s Min kamp (My Struggle) from 2009-2011 ticks in at 3622 pages according to its Norwegian publisher and spans the first 40 years of the life of the narrator (who is very close to the author). I can’t find a word count, but it would be around 1 million words, on par with both Proust and Powell.

    (I haven’t read it.)

  3. I haven’t either — I suppose I may someday, but at the moment my feeling is “life is too short” (ironically, considering the work in question).

  4. David Eddyshaw says

    The Verdurins’ secret plot to aid Saniette is one of my go-to examples for Proust’s greatness. People you’ve get pegged as vile are much more complex than you thought.

    Anderson’s notion that this and similar things in Proust reflect poor characterisation skills reminds me of the low opinion of Tolstoy held by a literate and sophisticated relative of mine, on the grounds that his characters are “inconsistent.” Yup. Like real people.

    Powell and Proust do share a distancing of both themselves as authors and of their narrators. I had a sort of micro-revelation about Proust when I was thinking one day “there’s a sort of airlessness about this … oh, asthma!” Like the homosexuality, it’s part of what the novels are about; it’s both a literal theme, and part of the metaphorical apparatus Proust is deploying to describe his world.

    Jenkins’ extreme reticence about his own affairs throughout always stuck me as weird. The fact that Anderson needs to make a point of minimising it is quite telling in itself.

    I do agree with Anderson that Powell is underrated, and for reasons pretty much extraneous to his actual work. I think he overeggs the pudding with his Proust, though.

  5. Wikipedia:

    Faure died suddenly from apoplexy in the Élysée Palace on 16 February 1899, while engaged in sexual activities in his office with 30-year-old Marguerite Steinheil. The priest coming for a last benediction asking “A-t-il toujours sa connaissance?” was reportedly answered “Non, elle est sortie par l’escalier de service” (a wordplay in French: “A-t-il toujours sa connaissance?” means “Is he still aware?”, but can be understood “Is his acquaintance still there?”, thus the answer “She left via the service staircase”).

    It has been widely reported that Felix Faure had his fatal seizure while Steinheil was fellating him, but the exact nature of their sexual intercourse is unknown and such reports may have stemmed from various jeux de mots (puns) made up afterward by his political opponents. One such pun was to nickname Mme Steinheil “la pompe funèbre” (wordplay in French: “pompes funèbres” means “death care business” and “pompe funèbre” could be translated, literally, as “funeral pump”). George Clemenceau’s epitaph of Faure, in the same trend, was “Il voulait être César, il ne fut que Pompée” (another wordplay in French; could mean both “he wished to be Caesar, but ended up as Pompey”, or “he wished to be Caesar and ended up being blown”: the verb “pomper” in French is also slang for performing oral sex on a man); Clemenceau, who was also editor of the newspaper L’Aurore, wrote that “upon entering the void, he [Faure] must have felt at home”.[4]

  6. David Eddyshaw says

    There are a number of actual linguistic assertions in the Anderson piece that struck me as a bit iffy, notably his assertions that the style of English is “paradoxically” more Latinate than that of French, and that French syntax is more verbose (what does that even mean?) than that of English. (His comment that French translations of English works have more pages doesn’t prove anything in itself about syntax, even if it’s true. How about words being longer on average, or there being more redundancy in the spelling system?) Or the stuff about English vocabulary being “twice” the size of French because of the Germanic and Romance components. These seem like the rumblings of someone who is unaware that linguistic questions can actually be, like, you know, investigated.

  7. mollymooly: Thanks for that educational and hilarious comment!

    David Eddyshaw: Yes, I’m afraid I’ve pretty much become numb and resigned to that sort of sheer linguistic ignorance among the punditry.

  8. The estimated length of Min kamp is 1 million words, so somewhat less than either Powell or Proust. The source probably isn’t all that reliable, though. However, it is a question (more so than with P or P) whether it is a novel or a memoir.

    I was struck by the line in WP that it was published in German translation under the name Mein Kampf at the insistence of the publisher, to which the author acquiesced. On the one hand, what else? On the other ….

  9. Yeah, that’s a tough one. A koan, if you will. “What is the German title of a book that would be titled Mein Kampf in German?”

  10. I was struck by the line in WP that it was published in German translation under the name Mein Kampf at the insistence of the publisher

    but they didn’t use Mein Kampf at all: the volumes were given rather portentous single-word titles (Sterben, Lieben, Spielen, Leben, Träumen, Kämpfen – Dying, Loving, Playing, Living, Dreaming, Fighting) and at least the paperbacks now refer to the series as “Das autobiographische Projekt”
    I guess they probably didn’t want to imitate the English titles as e.g. the French and Italian publishers did

    by the way, I enjoyed vols. 1 and 2 and forced myself to read 3 and 4, then stopped

  11. Trond Engen says

    The title is quite deliberate, so what else indeed. I think the story must be that the publisher chose not to use an overarching title in German — or maybe the Norwegian title. Mein Kampf is not used on the German publisher’s website. Neither does die Deutsche Nationalbibliothek know it (at least not by that author…), but “Karl Ove Knausgård Min kamp” can be found.

    I can’t speak for the other estimate of 1 million words. Mine is simply 3622 pages of 275-300 words each, so not very reliable.

  12. by the way, I enjoyed vols. 1 and 2 and forced myself to read 3 and 4, then stopped

    ditto for me, except that I didn’t get as far as vol. 4

    I keep thinking I may get back to it but the thought recedes as time goes by

  13. You Knausgård defectors have significantly decreased the likelihood of my even starting the struggle. Takk!

  14. You Knausgård defectors have significantly decreased the likelihood of my even starting the struggle.

    I still have ambivalent feelings about the whole thing, because the first two books are actually structured quite intricately, and then the flood-gates open, so to speak. It doesn’t help that apparently he finished the four others quickly – I’d recommend the first volume as a stand-alone book, quite honestly.

    (wikipedia: “Knausgård had finished two volumes when the first book was released. He had been planning to finish the six volumes within the year, preferring to work under harsh deadlines to combat his writer’s block.”)

  15. The Knausgård title somehow makes me think of Spike Milligan’s book about his experiences in the British Army in WWII–“Adolf Hitler: My Part in His Downfall”. (Seven volumes in total, but I don’t know the word count.)

  16. David Marjanović says

    What is “the iterative form of the past imperfect”???

  17. David Eddyshaw says

    It probably means he’s noticed that the French imperfect doesn’t neatly match up with any English past tenses. If he means anything in particular, it’s probably that the thinks P uses (say) chantait a lot more in the sense of “sang” rather than “was singing”; I would think this would be true of practically anybody not given in his letters to recounting little anecdotes of his past life all the time. To be fair, though, it would (if true) be germane to his assertion that Proust is actually pretty vague about time. Why he would regard that as a sign of Proust’s literary inferiority is probably best left unprobed. Some of us just do get annoyed by studied vagueness, I suppose. I hope Anderson never has a go at Murasaki.

  18. marie-lucie says

    JC: Artamène ou le Grand Cyrus (Madeleine de Scudery alias George de Scudery, the 1650s)

    The author of the magnum opus was Madeleine de Scudéry, and she had an older brother Georges de Scudéry, a largely forgotten dramatist. Their last name had three syllables, thus Scu-dé-ry, not two as Scude-ry would have.

    mollymooly: Pompes funèbres (as in “pomp and circumstance”) refers to the funeral business in general, especially the ceremonial aspect. “Death care” (if there is such a set phrase) is only part of it.

    David M: What is “the iterative form of the past imperfect”???

    This phrase sounds like a mashup by someone who perhaps dimly remembers technical words learned in school but never understood them properly, especially the difference between form and function. There is no “past imperfect”, the imperfect itself is a past tense. It can have an iterative (repetitive) meaning but there is no special iterative form. In English the iterative meaning can be expressed by used to or would or just the addition of an adverb such as often.

  19. m-l: George for Georges was a typo on my part, Scudery for Scudéry resulted from using an anglophone source, but Madeleine really did publish the books under her brother’s name, so “alias” was correct.

  20. marie-lucie says

    JC, thanks for the explanations. I did not know that Madeleine had published as Georges, as she is still so well-known while her brother is not (I did not know about him before reading your comment). Is “alias” the right term here, since she did not make up a name but attributed her work to him?

  21. I think so, yes. Alias as a noun is just a false name, whether it’s someone else’s name or not. On one of those blogs where “real names” were insisted on, I privately suggested to two protesting participants that they should use the names “Duchesse de Guermantes” and “Dulcinea del Toboso” respectively: the host made no complaint.

  22. marie-lucie says

    Great idea!

  23. Spike Milligan’s book about his experiences in the British Army in WWII–“Adolf Hitler: My Part in His Downfall”.

    That was the first volume: subsequent volumes included
    Rommel: “Gunner Who?” (Also published, confusingly, as “Rommel?” “Gunner Who?”)
    Monty: His Part in My Victory
    Mussolini: His Part in My Downfall

  24. And I second the applause for molly’s educational comment on dirty French puns.

  25. January First-of-May says

    is considerably longer than either Powell or Proust

    The longest single work of literature that I have read so far – if perhaps only for a rather broad definition of “literature” – was probably Ashes of the Past, a surprisingly detailed Pokémon fanfic. It is currently about 1.7 million words long (with more expected to be written), and chronicles the adventures of the titular Ash (and some other people) over what appears to be around four years (I hadn’t kept up with the specific timeline).

    The only slightly shorter Taylor Varga (around 1.5 million words? not sure), which happens to be one of my favorite fanfics (in any fandom), somewhat infamously only covers a time period of two and a half months… well, as infamously as it gets, anyway, considering that it’s a crossover of an obscure web novel (itself well over 1.5 million words long, though I haven’t read it) and an even more obscure anime (though almost none of the anime’s plot elements actually come through, if mostly on the account of it being so short as to barely have any).

  26. I enjoyed reading this, having read a fair bit of Proust but no Powell. I did think the Powell quotes made his style seem a bit stodgy though. Is that fair?

  27. Yes, that’s fair. He’s a wonderful novelist in terms of character and plot, but not really a stylist; there’s nothing wrong with his sentences, but they’re serviceable rather than gorgeous.

  28. And I found a perfect follow-up to that comment in Charles McGrath’s review essay on the same biography in last week’s New Yorker:

    Powell’s other great invention is the novel’s prose style. The writing in his prewar novels is crisp, ironic, and bright almost to the point of brittleness. But the sentences in “Music of Time” are often long and labyrinthine, heavily qualified and with dangling modifiers all over the place. Here’s an extreme example, from a scene in which a woman at a dance dumps sugar on Widmerpool:

    Barbara now tipped the castor so that it was poised vertically over Widmerpool’s head, holding it there like the sword of Damocles above the tyrant. However, unlike the merely minatory quiescence of that normally inactive weapon, the massive silver apex of the castor dropped from its base, as if severed by the slash of some invisible machinery, and crashed heavily to the floor: the sugar pouring out on to Widmerpool’s head in a dense and overwhelming cascade. . . . Widmerpool’s rather sparse hair had been liberally greased with a dressing—the sweetish smell of which I remembered as somewhat disagreeable when applied in France—this lubricant retaining the grains of sugar, which, as they adhered thickly to his skull, gave him the impression of having turned white with shock at a single stroke; which, judging by what could be seen of his expression, he might very well in reality have done underneath the glittering incrustations that enveloped his head and shoulders.

    Neither Spurling nor anyone else has really accounted for the stylistic change. It may owe something to Powell’s lengthy immersion in seventeenth-century prose during the war years, when, unable to write fiction, he worked instead on a book about the biographer and antiquarian John Aubrey. And it may even derive a little from the classical authors Powell studied at school. The sentences sometimes read like Latin: you have to untangle all those clauses to figure out what goes with what. Writing like this, dense and layered, is the opposite of traditional English terseness and understatement. It barges ahead as if heedless of all the rules. But in its density and its slowness it adds a kind of sombre, almost Proustian counterpoint to what might otherwise be just a fleeting procession of events.

    The whole essay is well worth reading.

  29. Jerry Friedman says

    First time posting here—and a little late, but I’d like to mention that fantasy and science fiction series can reach word counts a good bit higher than those cited above. According to Longest Fantasy Book Series on reddit, the three longest in English are over 5 million words. The Wheel of Time by Robert Jordan (finished by Brandon Sanderson) is only 4.4 million words long, but I think it might qualify as a single novel. I’m not sure, though, since I barely made it to 2 million.

    This page from 2009 also includes science fiction and a few non-speculative series. There are a good number over a million words. I’ll just mention Gene Wolfe’s “Solar Cycle” at 1.37 million, whose first four volumes are by far my favorite of the megaseries I’ve read and incidentally the most influenced by Proust, I’m told.

  30. Thanks for that, and you’re never too late around here!

  31. @Jerry Friedman: I feel I ought to have liked The Book of the New Sun more than I actually did. It was engaging, and there were plenty of puzzles—many of which I figured out in the course of reading, but by no means all. However, I cannot imagine ever having a particularly desire to read it again.

  32. Jerry Friedman says

    @Brett: It’s one (or four) of the books I’ve reread the most, including this year, but that just goes to show.


  1. […] Hat quotes from a beautiful Perry Anderson essay at the LRB about […]

Speak Your Mind