In 1921 the Scottish poet Edwin Muir and his wife Willa gave up their jobs in London and went to live on the Continent. The dollar was strong; they hoped to make ends meet by reviewing books for the American periodical The Freeman.

After a nine-month spell in Prague, the Muirs moved to Dresden and began to learn German. Willa did some schoolteaching while Edwin stayed at home reading the latest German-language writers. When hyperinflation struck Germany, they moved to Austria, then to Italy, then back to England. There they put their newly acquired German to use and became professional translators. For the next fifteen years, until the outbreak of war, they were, in Edwin’s words, “a sort of translation factory.” Together they translated over thirty books; Willa did a further half-dozen by herself. “Too much of our lives was wasted…in turning German into English,” wrote Edwin afterward, ruefully.

With one of their first projects, a translation of Lion Feuchtwanger’s Jud Süss, they struck it lucky: the book became a best seller, and their London publisher asked what other writers they could recommend. Edwin had been reading Franz Kafka’s posthumously published Das Schloss. “It is a purely metaphysical and mystical dramatic novel,…quite unique,” he wrote in a letter. His and Willa’s translation was published in 1930. Despite selling only 500 copies, it was followed by further Kafka translations: The Great Wall of China (1933), a collection of shorter pieces; The Trial (1937); America (1938); and, after the war years, In the Penal Colony (1948). These translations, despite their many defects, have dominated the English-language market since then. Now we have a new version of The Castle by Mark Harman of the University of Pennsylvania, standard-bearer for a retranslation of the entire Kafka corpus commissioned by Schocken Books from respected individual translators. (Breon Mitchell’s version of The Trial will be issued later this year.)

Edwin Muir saw his task as not only translating Kafka but also guiding English readers through these new and difficult texts. The Muir translations therefore came armed with forewords in which Edwin, relying heavily on Kafka’s friend and editor Max Brod, explained what Kafka was all about. His forewords proved highly influential. They proposed Kafka as “a religious genius…in an age of scepticism,” a writer of “religious allegory” preoccupied with the incommensurability of the human and the divine.

Inevitably the conception of Kafka as a religious writer influenced the choices the Muirs made as they translated his words. The English versions they produced conformed to the interpretation supplied in the forewords. So it is not surprising that Kafka’s first English-speaking readers accepted without demurral the Muirs’ version of him.

The interpretations embedded in the Muir translations—particularly of The Castle and The Trial—have long been a source of concern to Kafka scholars. In the United States in particular, the Muirs’ 1930 translation of The Castle, implicitly packaged with the Muirs’ 1930 reading and reprinted time and again, has seemed to hold an unfair monopoly (in Britain a new translation of The Trial appeared in 1977, and of The Castle in 1997).

There are other reasons as well why the Muir translation and the Muir monopoly have assumed a faintly scandalous air. The 1930 translation was based on the 1926 text given to the world by Brod, and this text was heavily edited. Brod was the one who made decisions about which parts of Kaf-ka’s fragmentary manuscript should go into the printed text and which not, and about where the chapter divisions should fall. Brod also augmented Kafka’s light, even minimal punctuation. Further errors were introduced by the printers. Thus the Muirs, through no fault of their own, were working from an original that was, by scholarly standards, unacceptable.

The main challenge facing a translator of Kafka, in the eyes of the Muirs, was to reconcile fidelity to Kafka’s word order—which was of course subject to the rules of German grammar—with the ideal of a natural-sounding, idiomatic English. To Edwin Muir, Kafka’s word order was “naked and infallible…. Only in that order could he have said what he had to say…. Our main problem was to write an English prose as natural in the English way as [Kafka’s] was in his own way.”

Naturalness is a concept not easily pinned down, but to Edwin Muir it appears to have included freshness of phrasing and lexical variousness. Thus, paradoxically, the Muirs are often more vivid than Kafka, whose German tends to be restrained, even neutral, and who is not afraid to repeat key words again and again.

Furthermore, although the Muirs’ mastery of German—particularly Willa’s—was astonishing, given the fact that they were more or less self-taught, and although Edwin in particular had read widely in contemporary German and Austrian writing, neither had a systematic grounding in German literature, so their ability to pick up literary references was rather haphazard. Finally, there were aspects of German or Austrian life, each with its own specialized vocabulary, that they knew only sketchily.


One of these—unfortunately for them as translators of Kafka—was the law and legal bureaucracy. I give one instance. In The Trial Josef K. tells the men arresting him that he wants to telephone “Staatsanwalt Hasterer.” The Muirs translate this as “advocate Hasterer.” Readers brought up in the Anglo-American legal system will assume that K. is asking for permission to call his lawyer. In fact K. is trying a bluff, threatening to call a friend in the prosecutor’s office.

In The Castle in particular, the Muir translation includes scores, perhaps hundreds, of errors of detail which, while they may not be important individually, have a cumulative effect, putting readers on an insecure footing, driving them back to check the original at every crux of interpretation.

A few instances may give an idea of the range of the Muirs’ inadequacies. Reading Kafka, Edwin Muir observes, is much like reading “a travel book which recounts minutely the customs, dresses and utensils of some newly discovered tribe.” But when it comes to the everyday material culture of Central Europe, the Muirs are uncertain guides. The Strohsack on which K. beds down in the village inn, for instance, is not a bag of straw (as they say) but a palliasse, a straw mattress.

The curious telephone system of the Castle, and the related telephone etiquette, also seem to defeat the Muirs. To them, telephonieren (to telephone) is the ringing sound a telephone makes. When the Castle officials disconnect the ringing mechanisms on their receivers, the Muirs say they are “leav[ing] their receivers off”; and when they reconnect them, the Muirs say they are “hang[ing] the receivers on.”

If the legal system of the ex-Austrian Empire is strange to them, the practices of Kafka’s fictional Castle officials are even stranger. Without expanding words into phrases or adding footnotes, it is hard to explain to the English-speaking reader exactly what is meant by the claim that the only road to Herr Klamm of the Castle leads through the Protokolle of his secretary, or quite what the Briefschaften are that Castle messengers ceaselessly carry back and forth. Nevertheless, the Muirs are too often content with indicating only roughly what is meant, for instance by replacing a German word with its English cognate (Protokolle by “protocols”) or with a term of vaguely similar denotation (Briefschaften by “commissions”) and hoping that enough of the meaning comes across. In this respect the Muirs’ standards are simply too rough to qualify them as interpreters of The Castle, which is partly, though not wholly, a fantasy of bureaucracy run wild.

Sometimes the Muirs’ guesses at Kafka’s meaning are mere stabs in the dark. Kafka writes, for instance, of officials who revel in their despotic power over petitioners, “against their own will [loving] the scent of wild game like that.” Though he was the least ideological of writers, Kafka had an acute feel for the obscene intimacies of power. Hinted at in his striking metaphor is a bestial, predatory appetite in the officials, sometimes submerged, sometimes baring itself. The Muirs, missing the point, write that the officials “in spite of themselves, are attracted by those outlaws.”

When Kafka is obscure enough to defeat any but an inspired reader (what is a clinging street, eine festhaltende Strasse?), the Muirs’ tactic is to take a guess at what Kafka might have intended, rather than—the last honorable recourse of the baffled translator—to fall back on word by word transposition. Their guesses are not always convincing—here “the obsession of the street.”

And sometimes the Muirs simply fail to see what is before them. “Her blank loveless gaze,” writes Kafka. “Her cold hard eye,” write the Muirs, missing the ambiguity of “loveless” (lieblos).

At a broader level, there are occasions when the Muirs, whether consciously or not, sacrifice fidelity to Kafka’s text to a vision—their own vision—of the whole. To the Muirs, taking a hint from Brod, Surveyor K. is a sympathetic pilgrim figure. So when K. claims at one point to have left a wife and child at home, yet at a later point wants to marry the waitress Frieda, the Muirs save their embarrassment by eliding the wife and child.

Finally, however “natural” the Muirs’ English may have been in its day, it is now dated—as we might expect, given that it is nearly seventy years old. Unless readers make conscious allowance for this, they will not know how to interpret moments when the Muirs’ language seems to take a dip into the past: “stoutly-built” rather than “powerful”; “guttersnipe” rather than “tramp”; “remiss in industry” rather than “lacking diligence.” Has Kafka himself used outdated forms, the inquisitive reader might ask, and are the Muirs signaling this in their word choice? The answer is, in each case, no. Although there are levels of formality in the language of The Castle, there is no historical dimension built into it—no systematic recourse to old usages or up-to-the-minute idioms.


Kafka began writing The Castle in early 1922, at a mountain resort where he had been sent for a rest cure (his tuberculosis had been diagnosed in 1917). The cure did little good, and he returned to Prague. In July he took early retirement from the insurance company for which he worked. His condition deteriorated further, and he abandoned The Castle. In all he had spent seven months on it, not completing even a first draft.

After Kafka’s death in 1924, Max Brod edited and published the three novel fragments his friend had left behind. Alarmed by what he considered to be crass misinterpretations of the first of the three to appear, The Trial, Brod accompanied The Castle with what the American scholar Stephen Dowden, in his survey of the reception of The Castle, pithily calls “a preemptive hermeneutic strike on behalf of his own views” in the form of an afterword.* This afterword deeply influenced the Muirs’ reading of the novel, and in due course their translation of it.

The 1926 edition was the one from which the Muirs worked. In subsequent editions (1935, 1946, 1951) Brod brought forth from his stack of manuscripts further episodes and fragments, as well as variants and deleted passages. These additions to the text have for some time been available in English, appended to the Muir translation in dependable versions by Eithne Wilkins and Ernst Kaiser, best known as translators of Robert Musil.

For his project of publishing the Kafka manuscripts posthumously, Brod marshaled a public declaration of support by respected intellectuals like Thomas Mann, Martin Buber, and Hermann Hesse, who thereby seemed to lend their weight to the readings of Kafka that accompanied the successive volumes. Thus the reading of The Castle set out in Brod’s 1926 afterword came to dominate the reception of Kafka in the German-speaking world well into the 1950s; in large part through the agency of Edwin Muir, a version of the same reading held sway in the English-speaking world.

In Brod’s reading, K., the new Faust, is driven no longer by a desire for ultimate knowledge but by a need for the most basic prerequisites of life—a secure home and job, acceptance into a community. To the later Kafka, says Brod, these simple goals had come to have a religious meaning. The minimal grace that K. seeks from the Castle is permission to settle down, to cease to be an outsider.

Knowing as he did Kafka’s sense of himself as a lifelong outsider, knowing of Kafka’s repeated attempts to get married, undermined each time by an incapacity to imagine himself in the role of husband and father, one can see why such a reading should have occurred to Brod. One can see too why it should have resonated with the public mood in postwar Germany and Austria, where the economy was stagnant, where neither Church nor State could give direction to life, and where widespread hopelessness and a sense that mankind had lost its way prevailed.

Brod’s optimistic reading of The Castle (influenced no doubt by his own Zionism) makes of Kafka—whom Brod revered, yet utterly failed to understand—a rather simple and conservative thinker who responds to the challenges of modern life with a call for a return to old verities. It is thus no idle undertaking to rescue Kafka from Brod’s version of him, or at least to complicate and deepen Brod’s version enough to situate his Kafka within the history of his times. If, in the English-speaking world, this means scraping off from the Muirs’ version of The Castle those accretions that belong only to Brod and Brod’s vision, or even translating The Castle anew, as Mark Harman has done, this is a task richly worthwhile.

Indeed, even Edwin Muir began to have doubts about a religious-allegorical reading of The Castle after he had met and spoken to Dora Dymant, Kafka’s last companion. Were it not for the fact that Kafka annihilates so thoroughly the boundaries between the clownish and the evil, one might call The Castle not a religious but a comic novel. The shadow cast by the Castle may indeed be dreadful, but K., the hero of the tale, caught in a labyrinth of bureaucratic hocus-pocus, has more pressing concerns than coping with dread, as he tries busily, sometimes frantically, to get his papers in order and keep the peace among the many women in his life.


The Brod who delivered The Castle to the world was of course no ordinary editor. His friendship with Kafka was close and deep, dating back to their student years together. There is no reason to doubt his claim to have been told Kafka’s plans for the conclusion of The Castle (namely that on his deathbed K. would receive a judgment from the Castle rejecting his petition to settle in the village, yet permitting him to live and work there).

So thoroughly has Brod been taken to task as the jealous impresario of the Kafka reputation and the founder of the dreary science of Kafkology (the term is Milan Kundera’s) that it is as well to remind ourselves that Brod saved Kafka’s manuscripts from destruction not once but twice: the first time by disobeying Kafka’s instructions that they be burned “unread and to the last page” after his death, the second time in 1939, when Prague was occupied by the Nazis and he fled to Tel Aviv bearing the papers with him. In 1956, when war threatened in the Middle East, Brod had the bulk of his collection conveyed to Switzerland. From there it traveled to the Bodleian Library in Oxford, where the eminent Germanist Malcolm Pasley reedited The Castle from scratch, producing in 1982 the two-volume critical edition on which Harman’s new translation is based.

Pasley’s description of the manuscript, included as an afterword to Harman’s translation, gives us a useful glimpse into the unique problems faced by the translator—or retranslator—of The Castle. Kafka seems to have planned to write the entire novel (which in its present unfinished state runs to some 100,000 words) in the form of a tightly knit, temporally consecutive draft which would require neither substantial additions and deletions nor restructuring. The novel was not written in chapters, though some chapter divisions emerged as the text grew. Nor, after the first third of the book, did Kafka carry out much paragraphing. His punctuation overall is extremely light, serving in the main to mark rhythm and cadence. Commas, rather than full stops, separate sentences.

Kafka’s language is generally clear, specific, and neutral; in the novels—The Castle in particular—it can seem monotonous if one is not caught up by the forward drive of the sentences, their urgent but rather abstract energy, for which Kleist’s prose provided Kafka with a model. Brod recalled that “anyone who had the privilege of hearing [Kafka] read his own prose to a small circle, with a rhythmic sweep, a dramatic fire, a spontaneity such as no actor ever achieves, got an immediate impression of the delight in creation and the passion that informed his work.”

In its spareness and apparent matter-of-factness, Kafka’s language has been claimed to be typical of the German of Prague, and particularly of the German of assimilated middle-class Jews, but the claim is questionable. It is more likely that it was influenced by the precision of good legal prose, the medium in which Kafka worked day by day. The manuscript of The Castle exhibits a number of Prague German usages. Though Kafka did not change these to standard German in the course of his revisions, he clearly did not intend them to stand in the text as indicating the use of dialect, and (correctly, to my mind) neither the Muirs nor Harman translate them into nonstandard English.

Nor, after the opening chapters of the book, does Kafka maintain any socially realistic distinction between levels of language: the people of the village as much as the Castle officials seem able to produce exegetical monologues on the most trivial of questions at the drop of a hat. Thus the translator, once he or she has found a variety of English that, to his or her mind, does justice to Kafka’s variety of German, does not often have to change gear. Indeed, the temptation to be resisted is to introduce a linguistic variousness that is absent in the original.

This necessary rigor raises certain problems. In the second half of The Castle, in particular, Kafka slips on occasion into some very tired-sounding prose. Here is an extract from the reported narration of young Hans Brunswick. I give it in Harman’s translation, which reproduces Kafka’s syntax and gives an indication of how light Kafka’s punctuation is.

Father…had actually wanted to go and see K. in order to punish him…only Mother had dissuaded him. But above all Mother herself generally didn’t want to speak to anyone and her question about K. was no exception to that rule, on the contrary, in mentioning him she could have said that she wished to see him, but she had not done so, and had thus made her intentions plain.

In producing English sentences as slack as Kafka’s own, Harman has in principle made the right decision (“I have tried to keep [it] as murky in English as it is in German,” he observes of a comparable passage). Nevertheless, it is only Kafka’s classic status that gives grounds for such a decision: translating a more run-of-the-mill writer, one would be eminently justified in lightly and silently fixing up the original.

There are several passages in The Castle where Kafka, groping to record moments of transcendental insight, visibly reaches the limits of expression. Passages such as these provide the harshest test of the translator, demanding not only an ability to follow the utmost nuances of phrasing but, beyond that, an intuitive sense of what the resistances are against which the language is pressing. A case in point is the passage, only a sentence long, describing the first lovemaking of K. and Frieda. It is a sentence to which Milan Kundera has already given microscopic attention, using it to test Kafka’s French translators, who by his standards fail badly. How does Harman fare?

“Hours passed there,” writes Harman,

hours breathing together with a single heartbeat, hours in which K. constantly felt he was lost or had wandered farther into foreign lands than any human being before him, so foreign that even the air hadn’t a single component of the air in his homeland and where one would inevitably suffocate from the foreignness but where the meaningless enticements were such that one had no alternative but to go on and get even more lost.

In the metaphor at the heart of this passage, the poles are die Fremde and die Heimat, foreign country and home country. To capture in English all the associations of these rich words is a hopeless task: fremd has all the associations of English strange, yet Kafka’s phrase in der Fremde is also a perfectly normal expression meaning “abroad.” In his syntax, Kafka seems intent on making the strange or foreign experience into which K. is drawn with Frieda seem ever stranger and more threatening as the sentence proceeds: “K. constantly had the feeling he had got lost or was (had gone) further in(to) die Fremde than anyone before him, a Fremde in which even the air had no component of the air of home [Heimatluft, a neologism], in which (in the air? in the Fremde?) one must [inevitably] suffocate of Fremdheit, and (yet) in (in the midst of?) whose crazy enticements one could do nothing but go further, get further lost.”

In his rendering, Harman copes well with Kafka’s strategy of overlaying K. as a wanderer coming from the comforts of Heimat into the unsettlingly fremd domain of the Castle with K. wandering into disturbing spiritual territory in the sudden act of sex with Frieda; but rather less well in conveying the impression of a man straining to get a grasp on fremd experience in a sentence faltering in its sense of linear direction and on the point of losing its way.

As for the Muirs, they miss the boat entirely, partly because they overtranslate (“K. constantly had the feeling” becomes “K. was haunted by the feeling”); partly because they evade difficulties—Kafka’s cryptic unsinnigen Verlockungen, which Harman translates rather woodenly as “meaningless enticements,” becomes in their version simply “enchantments”; but mainly because the English sentence they construct is too musically balanced.

In general, in all the large-scale decisions he has had to make—whether or not to remain faithful to the original in its moments of weakness, whether or not to follow Kafka’s punctuation and Pasley’s chapter divisions, what variety of English to aim for—Harman has done well. Though his renderings of the most baffling moments in the text rarely strike one as inspired, he can certainly claim to have produced a version of the novel that is semantically accurate to an admirable degree, faithful to Kafka’s nuances, responsive to the tempo of his sentences and to the larger music of his paragraph construction. For the general reader as for the student, it will be the translation of preference for some time to come.

If there are lapses, such lapses do not point to any larger deficiency. If only to give an idea of their scale, I list a few.

(1) If the Muirs’ English has become a little dated Harman’s is, in places, unnecessarily colloquial: “stomped”; “what he had gone and done”; “all that great.” It also sometimes descends into unjustifiable clumsiness. The Muirs’ “Painful, even unbearable” is certainly better than Harman’s “unbearably excruciating.” Word choice is occasionally slightly off target: “the village environs” is better than “the village surroundings.”

(2) There are moments when Harman falls into mindless late-twentieth-century jargon: “Perhaps K. was being mistakenly positive now, just as he had been mistakenly negative with the peasants.” Better might be “Perhaps K. erred here on the side of good nature, where with the peasants he had erred on the side of bad nature.”

(3) The word Heimat (home, homeland) creates problems that Harman does not always solve. Though the Castle and village have a telephone system, it is not used for communication with the outside world, nor do the natives seem to venture abroad. K., in contrast, speaks of himself as a wanderer. Nevertheless, there is no suggestion that he is a foreigner in the usual sense of the word. In the world of The Castle there seem to be no national boundaries. So when, in Chapter Two, K.’s thoughts return to his Heimat, Harman’s translation “homeland” is too specific, particularly when we go on to read that his Heimat has a square and a church. “Home” or even “home town” would be safer.

(4) Harman is not immune from the temptations of overtranslating that so mar the Muir version. To translate Ihre schmutzige Familienwirtschaft as “your dirty family shambles” introduces ideas of blood and death not present in the original. “Your sordid family setup” would be better. Similarly, “freaks” (Wilkins/Kaiser) translates Ausgeburten rather more soberly than “evil spawn.”

(5) Sometimes fidelity to Kafka’s word order becomes mechanical and yields what sounds like a parody of German: “a large, heavy, and now open gate.” The German is sometimes followed even when there is a close English idiom: “a crying injustice” is better than “a screaming injustice.”

(6) It is not always clear that Harman recognizes Prague usages. For instance, the huge, three-doored item of furniture in the landlady’s office in Chapter 25, ein Kasten, is probably a wardrobe, as Wilkins/Kaiser have it, rather than a cabinet, as Harman has it.

(7) There are odd moments of inattention: “I entrusted a message to you not so you would forget it or garble it on your cobbler’s bench” should surely read “not so [that] you would forget it on your cobbler’s bench and garble it.”

As these instances indicate, the issues over which one might want to criticize Harman are limited in range and small in scale. More significant ones emerge at moments when Kafka’s own prose is at its worst, or—to say the same thing—when it is hardest to ignore that what Kafka has left us is a draft, not a finished work. “It would after all have been better to have sent the assistants here, for even they would have been capable of conducting themselves as he had done,” writes Harman, following faithfully in the footsteps of a sentence that reads as though Kafka wrote it in his sleep. The trouble here is not only that the writing is slovenly but that Kafka’s meaning is unclear. The Muirs clarify the meaning, but only at the cost of importing a notion (foolishness) of which there is no trace in the original: “Really he would have done better to have sent his assistants here, they couldn’t have behaved more foolishly than he had done.” The problem is finally intractable.

Besides Malcolm Pasley’s afterword on the manuscript sources of The Castle, the new volume comes armed with a useful note by Arthur H. Samuelson on the publication history of Kafka’s works, and an eleven-page Translator’s Preface in which Mark Harman discusses Kafka’s language and some of the problems it presents, illustrating and justifying his own approach, and quoting some telling examples of the inadequacies of the Muir version.

It is of course to be expected that Harman should be hostile to his rivals (upon whom he delivered an even more trenchant attack in an article published in New Literary History in 1996). Everything he says about the Muirs is true: their religious interpretation of the book did indeed bias their rendering of it; they did indeed introduce transitional links to make Kafka’s daring leaps more easily negotiable; they did indeed, in general, “tone down the modernity” of The Castle.

But to say of Edwin Muir that “[his] literary sensibility…was molded by nineteenth-century figures such as Thackeray and Dickens” goes too far. Nor were the Muirs merely, as Harman describes them, “a gifted Scottish couple.” The Muirs, together with their publisher Martin Secker, introduced a difficult, even arcane German modernist to the English-speaking world years earlier than one might reasonably have expected this to happen. Edwin Muir himself was a considerable poet, not quite of the rank of Yeats or Auden, of whom he sometimes reminds one, but a modern master in his own right, certainly no hankerer after a lost nineteenth-century world. If Dickens left an imprint on the Muirs, he left an imprint on Kafka too; as for Thackeray, one searches in vain for any residue of him in the Muir translation.

Harman calls the Muirs’ translations of Kafka “elegant” and commends them for their “smoothly readability.” The compliment is backhanded, and intended to be so. His own English, Harman says, is “stranger and denser” than theirs. Harman would do well to recognize that, if a striving toward elegance—fluency is the term I would prefer—marks the Muirs’ translation as of its time, then in its very striving toward strangeness and denseness his own work—welcome though it is today—may, as history moves on and tastes change, be pointing toward obsolescence too.

This Issue

May 14, 1998