"Even Homais Nods": Nabokov's Fallibility, or, How to Revise Lolita
In Nabokov’s published screenplay, Act 1 opens: “The words LAST DAY OF SCHOOL are gradually scrawled across the blackboard.” 28 Dialogue confirms it as the last day of school for Dolly and her classmates, and consistent time cues move the action forward by degrees to the next day, the day of Humbert’s arrival in Ramsdale and his discovering that the McCoo house—in which he had hoped to enjoy Ginny McCoo’s proximity—has just burned down. He finds himself steered to the Haze house and is about to reject it when he sees Lolita. While Charlotte happily pays for the taxi and installs his belongings in the house, Humbert chats up Lolita. He eagerly agrees to help her with her homework, but then she shrugs, “Well, there’s not much today. Gee, school will be over in three weeks” (43). Shortly, Charlotte returns and calls Lolita to the phone: “It’s Kenny. I suspect he wants to escort you to the big dance next month” (45). Two pages later, Nabokov headlines a jump in time: “THREE WEEKS LATER, THE DAY OF THE SCHOOL DANCE” (47). This appears not to be an intertitle, simply an objective indicator for the film’s potential director, whether in studio or study. The dance ensues, with a cameo appearance by Clare Quilty.
Two clearly marked sequences, then, reinforce themselves and remain stubbornly incompatible with each other. In one, Humbert arrives at the Haze home on the last day of school. In the other, he has been at the Haze home for three weeks when Lolita’s school year ends. In the time problem in the novel, there are almost three hundred pages between the incompatible dates, which involve a single indicator in each case. Here in the screenplay one elaborate series of time markers leads almost immediately into another quite incompatible with it.
Surely, some will chorus, a writer as attentive to detail and as wily as Nabokov could never have left such a glaring inconsistency without meaning it. (As far as I know, the discrepancy has never before been remarked on in print, more than twenty years after the screenplay was published. Inconsistencies tend to become “glaring” only when someone points them out.) Is one of the two time sequences unreal, invented perhaps by Humbert? If so, which one? Or is the whole sequence proof that Humbert is Quilty, or that Quilty is only Humbert’s double, since both arrive on Lolita’s (different) last day of school?
It is easy, all too easy, to invent fancy interpretations of this kind. Twentieth-century criticism has become expert, if that is the word, in strategies for retrieving a “higher” consistency from seeming inconsistency—although this often resembles a craft skill, an easily-acquired habit, rather than real inquiry after explanation. Readers inclined, in this so-called “postmodern” era, to suppose a story will slyly undermine itself overlook other problems and possibilities. In the Lolita screenplay the inconsistent time sequences cannot be easily explained as Humbert’s invention, since the time indicators are objective, supplied in one case by Lolita’s classmates before Humbert arrives in Ramsdale (although of course Humbert might have invented this scene too, to say nothing of Lolita, and himself—this road can quickly lead to bog and fog), and in the other case by Nabokov himself in a “stage” direction.
But the screenplay’s incompatible time-schemes can easily be explained as a mere mistake. Nabokov composed a long first attempt at a Lolita screenplay in the spring and early summer of 1960. 29 It had a prologue, Humbert’s killing of Quilty, and three acts, the first of which starts with Lolita on her last day at school, the day Humbert arrives. Nabokov here introduces Quilty to Ramsdale and to Lolita by way of Quilty’s uncle, the dentist. But he then had the idea of reinstating a scene he had envisaged when he first composed the novel, of McCoo engaging Humbert in an irrelevant and comically spooky guided tour of the house where he would have had him as a lodger had lightning not just burnt it down. 30 Nabokov therefore wrote an alternative version of Act I that begins with Humbert’s arrival in Ramsdale and the scene at the McCoos’. This version then moves straight to Lawn Street and Humbert’s conversation with sun-bathing Lolita, who runs off to talk on the phone to Kenny, the boy who will take her in three weeks’ time to the end-of-school dance—which in this alternative version Nabokov uses as the means of introducing Quilty in Ramsdale.
Nabokov appears to have allowed Kubrick to decide between the two versions, for he offered both, one paginated normally, the other as “alternative 1,” “alternative 2,” and so on, together with the rest of the typescript. When Kubrick protested that he needed to cut drastically, Nabokov offered a much shorter and more filmable version, in which Act I drew entirely on the alternative version. 31 In late 1970, anxious to publish his screenplay before both Alan Jay Lerner’s Lolita musical and the deadline for his multi-book McGraw-Hill contract, Nabokov looked back over his screenplay, which now included three versions of Act I, the original, the alternative, and the abbreviated. 32 He decided he needed to introduce Lolita in Ramsdale before Humbert’s arrival, so returned to the original opening of Act I, but eliminated her clumsy encounter in the dentist’s chair with Quilty and opted instead for Humbert at the McCoos’, on the piazza with Lolita, and at the school dance with Quilty.
Nabokov conflated, for sound enough reasons, what seemed the best bits of each version, but he did so perfunctorily. He left very spare instructions for his secretary, Jaqueline Callier, to amalgamate the blocks as she retyped the manuscript. That he was anything but in command of dates in his own life, let alone in five versions of Humbert’s (the novel, and the ancestral, the alternative, the abbreviated and the amalgamated screenplays), is almost comically indicated in his record at the top of his instructions: “Added from Brown to Blue Oct 1930 Screenplay.” The blue folder contained the shortened and revised version of the screenplay, which he had submitted to Kubrick in September, now remembered as “October,” 1960, here misrecorded as “1930.” Perhaps the “1930” date proves that everything subsequent, including the novel Lolita and Nabokov’s meeting with Kubrick, was his invention? Or could we accept the simple proposition that he made a mistake here, as in the screenplay itself, and, perhaps, in a much smaller way, in the novel?
Let us turn now to Lolita itself.
The argument that the final scenes of Humbert’s story—his meeting with the married Lolita and his murder of Quilty—are his invention or fantasy depends on a single piece of evidence: that Humbert says on the last page of his book “When I started, fifty-six days ago, to write Lolita, first in the psychopathic ward for observation, then in this well-heated, albeit tombal seclusion, I thought I would use these notes in toto at my trial, to save not my head, of course, but my soul” (310). According to John Ray, Jr.’s foreword Humbert “died in legal captivity, of coronary thrombosis, on November 16, 1952, a few days before his trial was scheduled to start” (5). Flipping back fifty-six days, we arrive at September 22, the day Lolita’s letter reached Humbert. Since the visit to Lolita, the discovery of the name of her abductor, and the murder of Quilty all take place over the next three days, when Humbert says he has been writing in a psychopathic ward, they are therefore, according to the revisionists, fabrications or delusions of Humbert Humbert.
In view of Nabokov’s fallibility, it seems much sounder, let alone much more economical, to call into question a single numeral than to doubt the detailed reality of a whole series of major scenes. It seems especially peculiar to suppose that virtually everything in the last eighth of the novel is fabricated, except for the first nine words of the sentence quoted above: “When I started, fifty-six days ago, to write Lolita… .” Why, if even the trial mentioned in this sentence is Humbert’s fiction (as it usually is for the revisionists), 33 if the psychopathic ward too is a fraud (as it is for Dolinin and sometimes for others), is the “fifty-six” swimming in this sea of falsity to be fished out as incontestable fact?
That Nabokov could err in dating his own life and the lives of his characters has been amply demonstrated already. But let us be clear that—as the revisionists know—Lolita itself is no zone of immunity. To take the life first: in his November 1956 “On a Book Entitled Lolita,” Nabokov wrote that he had not “reread Lolita since I went through the proofs in the winter of 1954.” He altered this to “in the spring of 1955” for the Annotated Lolita (318, 439), although in fact he received the first proofs only well into the summer of 1955 (July 12).34 Only sixteen months after the event, in other words, he had been inaccurate by about eight months.
Within Lolita’s fictional world, he also made at least one incontrovertible error that immediately casts doubt on the value of the “fifty-six days.” The morning Humbert comes down to check the mail is “early in September 1952” (266, 426), according to the 1955 and 1958 editions. Yet three pages later we find “The letter was dated September 18, 1952 (this was September 22)” (269)—“this” being the day he receives the letter, hardly “early” in the month. Somebody, perhaps Nabokov himself, muted the mistake in the French translation, where the text has “vers la mi-Septembre.”35 In 1965 Nabokov altered the Russian translation by removing the “September 22” from the next chapter and bringing it forward to replace the vaguer initial reference: “for that particular morning, early in September 1952,” became “ibo v to utro, 22-go sentiabria 1952-go goda” (“for that morning, September 22 1952”).36 Several years later Nabokov supplied to Alfred Appel, Jr., for the Annotated Lolita, the correction to “late in September 1952” (266).
The undoubted mistake here that persisted in Nabokov’s manuscript, typescript, and through readings of at least two sets of proofs (Olympia’s and Putnam’s, though he also read the Crest edition, and presumably the Weidenfeld and Nicolson and Corgi editions) shows how little credence can be given to the unsupported testimony of “fifty-six days.” Here a discrepancy of two weeks or so occurs within a space of three pages, and was not picked up by Nabokov for over ten years. The “fifty-six days” as evidence depends on an error of only three days over a gap of over three hundred pages.
Not only on that, some might say. There is one other relevant change in the Russian—where, as Gennady Barabtarlo shows, dates have several times been supplied where there were none in the original.37 In the English edition, Humbert, after finding it impossible to trace Lolita’s abductor, begins a new chapter: “This book is about Lolita; and now that I have reached the part which (had I not been forestalled by another internal combustion martyr) might be called ‘Dolorès Disparue,’ there would be little sense in analyzing the three empty years that followed.” (255) In the Russian the last thirteen words become “ … podrobnoe opisanie poslednikh trekh pustikh let, ot nachala iiulia 1949 do serediny noiabria 1952, ne imelo by smysla” (234: “a detailed description of the last three empty years, from the beginning of July 1949 to the middle of November 1952, would make no sense”). Alexander Dolinin cites this in support of his theory, as proof that there was nothing but blankness—no meeting with Lolita, no encounter with Quilty—in Humbert’s life from Lolita’s disappearance until this point near the end of his writing his book. Humbert is, “therefore, at home, at his writing desk but not in prison awaiting trial, as he has tried to convince his gullible readers.”38 But if this were a major piece of evidence in the reading that Nabokov wanted to imprint on his book, why did he then not transfer it back into the Annotated Lolita when he incorporated corrections there that he saw as necessary in the Russian?
If the probability of a mistake in Nabokov’s numerals seems high from the first, the improbability of the revisionist view seems overwhelming.
Even a reader unaware of Nabokov’s capacity for error will see an immediate objection to the revisionist theory: what of John Ray, Jr.’s foreword? The November 16 date for Humbert’s death, on which the whole case rests, comes from Ray, but Ray also confirms just what the case tries to deny—what Dolinin denies outright—that Humbert was in prison awaiting trial when he finished his manuscript and promptly died. Tekiner suggests that—since, presumably, Ray does not explicitly mention that Humbert’s trial is for murder—the trial could be for his treatment of Lolita.39 But Ray declares that “References to ‘H.H’’s crime may be looked up by the inquisitive in the daily papers for September-October 1952” (6). There is nothing here or anywhere else in the foreword to imply that the inquisitive will find that the newspaper accounts utterly contradict Humbert’s.40
In his long discussion of the implications of the “fifty-six days,” Dolinin of course refers to the “November 16” date from which the countback starts, but curiously he never mentions the person who supplies that date, and never addresses Ray’s assumption that Humbert’s story coincides with the known facts of the case, the details of the murder listed even in the newspapers. But to ignore evidence does not make it go away.
Connolly at least takes note of the conflicting evidence, even if only to will it into oblivion when he suggests that Humbert may have invented Ray’s foreword.41 But if that is the case, then of course Humbert does not die on November 16, 1952, and there is no firm date from which to count back fifty-six days, and the discrepancy on which the whole case rests becomes nonexistent or meaningless.
Nabokov intended to indicate that Humbert died just after putting the last words to his manuscript. That is why he supplied the number of days Lolita took Humbert to compose, and why he has Dr. Ray supply the date of Humbert’s death, and why he explains in his interview with Alfred Appel that in Humbert’s final paragraph he meant “to convey a constriction of the narrator’s sick heart, a warning spasm causing him to abridge names and hasten to conclude his tale before it was too late” (437). If there is a discrepancy between the number of days Lolita took Humbert to write and the number of days until Humbert’s death, that seems an error all too easy to make. Either Nabokov simply used the wrong starting point, counting from September 22 (Humbert’s receipt of the letter), the one concrete date given in the novel’s concluding sequence of events, rather than from September 25 (the murder), which has to be inferred from the text, or he counted correctly but he—or the typesetter—put “November 16” rather than the intended “November 19” for Humbert’s death, making no more than the very common slip of 6 for 9. If the text now read “November 19,” the argument for Humbert’s having invented the last fifty pages of Lolita would immediately collapse. Surely it is too much to base a major reinterpretation of a novel on a single typographic character?
Nabokov always aims for exactitude. He does not allow us simply to lean on evidence, as the revisionists have to do; he makes it click into place. He has made a mistake in the dating, but what he has tried to do has his customary precision and point. Humbert admits that he has “wanted,” as he says in his final paragraph, just as he feels his heart twitch, “to exist at least a couple of months longer” (311) than Quilty. In fact, since he will have only a few more hours or even minutes, he will have outlasted Quilty by fifty-six days, or eight weeks: exactly two lunar months, but still just short of two strict calendar months. Playful Aubrey McFate as it were pretends to grant Humbert the two months he had asked for, then cuts him short, denying even that small request. That is the very exact, very Nabokovian irony of these final dates, except that somewhat—but not completely—uncharacteristically, and all too humanly, he has made a slight error.
I would not have written this article if only one critic had proposed the revisionist hypothesis. I would have stopped here if two or three had propelled me into print. But with six already advancing the argument, another thinking about doing so, and others inclined to entertain it, I will continue.
Dr. Ray’s foreword records that Mrs Richard Schiller dies in childbirth in Gray Star, “a settlement in the remotest Northwest” (6). How has she reached there, if Humbert does not respond to her letter that says “I’m going nuts because we don’t have enough to pay our debts and get out of here. Dick is promised a big job in Alaska” (268)? Why does Nabokov in the afterword think of Gray Star as “the capital town of the book” (318) if Lolita does not die there in childbirth? (Gray Star, presumably, is Juneau, Alaska’s capital, in allusion to the old cartographic convention of stars for capital cities, but also a play on Juno, the goddess of marriage.) If Ray’s foreword is accepted—and to repeat, if it is not, “November 16” disappears as evidence, and takes with it the whole revisionist argument—it explicitly or implicitly confirms Lolita’s letter, Humbert’s visit to her, Quilty’s murder, and Humbert’s composing the manuscript in prison while awaiting trial for the killing, all the things the revisionists try to discredit.
So too does Lolita: A Screenplay.
While the screenplay reinvents minor details of the novel, its main alterations seem designed precisely to convey what Nabokov regarded as crucial to the novel but likely to be lost without considerable adaptation.42 First, and most important, Quilty’s shadowy presence throughout the novel, which readers can discover only after Humbert has himself dropped the name, is signaled in the screenplay by opening with a flash forward to the murder scene, and by then making him more prominent, once the narrative returns to the beginning, from the time of Humbert’s arrival in Ramsdale (at the school dance, where Quilty is presented as author of The Nymphet; at the Enchanted Hunters, where he is named as the drunken guest; at Beardsley School, where he is again named as author of The Enchanted Hunters). Second, the Edgar Allan Poe allusions, which Nabokov stresses, at the cost of some strain, through Humbert’s scholarly work and sometimes even Lolita’s schooling. Third, John Ray, Jr.’s position as frame to and external commentator on Humbert’s confession. Dr. Ray becomes the sometimes comically obtrusive narrator of the whole film.
Humbert cannot narrate the film, as he does the book, for his utter ignorance of the identity of Lolita’s abductor until the end is still crucial to the story. In the novel, he could introduce Quilty’s presence and yet keep his identity hidden until the right moment, thereby having the satisfaction of keeping the reader in the darkness he had himself found so unlaughable. In the film, he could not be the narrator and allow Quilty to be seen on screen without repeatedly disclosing his present awareness of Quilty’s role. By removing Humbert from the narration of the film and flashing forward right at the beginning to the murder, Nabokov alerts us to the identity of Humbert’s foe from the start, and therefore makes us vividly aware, whenever we later catch sight of Quilty, of Humbert’s failure to recognize his rival until the very end.
The screenplay opens with Lolita telling Humbert where Quilty lives—showing him, in fact, a magazine photograph of Pavor Manor which then comes to life as Humbert arrives and promptly, wordlessly, kills Quilty. Immediately afterwards, the camera cuts to:
Dr. John Ray
After Ray explains that Humbert’s memoir is “mainly an account of his infatuation with a certain type of very young girl,” the camera cuts to:
Humbert’s Cell in The Tombs
Obviously, there are differences, but they seem designed primarily to make the major effects of the novel possible on the screen. Dr. Ray exists objectively before our eyes, and he describes Humbert’s composing the manuscript in prison after committing murder (he does not explicitly specify Quilty as murder victim: does this leave a loophole for the desperate revisionist?). By indicating Humbert’s reference books, Nabokov establishes his character’s effort at reliability in retelling his past. And he lets us see the murder before Humbert sets down his story, even lets us see Quilty asleep in Pavor Manor before Humbert first appears on screen, before Humbert reaches the Manor himself. The murder, unequivocally, is not a product of Humbert’s narration.
The scene of Humbert’s reading Lolita’s letter, of which the revisionists make a great deal, is replaced by a parallel scene in the screenplay. Understandably, Nabokov has excised Rita from the screenplay as an unnecessary complication, and instead shows Humbert, after he loses Lolita and all trace of her abductor’s trail, teaching once again at Beardsley College. There he meets Mona Dahl, who quizzes him—years have passed—about Lolita. As Nabokov notes after this scene in an explanatory aside unimpeachably immune from revisionist skepticism: “It should now have been established that Mona has had a letter from Lolita, apparently asking her to find out if it is safe for her, Lolita, to write to Humbert” (198). Humbert picks up his mail at the university post office and heads straight to an examination he is to invigilate. He opens the letter, hears, just as in the novel “a small, matter-of-fact, agonizingly familiar, voice” (199)—and after reading through Lolita’s letter, he dashes, dazed, from the exam room.
In the novel, Humbert prepares for his unpreparedness for Lolita’s letter with the great passage about endowing “our friends with the stability of type that literary characters acquire in the reader’s mind. No matter how many times we reopen ‘King Lear,’ never shall we find the good king banging his tankard in high revelry, all woes forgotten, at a jolly reunion with all three daughters and their lapdogs. Never will Emma rally, revived by the sympathetic salts in Flaubert’s father’s timely tear” (267). Obviously something new is needed for the screenplay, both to prepare us for the surprise and to show Humbert’s unpreparedness: hence the device of introducing Mona’s questions, whose import we can see but Humbert cannot, and Humbert’s blandly opening the letter (“from a Mrs. Richard Schiller—some graduate student, I presume” , he had moaned in the mailroom) in the midst of the examination. And just as the novel stresses the shocked suddenness of Humbert’s response—he leaves without even waking Rita from her solid morning sleep—so does the screenplay, when Humbert lurches away from his post as invigilator. For all the changes in the treatment of Lolita’s letter, Nabokov has sought cinematic ways of stressing its credibility and of eliciting the same key responses in us and in Humbert.
33. Tekiner, p. 468, writes however that “The chronology implies that Humbert is in jail for his actions toward Lolita, rather than Quilty,” but does not explain how or why Humbert has been tracked down, or at what point prior to the supposed arrest for his treatment of Lolita (an arrest, of course, entirely without textual foundation) he began, as this conjecture would require, to suppress what was really happening to him, or why the conjecture does not square with the foreword (see below). Toker 218 suggests as one possibility (though she seems to prefer another) that “Humbert may have been arrested on the same day, almost immediately after reading Dolly’s letter, and placed in a psychiatric ward ‘for observation’ … prior to being scheduled for trial,” but though she rules out the murder of Quilty she does not suggest why he is being tried.
37. Aerial View: Essays on Nabokov’s Art and Metaphysics (New York: Peter Lang, 1993), pp. 135-38. The dates become more specific for several reasons: because Nabokov’s style evolved consistently towards greater chronological detail; because he felt he needed to identify for Russian readers in the late 1960s a period that was more self-evident to the Americans for whom he was writing in the early 1950s; and to correct inconsistencies he had noticed.
42. Toker, p. 210, realizes the awkwardness of the screenplay to her case, but rules it out as “a totally new work… . The screenplay, therefore, cannot be used to settle moot points in the novel.” It is indeed, as Nabokov says, “a vivacious variant” (Lolita: A Screenplay, xiii) on the novel, not a bland transposition, but as the examples will make clear, the screenplay strives even in its changes to be true to the novel.
Zembla depends on frames for navigation. If you have been referred to this page without the surrounding frame, click here.