SOURCE: "Five Ways of Looking at Aprille (with Apologies to Wallace Stevens): Analysis of Storytelling in the Twenty-First Century," in Eye on the Future: Popular Culture Scholarship into the Twenty-First Century, edited by Marilyn F. Motz, Bowling Green State University, 1994, pp. 91-106.

[In the following essay, Yaross Lee uses Keillor's story "Aprille" to analyze the effect of medium on a story.]

A slippery problem facing scholars of popular culture involves how to analyze examples that exist as multiple texts or performances rather than as a single stable artifact. Stable artifacts include the texts of popular fiction, tapes of radio or television broadcasts, and theatrical films or videotapes. The comic strip or book is somewhat less stable, since a scholar may have to grapple with the historical authority of the newspaper feature page versus the narrative authority of the anthology or comic book, but one can make a case for studying either version or both. A similar problem exists for some television series, which broadcast videotapes of live performances before a studio audience. All in the Family (1971–79), for instance, exists in two video forms: the master tapes of the live performances and the edited tapes of the broadcast series. Although the edited tapes captured most of the live performance's spontaneity—various pratfalls and glitches in performance became evidence of the taped show's authenticity—the cuts not only kept the show within necessary time limits, but also altered its substance: the off-color remark or gesture, long laughter from the audience. Still, a scholar can distinguish between the master and broadcast tapes in much the same way as between an author's manuscript and the published text. Despite the limitations of the broadcast tapes, which (among other things) obscure whether the audience laughed for 15, 30 or 55 seconds—they nonetheless are the authoritative texts for studying the work of producer Norman Lear and his cast. They represent commercial television, whereas the master tapes record only a studio performance.

By contrast with stable artifacts such as these, consider standup comedy and public storytelling. Both involve multiple performances that vary the material in many ways. Some variations are minor; others, substantive. Some represent a refinement; others, simply a variation with only subtle shifts in sense. Folklorists have made progress on but not really resolved the interpretive problems that result from multiple texts. Richard Bauman and Sandra K. D. Stahl have focused on the relation of each telling to its narrative context, for example, accounting for differences in the tellings in terms of the different narrative events. Contextual criticism thus downplays variations in phrasing and concentrates instead on theme, structure, and cultural significance. Seeing narrative details mainly as elements of a cultural message, however, gives the criticism a didactic thrust more appropriate to fairy tales than to the monologues of Johnny Carson or Spalding Gray. Under the circumstances, it is not surprising that even folklorists have found the thematic and structural approaches limiting. As Dell Hymes observed in his examination of a pair of transcriptions, "Each telling makes use of common ingredients, but it is precisely in the difference in the way they are deployed and shaped that the meaning of each is disclosed."

Variations matter even more in popular performances, which have become commodified as commercial art. Whereas the traditional folk narrator served as a medium of transmission, today's standup comedian or storyteller is the author of the material as well as its performer. The oral narratives of the standup comedian or public storyteller may sound like folklore, particularly the genre known as the personal experience narrative, and the anecdotes may in fact have originated in experience, yet the comedian and professional storyteller have no obligations to truthfulness. Audiences grant these performers the novelist's license to invent, and willingly suspend their disbelief. Variations from performance to performance are therefore more substantive than variations among performances of a folktale, even though traditional tellers commonly had their own, distinctive ways of telling a tale.

They are also more significant than variations in the performances of a scripted play. Whether introduced accidentally or deliberately by actors, director, crew, or performance space, variations on a script give different audiences different experiences of the play, but theoretically, at least, the performance remains somehow distinct from the play itself. Just as the musical performance is recognized as the approximate rendering of the ideas in the score (Dehnert), the dramatic performance is recognized as an approximate rendering of a definitive text. Such a text does not even exist for standup comedy and public storytelling, although we tend to behave as if it does. For a recent example of this phenomenon, consider the efforts by Lenny Bruce's producer, Don Friedman, to recreate the comedian's 1961 Carnegie Hall performance in commemoration of what would have been Bruce's 67th birthday. The fifteen actors who auditioned in July 1992 based their impersonations not on Bruce's own monologues, which were of course ephemeral, nor even on Bruce's own recordings, which are more stable, but on the stable 1974 text of Dustin Hoffman's re-enactment of unstable routines in the film Lenny.

As oral genres grounded in colloquial talk and tending toward improvisation, standup comedy and storytelling reverse the stage play's implied priority of written text over live performance. And not only is the performance more definitive than the written text: the various oral performances of a single story or routine vary in authority in relation to one another. Unlike the multiple video-texts of the television broadcast, one public performance will not have more authority than others, unless some are designated rehearsals or trials. Unlike the successive drafts of a manuscript, the most recent performance does not always stand as the artist's last word.

A particularly illuminating case in point is Garrison Keillor's 1986 story "Aprille," which nearly defies classification. The story takes its inspiration from The Canterbury Tales, builds its theme on a passage from the New Testament, and blends his wife Ulla Skaerved's recollection of a childhood game on the bus with half a dozen fictional Lake Wobegon anecdotes—all the while purporting to tell his own experiences. So, "Aprille" is not folklore, not even "literary" folklore like the personal experience narrative, though it imitates folklore: the monologue is a professional performance presenting fictional personal experiences of narrator Garrison Keillor of Lake Wobegon, Minnesota (a fictional town)—all of these created by writer Garrison Keillor of Anoka. Nor is "Aprille" a short story, since it was composed primarily for oral performance, not for print. Nor is it television, since the Disney cameras that recorded the show (originally transmitted live via cable and later shown over public television) shot it from the vantage point of the studio audience—as a radio program being witnessed by an audience, not as a show that was meant to be telecast. But the monologue is not conventional radio, either, since the speaker engages in intimate conversation and lies instead of the public and factual material that make up normal radio talk. All these modes of storytelling contribute to the tale's humor by inviting and then frustrating our generic expectations.

More important, the tale itself invites us to examine relations between oral and written storytelling, since Keillor structured "Aprille" around the ten-line opening of Chaucer's "Prologue," giving the oral story a thoroughly literary grounding. The recitation not only reminds us that The Canterbury Tales itself presented purportedly oral stories as written texts, but also calls attention to the difference between the text-bound activity of memorization and the performance-based activity of improvisation in Keillor's own narration. In addition, as Chaucer's words set the springtime Minnesota scene for the main story—how Lois Tollerud, a young woman troubled by the existence of evil in the world, does not find her faith on Confirmation Day, yet nothing happens as a result—Keillor also establishes a series of thematic parallels among the three pilgrims of the tale. Lois, Keillor, and Chaucer all undertake spiritual journeys that turn into occasions for storytelling, and in "Aprille" as in The Canterbury Tales, the pilgrims' stories become more important than the religious goals inspiring them.

Despite the seriousness of its themes and structure, "Aprille" remains typical of Keillor's work: it is also a very amusing story. Narrator Keillor's own pilgrimage fails when he arrives in Lake Wobegon and finds that his whole family apparently ran out the back door, at which point his journey to see them devolves into a quest for a toilet. His goddaughter Lois is also on a quest (she's looking to regain her faith in God), but at first they find only each other. As one anti-climactic anecdote leads to another, the humor builds in a conversational, apparently unstructured way that belies Keillor's intense labor on it.

The monologue was the centerpiece of a performance celebrating the grand re-opening of St. Paul's World Theater on Friday, April 25, 1986, and it was broadcast live the following night during his regular Saturday evening radio show, "A Prairie Home Companion" (1974–87). Keillor's popularity in 1986, soon after he agreed to allow the Disney Channel to cablecast the weekly show, made the April 26 broadcast extremely important to him, and even after the opening "concert" (i.e., non-broadcast) performance on Friday evening he continued tinkering with the details and themes of the story all the way up to broadcast time on Saturday night. Before each oral performance, he worked at his word-processor, printed out a draft, and edited it by hand. The result of this process is five variants of the tale: the Friday computer printout with handwritten emendations (version 1, 4/25/86), the Friday evening broadcast (version 2, 4/25/86), the Saturday computer printout with handwritten emendations (version 3, 4/26/86), the Saturday evening broadcast (version 4. 4/26/86), and the published Leaving Home story (version 5).

The first text (version 1) contains all three of the main stories—Lois's lost faith, Einer Tingvold's lost binoculars, and young Gary's fear of being isolated from family and friends—along with a fourth anecdote about the Tolleruds' agnostic Uncle Gunnar, which remained in all three written versions but was never included in an oral rendition. But several crucial details in Lois's story changed after the first performance, and the ending of the narrative continued to evolve through the second performance (version 4). Several of the most significant changes in his second performance did not, however, find their way into the published version of "Aprille," which appeared in Leaving Home, a collection of Lake Wobegon tales published shortly after "A Prairie Home Companion" went off the air. Although Keillors introductory "Letter from Copenhagen" described Leaving Home's stories as "written for performance on the radio," this fifth, published version of "Aprille" is neither a transcription nor a reworking of either oral performance, but rather a minor revision of the second printout (version 3).

Together the five variants of "Aprille" illustrate the problems inherent in analyzing the unstable texts of oral popular culture and point the way toward more systematic and sensitive analysts of storytelling in the twenty-first century. We need to find techniques appropriate to these texts and to understand the increasingly important roles of print, broadcasting, computers, and audio tape in defining them.

"Aprille" illustrates the fundamental difficulty of identifying exactly what is the story, since unlike the successive drafts of a manuscript the five variants of the tale do not exist in simple chronological relation to one another. Additions to the written text carry over thematically, if not always literally, into the next oral telling, yet improvisations in the oral tale are seldom incorporated into the next draft. The process thus underlines the greater importance of the oral versions, and their greater autonomy, and suggests that the final written version of the story (version 5) remains less significant than the second oral performance (version 4).

The first printout, for example, shows Keillor working to expand the description of springtime in Lake Wobegon at the beginning of "Aprille." Next to the typed remarks about trees and birds, he added a handwritten note in the margin: "The NBFs are washing their sheets, a sure sign" (version 1). In the first oral performance that same evening, he provided an even fuller description, saying, "this last week the Norwegian bachelor farmers washed their sheets, which is a sure sign that the cold weather, cold weather is over, and they're starting now to think about the danger of, the danger of infection" (version 2). The repetition of the phrases "cold weather" and "the danger of" work like "um" or "uh," as voiced pauses suggesting that Keillor invented the phrasing on the spot, and this implication of spontaneity asserts the authenticity of the storytelling event: the speaker hasn't prepared his remarks in advance. In this case the phrasing almost certainly was spontaneous, however, since none of the other variants repeats it exactly.

The next day, the second printout (version 3) shows Keillor integrating into his electronic text the marginal note from the first printout, but not the remark about "the danger of infection" from the previous night's oral performance. Instead, the text notes merely, "and now the Norwegian bachelor farmers are washing their sheets—" (version 3). The remark remains in that form in version 5, the version published in Leaving Home 15 months later. In the second oral performance (version 4), however. Keillor altered not only the phrasing of the item but also its placement and details. He postponed mentioning the Norwegian bachelor farmers until after he had described Lake Wobegon's springtime scenery in detail, finally noting, "the Norwegian bachelor farmers hung out their sheets, this last week, finally washing their sheets after those long winter months. Now it's finally safe to do em" (version 4).

These variations in such a simple matter demonstrate that Keillor's weekly monologue had more in common with a jazz performance than with an impromptu speech. That is, the monologue was a planned performance featuring spontaneous talk. The demands of live radio made this risky enterprise all the more precarious, since a live broadcast cannot be edited if the program runs too long. In all his years on this tightrope, Keillor fell only once, in Juneau, Alaska on July 12, 1986, when he just could not bring his monologue to a close, even though he had written a nine-page text for himself. Keillor worked out the components of "Aprille" carefully but nonetheless embellished and altered it in performance. His written drafts therefore represent prompts for the performance, suggesting the directions of his thoughts and his thematic intentions, but remaining subordinate as texts to the public performances.

For example, consider the different reasons that the bachelor farmers finally do their laundry. In the first telling they wash the sheets because it's warm enough for the bacteria in them to breed; in the second telling they wash them because it's finally warm enough to risk coming in contact with water. Keillor does not seem simply to have changed his mind here; the two jokes are equivalent. The point is the eccentric variation on the traditional obsession with spring cleaning. The exact rite is less important than the fact that a rite exists because the rite is more ritualistic than practical. In the case of the Norwegian bachelor farmers, spring cleaning accomplishes very little. Washing their sheets—for whatever reason—will not appreciably raise their level of hygiene. Perhaps more important for a theory of oral narrative, equivalent variation is possible because oral tellings do not supplant one another in the same way that successive drafts do. Whereas interpretation of writing rests on the principle of final intention, in which the authoritative text is the last of successive written drafts, interpretation of oral storytelling requires a principle of simultaneous authority, in which each telling has equal validity.

If the substance of these differences matters very little, the fact that the differences exist matters very much, indeed. The anecdote is full in the oral performances and sketchy in the printouts, in a progression typical of oral retellings. Richard Baumann notes that stories tend to grow when retold, especially when the teller commands the attention of the audience without having to seize it from other parties present—the difference, for example, between yarn spinning around a campfire or in conversation and the full performance of a storytelling festival, in which the featured storyteller is expected to demonstrate virtuosity. But the relationship between Keillor's written texts and the next telling reflects the complexity of the hybrid genre within which he works. Although details become more elaborate in the retellings, episodes may be rearranged, truncated, or eliminated. For example, both oral performances of "Aprille" mention the day of a rainstorm, and in the second performance Keillor elaborates the detail by making a show of searching his memory for the exact day—"[I'm] thinking about an afternoon, like—well, Tuesday, or Wednesday aft—Wednesday afternoon, after that tremendous rain that we had in the morning" (version 4)—but none of the written versions, including the version published in Leaving Home, refers to a specific day of the week. By contrast, Keillor apparently considered Daryl's Uncle Gunnar important enough to devote about 250 words to his eccentricities in each of the written texts, including the Leaving Home version, but Uncle Gunnar does not make even a brief appearance in either oral performance. We cannot know whether Keillor dropped the anecdote to keep the monologue within the allotted time (one aspect of his virtuosity was surely his ability to work within the finely calibrated time intervals of live radio), or whether he lost interest in the material but failed to delete it from the computer file that ultimately became the text in Leaving Home. Any analysis of the story must take this inconsistency into account, but it points to the theoretical difficulty of assuming that versions in two media stand in strict chronological succession to one another.

Other differences, however, clearly show the artist changing his material and sharpening his vision of it. Apparently dissatisfied with the tale he told on Friday night, Keillor sat down on Saturday and made two major changes in the computer text that became version 3. He deleted three suspected miracles that inspired Lois to keep an open mind about her lost faith, and moved the anecdote about Einer Tingvold to follow rather than precede the tale of playing strangers on the bus. Integrating these changes led to others, resulting in very different themes and meanings between the Friday and Saturday versions, oral and written.

In Friday's versions 1 and 2, the possibility of miracles tempers Lois's crisis of faith. The hot water does not run out as usual on Saturday night, her mother manages to get a stain out of Lois's confirmation dress, and her father's arm does not burn when his new sweater catches fire from the candles on her cake. These events strike her as possible evidence that God does exist, and they allow "Aprille" to end with some optimism for Lois and for all the faithful. Removing them, on the other hand, diverts the tale from questions of God's existence to questions of human faith. It also allows Keillor to emphasize the parallels between Lois and himself: she becomes terrified when her prayers echo without response, just as he had been frightened as a child when another Lois, his aunt, pretended during their game of "Strangers" that she did not know him.

As deleting the miracles diminished the tale's optimism, so moving the Einer Tingvold anecdote altered the story's original theme. An example of faith sure-to-be-found in versions 1 and 2 became a warning about faith abandoned in versions 3 through 5. With the story of how Tingvold threw away the next day's breakfast eggs and his own beloved binoculars just because he was frustrated that an unruly group of Boy Scouts wouldn't learn semaphore signals, Keillor provides a highly comic parable on the dangers facing Lois in her doubts about God on the eve of her confirmation. The anecdote's new position after the story of playing "Strangers." a game in which a young Keillor and his aunt Lois pretend not to know one another until the boy feels frightened and lost, emphasizes the consequences of throwing away one's religious values in the heat of a moment's disappointment. At the same time, the binoculars need not symbolize faith or even integrity to provide a reason for telling Einer's tale, since its details offer the considerable pleasures of comic retribution. As a result, "Aprille" not only implies the orthodox conclusions that Lois's faith in God matters less than God's faith in her, and that neither God nor faith will abandon her if she does not throw them away or try to become someone she is not, but also insinuates the downright blasphemous notion that faith is irrelevant to a good story. In this context, the parallel between Lois's story and Einer's also intensities relations among the three storytellers in "Aprille": Keillor, Lois, and Chaucer.

Eliminating divine intervention (or the perception of it) also required changes to the ending of the monologue. In the first oral performance, he started with the simple conclusion of his written text—"For the fourteen year olds of this world. I'm glad to get old so they can grow up and we'll see what they do with themselves" (version 1)—and extended it into a comment on how the world as a whole benefits from the courage of the young, "who having lost their faith could stand on the edge of darkness and wait for it to return" (version 2). By the next day, however, his written text proposed that loss of faith has less to do with God than with ourselves. People get caught up in games like "Strangers" and, after a time of playing at being someone else, forget who they really are. Nonetheless, God's grace remains as reliable as the signs of springtime: "the sweet breath, the tendre croppes, and the smale fowle maken melodye—God watches each one and knows when it falls, and so much more does He watch us all" (version 3). The power of Chaucer's poetry and Keillor's rhythmic line, a resolution to the question of evil, and the storyteller's virtuoso control over an apparently formless tale bring the third version of "Aprille" to a powerful close.

But although Keillor carried over this ending from version 3 to version 5, the text published in Leaving Home, he did not repeat it in version 4, his final oral performance. Apparently on inspiration in that Saturday night broadcast, he loosened the connection between spring and reborn faith. Instead, he proposed a more generalized power in "this world, and each other, and the people in it": "Well, I'm transformed by this world, the one that I look at. It's so beautiful. I believe that it has the power to make us brave, and to make us good. This world, and each other, and the people in it. It has the power to give us faith, the sweet breath of the wind and the tender crops, and the small fowles maken melodye that slepen all the nycht with open eye" (version 4). Whereas the written texts of versions 3 and 5 offer the traditional Christian insistence on faith in spite of doubt, the oral text of version 4 offers consolation through poetry, not doctrine.

My point is not simply that Keillor declined to revise the second written version to incorporate changes made in the second telling, or even (as I suspect), that at some level of consciousness he chose an orthodox conclusion for his stable written texts and a more ambiguous ending for his unstable oral texts. Rather, the issue is, what finally is "the story"? The lesson of "Aprille" is that scholars of storytelling need to answer that question on an individual basis for every tale and teller. For now we must conclude that the story is not, to borrow terms proposed by Barbara Herrnstein Smith, some "basic story" consisting of elements common to the five variants, nor a Platonic ideal of the tale constructed from bits and pieces of the variants. Moreover, a reliable analysis cannot look solely at the published text of an oral tale, no matter how much one would like to replace multiple unstable texts with a single stable one. Any reconstruction of "the story" must take into account sequential, substantive changes that amount to revisions, as well as accidental variations, like the bachelor farmers' laundry day, that represent what Erving Goffman calls "fresh talk."

"Aprille" is, therefore, a different story in each of the two media that its author worked in and in the different versions in each medium. The tale of faith lost and found clearly belongs to the stable world of print texts, as represented in Leaving Home. The more ambiguous quest for faith in the oral versions of "Aprilte" remains, appropriately enough, unresolved in the unstable texts of the performances. We cannot really understand "Aprille" if we do not know whether the ending envisions a world made wonderful by God or by the people in it, whether the answer to the problem of evil in the world lies in religious faith or in awe over nature. Yet the nature of oral storytelling means that instead of choosing one ending, we must somehow—like Keillor himself—embrace both.

The discontinuity between the fourth and fifth versions—that is, between the second performance and the published version in Leaving Home—reminds us that every oral telling begins afresh, whereas successive printouts from a word processor reflect only changes wrought upon the electronic text. The five variants together illustrate the creative process of a single mind, but as texts the oral and written versions of "Aprille" stand somewhat independent of one another. Instead of looking at the five versions as a single series, then, we need to see a more complex relationship among the variants. The three written variants (versions 1, 3, 5) have an evolutionary relationship to one another, yet the second printout has a special significance. It incorporates ideas stimulated by the first oral telling and plots out the most important changes incorporated into the second telling. For similar reasons, the second telling has somewhat more authority than the first, since Keillor incorporates into the Saturday night version ideas first used in the Friday performance and the Saturday draft. On the other hand, Keillor's practice of improvising his narratives like jazz performances rather than memorizing his texts like scripts means that every oral telling is an authoritative text somewhat autonomous from the most recent written text. An oral telling has an evolutionary relationship to previous performances yet stands independent of them.

The different endings of different versions thus offer yet another contrast between the unstable texts of oral stories, and such stable texts as Chaucer's poetry and the Bible. Oral stories are unidirectional, unlike written texts, which allow a reader to flip back and forth among the pages. And whereas the intent listener concentrates on the last joke only at risk of missing the next gem, a reader has the leisure to explore implications between the lines. In the context of three subverted pilgrimages, the dominant theme of the written tale, whether God loves Lois seems not so much paradoxical (the implication of the oral story) as irrelevant. In the end, the "Aprille" of the written texts characterizes religion as the narrative means to a narrative end—a subject for tales and an excuse for storytelling. Perhaps the goal of the oral yarn was to encourage religious faith, perhaps not: the end at once promotes and suspects piety. But in the written tale, which acquires a different emphasis, religion itself has a goal, and that goal undoubtedly is storytelling itself.

For this reason, transcription will never provide a satisfactory solution to the problems of studying oral performance. The process of transforming an oral text into a written one also gives the unstable oral text the appearance of a stable printed text. On that basis, audio tape is probably not the answer, either, though it represents a tremendous improvement over the inadequate transcript. Audio-visual CD-ROM technology, on the other hand, offers a superior means of storing, retrieving, and perhaps even quoting excerpts from oral narratives, since it can evade some problems of transcription by using the speakers already installed in personal computers (a sound board will improve the fidelity), and CD-ROM technology already in use makes it possible to search for specific phrases and hear them recited while the text also appears on the screen. With appropriate software, scholars could input their own material for such random access and analysis, which would offset the unidirectionality of live and conventionally recorded storytelling. Ideally, we could search not only for keywords but also for inflections, pauses, and mannerisms—all of them represented as digitized patterns—without the trouble of optical scanning or other forms of data entry. In this Utopian scheme, the oral source will continue to be examined as an oral source, not as a visual representation of an oral source. Newer technologies may offer better alternatives.

While transcription therefore threatens any oral performance, it is particularly damaging to "Aprille," whose oral performances rely on devices for which no print-based equivalents exist. How can the printed page express the unselfconscious tone of voice, the appearance of artlessness and spontaneity, or the perfectly timed pause? Keillor's practice of plotting out his monologues but not scripting or memorizing them gave the tales elements of literary composition while fostering genuine improvisation in performance. When he structured "Aprille" around a ten-line passage from The Canterbury Tales and lengthy verse from the New Testament, he intensified this already risky game of memory and improvisation. A printed quotation cannot convey the least degree of artlessness or spontaneity. Nor can it demonstrate Keillor's remarkable memory, which produces these effects in the first place. By contrast, in the oral performance of "Aprille," Keillor's memory not only sustains him through a long recitation in passable Middle English, but also enables him to embellish the prepared story on the spot. Without this display of memory, the story loses other crucial elements: our awe of the storyteller, apparently spinning this yam just for us as we sit enchanted by his gift; and our delight in language, perhaps the most powerful oral gratification. Seven pages of reading cannot match the rhetorical impact of Keillor's 30-minute performance—he called it a seance—for the very medium of print inhibits, though of course it does not entirely obscure, the tale's celebration of words—Chaucer's, Keillor's, and the Bible's.

In the oral performance of "Aprille," even scripture is oral. Several times the storyteller recites parts of Lois's confirmation verse, relishing its sounds, but he quotes the full text only once, when describing how Marilyn Tollerud inscribed all 31 words—"Be not conformed to this world: but be ye transformed by the renewing of your mind, that ye may prove what is that good, and acceptable, and perfect, will of God"—in blue frosting atop Lois's cake. One could hardly ask for a more oral use of a text than to turn it into food, and then to turn the food into a joke for oral performance. This gentle ridicule of the Tolleruds' celebration quite literally, and not so gently, reduces the Bible to a mouthful of words, of which Keillor's share on his square of cake reads "Con but for"—a hint at the humorist's con-game. The long recitation of the verse on the cake exploits the advantages of oral performance by demonstrating the teller's prodigious memory, all the more impressive at the end of the narrative when he reasserts his control over all his apparent digressions by reciting Chaucer's poetry.

But no moment of "Aprille" better illustrates the essence of oral yam spinning, and Keillor's brilliance at it, than his modulations and pauses at a key point in mid-story. (The phrasing varies from version to version, but the episode exists in all five variants.) Slipping from his role of observing narrator, he adopts Lois's point of view to relate her thoughts on good and evil during a solitary walk in the woods after lunch, while her family eats the cake. He continues to tell the story from Lois's perspective as he describes her frantic efforts to run away from a man standing by the road, whom "she knew … was put there by an evil force … and that this was Evil roaming the world, and looking for whomever it may devour" (version 4, 4/26/86). Tension increases as she falls, begging him for mercy. "'Please, please,' she said, 'don't do it.'" Keillor's voice drops to a whisper in the tension. Then he pauses, raises the pitch, and in a slightly bewildered tone, adds unexpectedly, "Which surprised me…." In the pause following this wrenching shift of perspective, we realize that our storyteller was the man standing there, that Lois made a mistake; the revelation rescues Lois and the tale comically as it evades the question of evil. Nonetheless, this exquisite moment from the oral performances, a rhetorical tour de force, does not appear in Leaving Home. In the printed version of "Aprille," a new paragraph just under Lois's words begins simply and directly, "I hadn't seen her for five years." Whether or not he had added the remark about his surprise, Keillor might still have revised the written text to startle his readers as he had shocked his listeners—that is, to translate the oral-aural experience into a literary one. But he didn't. In place of the stunning oral performance, the printed text offers the aesthetic compensations of print, careful construction and themes worth reflection.

The power of the monologues also derives from a number of non-narrative elements in their performance. Keillor's voice—intense, sincere, companionable—was a major factor in his stories' appeal, much enhanced by the intimacy of radio as a medium and his understanding of how to exploit it. The monologues portray one colloquial, confessional voice speaking directly to the individual radio listener—one companion chatting to another. The obviously live, unscripted performance added to the illusion of spontaneous conversation. So did the choice of "panned mono" technique for the stereo broadcasts, since this sound mix emulates live, directional, non-broadcast speech. But Keillor's genius lay in matching his message to the medium. As he worked out the Matter of Minnesota for oral storytelling, he borrowed techniques from oral narrative traditions. In these traditions, Walter Ong observes in "Writing is a Technology That Restructures Thought," oral stories link present and past, memory and orality, knower and known—all in the "simultaneous present" created by live narration. The fundamental orality of "Aprille," despite its literary structure and literate story material, points to the source of the Lake Wobegon monologues' appeal: they mimic our most beloved kinds of folklore.

"Aprille" and Keillor's other apparently improvised monologues, like those of many standup comics and professional storytellers, are invented stories masquerading as personal experience narratives. That is, they are literary fictions imitating folk legends and artistic performances imitating unpremeditated talk. Autobiographical experience inspired many Lake Wobegon anecdotes, yet Keillor consistently subordinated facts to his narrative purpose, fully fictionalizing them. Similarly, interaction with his audience helped shape the narration of his oral stories even though the rules of professional storytelling performances allow a studio audience only the limited responses of laughter and applause. In conversational storytelling, by contrast, listeners may ask questions if they're interested, and seize the floor from the speaker if they're not. They may also cue the performer in any number of nonverbal ways to alter the pace of the tale. But just as his stories are not folklore, neither are they "radio talk," as Goffman terms it. Announcers are supposed to talk about real events of moment in the real world, not invented anti-climaxes in an imaginary town, and at least part of the monologues' humor from the very beginning stemmed from their violation of listeners' expectations. Radio storytelling separates teller and listener even further than the professional story performance. A radio audience has no active role to play during narration. Neither the performer nor anyone else notices if a listener falls asleep, walks away, or even turns off the set (commercial radio has a vested interest in identifying programs that cannot hold their listeners, but public radio, lacking sponsors, has no such audience research). Despite such odds against him, Keillor at the height of his popularity drew an estimated audience of some four million listeners to the weekly broadcast of "A Prairie Home Companion," and they certainly were not passive. Each new story in the Lake Wobegon saga brought letters of commentary, suggestions for stories, and gifts to the storyteller.

Keillor's success as an oral storyteller came largely from his transformation of the personal narrative into a narrative genre appropriate for the electronic age. In that sense, he finally accomplished in the 1980s the return to tribal orality that Marshall McLuhan first predicted some 20 years earlier in Understanding Media, although it is his skill as a writer that makes this achievement possible. But of course Keillor's stories belong to the electronic hearth, and as a result they exemplify the ephemerality of communication in an electronic age. Radio stories leave no artifactual texts. Computers allow writers to write and overwrite the same text, bringing us many authoritative texts, even though some revisions will remain invisible, existing only on a disk. In this sense, despite its origins in such canonical literature as the Bible and The Canterbury Tales, "Aprille" belongs to the world of post-modernism, in which instability, ambiguity, and multiplicity rule.

In its liminal position between the literate and the oral, "Aprille" challenges scholars to move beyond reductive transcriptions that put oral texts into print. Instead, we must interpret the artifacts of popular culture—print, broadcast, computer texts, audio and video tape—in the terms of whatever media they use. For broadcasts and other unstable artifacts, that means acknowledging simultaneous authority of multiple texts rather than seeking the stable meanings of a writer's last intention. The list of popular narrative media will surely grow in the twenty-first century, and our theory must be ready to meet them.