The Next Sound You Hear

... May well be a Synthetic Creation of that Electronic Frankenstein the Tape Manipulator

My First Encounter with the Deepening Shadows of tape-recording make-believe occurred when, shortly after a book of mine had been published, I was invited to come and be interviewed on Monitor, NBC's marathon network radio program. I went, as per my letter of invitation, to the office of an amiable chap named Fitzgerald Smith. He chatted with me about the book for a while, and then said, "Let's go into the studio and tape the interview. I prefer to ad-lib my questions, but please be sure not to overlap me -- just wait a beat after each question before you start." I must have looked puzzled, for he added with a wry smile, "I don't go on the air, of course. If you do all right, we type up a list of the questions I've asked, and one of our high-priced names records the list when he has time. Then I'm edited out, he's spliced in, and that's what goes on the air."

One afternoon several weeks later, on the NBC network, I was apparently in brisk discussion with two well-known NBC reporters, Peter Roberts and Al Collins. Millions of listeners throughout the country heard these two and myself discussing my book; but the Messrs. Roberts and Collins wouldn't have known me if I jostled them in an elevator. It was enough to make a man's flesh creep.

I was intrigued enough to make a few inquiries, and discovered that this was but one of hundreds of such bizarre events -- if one can really call by the name of "event" that which never happened. Jascha Heifetz, for instance, had recently recorded a Bach two-violin concerto by himself -- the first two-armed man in history to play four arms' worth of music -- and Les Paul and Mary Ford, in an excess of egomania, had multiplied themselves in whole string bands and soprano choruses. A singer I know told me she had just taped the leading role of a Broadway musical in the New York studios of a major recording company, accompanied by an orchestra that had never left Europe and had done its playing a year earlier; moreover, she had even sung a brief duet with an ailing baritone who showed up a week later, when he was better, to do his half of it. And quite recently, as I learned, an interview with Gil Hodges, in which he mentions Post cereals' baseball trading cards, was sent out to hundreds of local radio sportscasters with blank time left so they could dub in, in their own voices, the questions to his answers, and thus impress audiences with apparent acquaintanceship with a baseball VIP.

It struck me that all this was more than a technological advance; in my own time, it was getting difficult to keep one's grip on the meaning of reality, and ever easier to ascribe historical actuality to happenings that hadn't occurred at all, but had been synthesized. It is true that illusions of various sorts have always been acceptable in the performing arts, from the masks of the ancient Greek actors to Mary Martin's flying across the stage in Peter Pan, but these have not involved any real trickery of the spectator. The civilized adult has always known he was watching an illusion, and accepted it as a part of the esthetic experience -- quite a different thing from being fooled about the nature of what he is perceiving, and deceived about who is doing or has done what. And if true of artistic performances, this has been true all the more so of reportorial or argumentative performing. Listeners have willingly accepted the rhetorical devices and gestures (continued on page 114) Next Sound (continued from page 95) of debaters, but they have also been sure that they were hearing the real words of real people, rather than synthetic speeches and composite discussions.

Now, however, the borderline between showmanship and outright deception is so blurred as to be, perhaps, impossible to define. A speech, an exchange of opinions, an artistic performance -- all events which have always had coordinates of time and space in the real world -- are now being assembled out of raw materials and given a set of coordinates retroactively. A couple of years ago for instance, one of the most famous concert pianists alive began to record a short bravura piece faster than he, or anyone, had ever played it. Unfortunately, he kept hitting fistfuls of clinkers, but his pride would not allow him to back down. The RCA music director in charge tactfully said nothing, and merely recorded it again and again; afterward, he spent agonizing hours cutting up the tapes, splicing together the good bits and scraps, and finally achieving a dazzling 13-minute, 72-splice performance.

As almost everyone knows, this and similar small miracles of creation are the result of the fairly simple mechanical processes involved in tape editing by which sounds are cut up and recombined as easily as arranging a bridge hand, and which, if skillfully done, are undetectable afterward. Until the last century, none of the verbal or musical sounds created by human beings were recorded (except in the indirect form of writing), and hence none were ever tinkered with in the modern fashion. Even after Edison invented his phonograph in 1877, there was none of this kind of Newthink; Edison's method involved cutting a groove on a rotating cylinder (and later on a disc), and the groove thus existing on a plane or curved surface was fixed and unalterable.

But if the spiral groove on a single solid surface could somehow be unwound and strung out, it might indeed be cut up and recombined as one wished. A tape recording is, in a sense, just that -- not a groove, of course, but a narrow strip on which sounds are arranged in linear fashion; with time thus strung out, the things that happened in it could be broken up and moved about.

None of this was in the minds of the early inventors of magnetic recording. By the 1920s, engineers in Germany, Italy and the United States were experimenting with both magnetizable wire and tape, in an effort to record sounds more faithfully and durably than they could on the phonograph record. Just as easily, too, the recordings could be erased by demagnetizing them, thus making the tape reusable.

Erasure is, in fact, so easy that it was long responsible for a chronic neurosis among sound engineers -- one had only to push the wrong button and, instead of playing back what he had, he instantly and irrevocably lost it. A host of myths has gathered around this fact, a typical specimen of which goes as follows: A dozen years ago an engineer named Willie Goldsmith, working for the Carnegie Hall Recording Company, pushed the wrong button while cutting a master from a tape recording of Jacques Abram playing a Chopin piano sonata. All was quiet for a few seconds: then Goldsmith reacted, but half a dozen fast measures were gone. So, alas, was Abram -- gone on a Western tour -- but the record company which had subcontracted the job to Goldsmith's employer was committed to releasing the disc in a matter of weeks.

Goldsmith's fix was complicated by the fact that in his teens he had been a piano student at Curtis Institute, and had once studied this very sonata. For two nights, he couldn't sleep, harboring both his secret and his temptation. Then he gave in to his dreadful craving: after rehearsing the missing measures for many hours, he stayed at the studio late one night and secretly recorded them. Out of a score of tries, he had one perfect take: this he spliced in, and finished the master disc, which passed the client's and the soloist's later inspections without any difficulty. He told his boss about it only years later, and neither Abram nor the record company have heard about it to this moment.

The Goldsmith-Abram collaboration would not have been possible if wire had continued to be the favored method of magnetic recording, as it was before and during World War II. Early tape was much inferior to wire in fidelity and durability, but editing wire involves tying tiny square knots in the hair-thin stuff. Splicing two notes precisely thus requires even more luck than skill; moreover, neither luck nor skill can prevent each knot from making a distinct click as it crosses the playback head.

In 1944 and 1945, however, U.S. Signal Corps officers discovered tape recorders and excellent tape in abandoned German positions. Coincidentally, both the Minnesota Mining and Manufacturing Company and the Armour Research Foundation were independently developing another and even better kind of magnetic tape. A handful of enthusiasts started demonstrating the virtues of tape-recording to broadcasters, and when Bing Crosby saw how perfect programs could be easily assembled from numbers done separately, he switched from live to taped shows during his 1947--1948 Blue Network season. By 1948 the switch was on both in radio and recording studios. Within a year tape became the medium of choice, and last year upwards of $40,000,000 worth of tape was sold for sound recording alone.

Today's sound tape -- improved, but substantially the same as that perfected by 1947 -- is a quarter-inch ribbon of cellulose acetate or plastic film, on one side of which is a coating of ferrous oxide in a resinous paint. To cut such tape, one needs only a scissors or, preferably, a single-edge blade (a splicing block helps, by making the angles of the cuts identical); and to patch it, one needs only a reel of splicing tape with which to hold the cut ends together. If the cut is made diagonally, either in the middle of a silence or in the middle of identical sounds, and the ends carefully brought together, the patch creates no click or noise as it is played back.

The most obvious advantage this offers is this chance to cut out anything embarrassing or dull, and this accounted for the first kind of editing done. In 1949, a Saturday-afternoon performance of the Metropolitan Opera was taped in its entirety and scheduled to be broadcast that evening. Lily Pons sang Lucia di Lammermoor, and near the end of the Mad Scene, she ran up to a high C, slipped, and landed ignominiously on her B flat. At the studio, a kindly tape editor removed the offending note (which, musically, was expendable), and that evening the radio audience heard the revised edition without knowing the difference. All but one gentleman, that is, who had attended the matinee, and, having an odd sense of humor, had assembled a group of musical friends that evening to hear the fractured aria and to make merry about it. When the moment came and went, all perfection, his rage and chagrin knew no bounds; within minutes he was on the phone threatening to sue the network, though it was not clear what for. The threat came to nothing, and editing went ahead from that simple beginning to become a fine art of anatomizing and rebuilding both music and speech. For the converse of removing an unwanted segment is to collect wanted segments and put them together. At first -- and most of us can remember those days of innocence -- this meant assembling single scenes into a whole drama, or selected interviews into a documentary show, but through gradual refinements of technique, it came to mean the selection, rearrangement and patching together of fragments as small as a sentence or a musical phrase -- and even of single words or notes.

To see how this is done, I watched Lee Hanna, Director of News and Special Events at WNEW Radio in New York, edit several hours' worth of interviews and commentary on mental health into a half-hour documentary. Hanna, a snubnosed, (continued on page 192) Next Sound (continued from page 114) boyish-looking man, stood before a waist-high Ampex machine with a pile of reels of tape on a nearby table. He flipped switches, sending tape spinning back and forth from one reel to the other as he listened for useful paragraphs, or sentences, and skipped by dull stuff or repetitions. Sometimes the sound rose to a squeaky gibbering as he went into high speed; sometimes it descended to a deep growl as he hand-turned the reels so as to locate the exact ending of a word. With red grease pencil he would mark the tape at this precise point, and cut it with a razor, on a splicing block; then would listen until he came to another sentence, or sometimes a phrase, that went with what he had before. Bringing the beginning of this and the end of the other piece together, he'd align them and deftly slap a bit of white patching tape over them; the result was a new and better bit of talk.

In this fashion, the mound of material gradually shrank to a "rough cut" of about 40 minutes of the best questions, answers, interviews and comments. In the process, a spaghettilike heap of "garbage" or discarded tape grew around Hanna's feet. He had also draped a number of short excerpts around his neck; these, he told me, were his "and-the" stockpile -- a collection of simple words he might later need to make complete sentences out of roughly cut speech. The next day, when I returned to watch. Hanna polish his material, he removed a dozen extraneous words and about as many uhs and ers from sentences, inserted three or four ands as well as some very brief pauses, and cleaned out a number of noisy intakes of breath. "Actually, this one hasn't been bad," he said. "But the other day I had to clean up one of our famous four-star generals. What a job! He has a terrific whistle on his final Ss. I had to make 15 splices in one little minute-and-a-half segment to do for him what his dentist should have done."

Even this, he said, was far from the ultimate in reconstruction; and he had another engineer, who is a speech specialist, show me how bad stammering could, with a good deal of skillful work, be turned into perfectly normal speech. Single words, moreover, could be rebuilt out of excerpted vocal sounds, providing one had enough knowledge of phonology. "Would," for example, has a relatively long lead-in and the "d" tapers off slowly. If not handled correctly, the word can start with a kind of thud or end with a "t" sound. "It's possible," Hanna summarized, "to do almost anything, especially if you have someone with a slow way of speaking, and pauses between his words or phrases." (Since even the external traces -- the splices themselves -- can be removed by copying an expertly edited job onto a "virgin" tape, most courts will not admit taped conversations as evidence.)

Most tape editors are pleased with their ability to alter and improve what has been said, but some people in the communications field view it with misgivings. One of them is Martin Weldon, Hanna's superior and Director of News and Special Events for Metropolitan Broadcasting, an 11-station network. "A broadcaster has a duty to his listeners not to edit anything so as to present a fiction in the guise of truth," Weldon says. "But he also has a duty to hold his audience and to avoid dullness. Unfortunately, there isn't any boundary between truth and fiction, but only a great gray area. As soon as we do more than take a representative excerpt, we're venturing into that gray area, but how far we go is up to our own judgment and conscience. We never knowingly pervert anyone's remarks, but even in editing out speech faults, there's some element of falsification. U Thant, for instance, isn't fluent in English; he speaks slowly and with long pauses in his sentences. We have to clean him up -- it's deadly, otherwise. Sick or bumbling politicos sound healthier and smarter, when we edit them, but we can't help it. Yet that kind of thing alters the public image of a man. It's fooling around with history, and I'm uncomfortable about it. But I don't know the answer. Once you and your competitors have such facilities, it's suicide to reject them. Why, there are commercial firms who supply hundreds of radio stations throughout the country with a steady flow of 'open-end' interviews -- taped, one-sided conversations which enable someone from a local station to seem to be interviewing a celebrity. And this is all free, if you don't count the casual mention on the tape of a particular brand of cereal or seat covers or frying pans."

The natural advantages of tape have likewise enabled it to invade the field of musical performing and work its curious changes on the attitudes of artists and audiences alike. The basic and unexceptionable advantage is the cutting of costs. Len Frank, president of Carnegie Hall Recording Company, estimates that recording costs would be at least double without tape. He explains that before tape, if an artist made the least error while recording a three-to-five-minute side, the whole side had to be done again. When LPs came in, with their 20-to-30-minute sides, the problem of perfect performances became almost insurmountable.

But tape made it unnecessary to have a perfect performance in the studio. The recording artist today can either make several complete takes, and do a few additional short repeats of difficult sections, or if he makes a mistake, he can simply break off, go back a few measures, and pick it up again. Later, the best sections are chosen and spliced together, and the master disc is cut.

The editing rooms in which this plastic surgery is performed are generally of a type: a more or less soundproofed windowless chamber equipped with one or more massive tape machines; a control panel from which one can govern the volume, tone, tape speed, and so on; a couple of high-fidelity speakers and several chairs. In the larger companies a tape engineer runs the equipment, at the command of a music director or producer who uses a score and his notes; in smaller companies one man may do the whole job himself.

Physically, the steps in music editing look about the same as those in speech editing, but a great deal more precision is required, since the editor must follow the composer's score. Most music editors are former musicians, for the ability to read music is almost imperative. One small record company got by for a while with a tape editor who had an excellent ear but little formal knowledge of music, until one day a review of their latest record arrived from a music magazine asking, rather querulously, "Since when does one omit a repeat in a Beethoven symphony?"

The usual procedure is for the music director to listen to each run-through and short take, with the score before him, making notes as to which passages, and even which difficult notes, are best performed in each; then he excerpts his choicest bits and splices them together into a whole. The possible artificiality of the resultant performance bothers every editor to a degree, but some consider heavy editing not so much a species of dishonesty as a form of artistic creation. Seymour Solomon, the president of Vanguard Recording Society, told me that he is apt to spend four hours of editing for every hour of actual recording, and to make anywhere from 30 to 250 splices in a four-movement symphony. "Most of what we edit out are imperfections that are perfectly acceptable in a live performance. But in a record you're stuck with them; after four or five playings you may find them intolerable. In editing, we try to create an idealized performance in which the artist is playing as he might on one of those rare nights when his mood is exhilarated, the audience is brilliant and receptive, and from the first note he feels in total control."

Though some purists may find this a Svengalilike attitude, there is no doubt that producing such an idealized performance requires a high degree of skill. Music being a good deal more organized and subtle a form of sound than human speech, even the most delicate differences in shading or intensity -- especially of a violin or voice solo -- may make a splice in the middle of the passage quite apparent; it might, for instance, sound like a rough attack on the first note of the spliced-in segment. Again, the sound of a kettledrum creates a series of sharp, lasting reverberations; if a splice is made immediately following a drumbeat, the reverberations may not quite match, giving an effect of a secondary thump or bump. But these and many similar problems can be gotten around by skilled manipulation. The sharp violin or voice attack can be softened, for instance, by cutting the tape in a long sloping splice rather than at the usual 45-degree angle; the new note thus feeds in at comparatively low power for an instant, and so loses its brusque attack. Reverberation, mismatched to a drumbeat, is fixed by choosing an earlier spot for the splice so that the reverberations possess their own drumbeat. The loudness of any passage can be adjusted by "riding the gain" (working the volume controls) while re recording the music onto another tape. Shrillness of a solo instrument or singing voice can be masked by rerecording the offending notes through an electronic filter, and splicing in these filtered tones.

Whatever the merits of these electronic manipulations, the mere spliceability of tape has yielded some valuable performances that would otherwise be nonexistent. Aging performers of great interpretive power, but flagging energy or stiffening fingers, can, like the late Wanda Landowska, systematically record their matured conceptions in relatively short takes which are later assembled; the ribbon of acetate and iron oxide becomes for them that fountain of youth so vainly sought by men in earlier ages. Sometimes only tape can make a total idea of a work come to life. Toscanini would never permit RCA Victor to release records of his 1951 broadcast of the Verdi Requiem because it was something less than the perfect version he had in his mind, but in 1954 an RCA Victor editor brought him tape recordings of all the rehearsals and of the broadcast itself, and Toscanini, after reviewing the mass of material phrase by phrase, put together a performance made up of more than a score of spliced pieces. The world now possesses it; if it is not an authentic recording of a great performance that actually look place, it is at least Toscanini's private conception of a great performance.

As in the case of speech editing, no one can clearly define the limits where this passes beyond the boundary of ethical editing into untruth. In pop music, to be sure, the outlandish sounds that have been so popular on records have been known by the public to be "gimmicked up." Such oddities as speeded-up voices, distorted and jangling rock-'n'-roll piano numbers, or the lustful groans manufactured out of feeble adolescent bleating, were known to be synthetic and valued for their very artificiality.

Now most of these teenage delights arc on the wane. Young and old alike, however, may be replacing them with what might be termed the supergimmick. Columbia's new deluxe superstereo series is using eight-track tape "to provide exciting stereo experiences," to quote Ernie Altschuler, a producer for that company. Eight-track tape makes it possible to record eight separate groups of instrumentalists or singers -- physically separated by acoustical barriers to keep the sounds from mixing -- with precise control of each track. Then when the sound is rerecorded for two-track stereo discs, the eight tracks are channeled in a great variety of ways onto the stereo tape. I heard a recording of a combination of Humoresque and Old Folks at Home from a Marty Manning album which made me feel as though I were listening to musicians who fleetly raced, on silent gym shoes, from one side of the room to the other every eight measures. And I thought how far stereo has traveled (I hesitate to say progressed, though many people think the word appropriate) from its original purpose.

In the field of legitimate music, where novelty is not the desired end product, the more traditional attitude toward the performer and his performance, though undergoing erosion, still persists. A few years ago Kirsten Flagstad, making a recording of Tristan and Isolde for HMV in London, decided not to try for the two high Cs in the first part of the second act. She got Elisabeth Schwarzkopf to come to the recording session and belt out the two notes just when Madame Flagstad and the orchestra got to them. Innocent record buyers couldn't tell the difference, but the story leaked out and there were cries of outrage throughout musical circles. The stunt has not been repeated since.

Yet the technical fixing of any soloist's performance -- even if there is no brand-switching involved -- might be cause for similar outrage. Some years ago a leading Metropolitan coloratura -- not Miss Pons, this time -- was recording that same perilous Mad Scene from Lucia, and simply couldn't sing a good high E flat at the end. Finally the director, suffering the pains of orchestral overtime, asked her to wind up her cadenza -- which, fortunately, occurs during an orchestral pause -- on a D, and this she did well enough. Musically it made no sense, but in the editing room the director snipped out the D, played it through a tape machine at slightly increased speed, and goosed it up into an E flat which he later described as "perfect in pitch, but a bit like a peanut whistle." The synthetic E flat was then blended and mixed with the orchestra's pickup chord, and the result was spliced to the first part of the aria. The record stands as a monument to artistic collaboration, of a sort.

Most performers view this kind of thing with disapproval, perhaps recognizing in it the potential loss of respect for all their kind, but the opportunity it offers to improve their performances is a temptation of such an order as frequently to overcome their qualms of conscience. "The attitude of a lot of artists has changed," Richard Mohr, a music director of RCA Victor says. "Their personal standards, probably without their knowing it, are lower. And young artists, whose reputations aren't yet secure, want to record perfectly ordinary numbers in short snatches, working over each bit before going on. It's a sign of Van Cliburn's integrity that he can let a mistake go by, preferring the validity of the real performance to 'just perfection.'"

Mohr tries to keep synthetic perfectionists in line, but some other record producers have collaborated with their artists to create performances the latter cannot match in a live concert. Most lovers of legitimate music are not satisfied with this make-believe world, and are aggrieved when they find themselves technologically tricked. At the same time, the tape era has accustomed them to a glossy perfection of performance, and to definitive and unchallengeable interpretations. "People are growing somewhat incapable of enjoying a living performance with all its imperfections, its uncertainties and its excitement," Howard Taubman, longtime music critic of The New York Times, and now its drama critic, has observed.

Not only may tape be destroying the excitement of the living performance, but it may well be the villain that does away with the living artist. That, at least, is the specter that haunts musicians when they hear of concerts of electronic music, that recent "art form," impossible before tape, which dispenses with some or all performing artists, and substitutes on the concert stage -- sometimes with, sometimes without, orchestral accompaniment -- the reel of acetate ribbon.

Electronic music is made up of sounds -- instrumental, man-made, natural or generated by oscillators -- first taperecorded, then electronically blended, mixed and otherwise transmogrified, and finally assembled by the concocter into a whole which the composer claims has meaning, although the results frequently defeat many listeners' efforts to comprehend it. Still, an increasing number of people all over the world have been attracted by the superhuman possibilities of this new music since its debut in 1948 in Paris, in the form of a radio concert of railroad noises and the like.

Early in the development of this novel art form, its devotees were often absorbed in the effort to use ordinary noises in extraordinary ways, or to make everyday sounds by heroic scientific methods. Engineers at the RCA Acoustical Laboratory, for instance, labored nobly over their "synthesizer" -- a science-fiction agglomeration of electronic components -- and succeeded in manufacturing a set of tremulous imitations of the sounds made by violins, oboes and other instruments, all of which they put onto a record. The Sounds and Music of the RCA Electronic Music Synthesizer. The pièce de résistance is a few lines of Tennyson's poem Sweet and Low, in which the spoken sounds were all made up out of whole cloth, if I may use so inappropriate an image. The "voice" -- one has to strain to recognize it as such -- has a weird unearthly sound, not so much ghostly as cadaverous: it sounds as though a corpse, not too well preserved, were straining to make its decaying throat and lips function once more. Poe would have loved it.

More recently, the trend in electronic and tape art has been not to duplicate old sounds, but to construct and use new ones. Here in the United States, perhaps the most substantial contributions are being made by the Electronic Music Center at Columbia, a joint project of Princeton and Columbia universities, with three elaborately equipped studios under the direction of four respectable professors of music (Otto Luening and Vladimir Ussachevsky of Columbia, and Milton Babbitt and Roger Sessions of Princeton). In the spring of 1961, the public was treated to the first concert made up entirely of Center-produced works. On the stage of McMillin Theater, at Columbia University, six loudspeakers with colored lights playing on them, plus 13 others more modestly stationed around the auditorium, offered the audience samples of the kind of music that one composer defends by saying, "Noise and sound [for 'sound' read 'music'] have equal rights ... The value 'zero' of the noise scale is sound."

In this field, Luening and Ussachevsky might be considered conservatives. "We do not turn away from fundamental musical structure or tradition," Luening explains, "but we do try to draw upon new dimensions of sound so as to enrich the instrumental medium." On the other hand, Edgar Varèse, who at 76 is considered the pioneer of the musical concepts that led to electronic music, has a more revolutionary view. "Even as a boy," he says, "I felt stifled by tradition. I was looking for a bomb that would blow open the musical world and let in sounds -- all sounds." But radicals and conservatives within electronic music use methods that, by orthodox musical standards, are bizarre and wildly uncomposerish. I watched Ussachevsky work one day, surrounded by an arcane jumble of recorders, wires, speakers, generators, control panels, filters and boxes of sounds with such intriguing labels as "Auto Races," "Exp. with African Metal Bars," "Abscess" and "Mysterious." Assisted by a music student and a tape engineer, he labored for three hours to create one phrase nine notes long, for a new composition, passing a few tones through filters and echo chambers, cutting them to proper length, snip by snip, splicing them together, and then adding overall reverberation. At the end of the third hour, he had achieved what sounded like a series of notes on Parisian taxi horns: "Deedle-ump, de-deedle-deedle-oomp!... mp!" it went. As I left, Dr. Ussachevsky was trying to eliminate that final ". . . mp!" which, apparently, was an undesired echo.

Varése's most celebrated piece of this genre, his Poéme Electronique, demanded similar labors. This composition was played at the Brussels Fair in 1958, in an auditorium designed by Le Corbusier, through 425 loudspeakers arranged all around the audience. Of its eight minutes, more than 90 percent is laboratory-made sound. "We made hundreds -- no, thousands -- of splices," Varése says with great pride. "I worked half a year on it." The result of his Herculean snipping and pasting is hard to describe in words; in part, it goes something like this: "Gongs. A kind of shriek. Pip-pip-pip. Squeaks. Grunting. Faint echoes, far away. Squiggles, giggles, twittering and gongs again. Scrunch! Boom! Loud hum. Feedback type of squeal. Shrieking wind, rising and exploding. Rocket blast-off..." And so on.

Clearly, magnetic tape has been the bomb M. Varése sought. But its explosion has done more than let in new sounds; it has made noticeable cracks in the old bulwarks of veracity and reality. Whether this will in the end prove to have been good for art and mankind is not yet certain. The enthusiasts feel that the cutting and the splicing and the rearranging of imperfect nature are the rudimentary tools with which new, previously unimagined art forms, free from the limitations of the past, will be created. Other people less cheerfully feel that the whole thing is one more symptom of a mass drift from the daylight of the rational into the night of the absurd. The reel of tape goes round and round, and where it will stop nobody knows.