I began working with two collaborators, Catherine lkam, a French video and installation artist, and Bill Raymond, a New York-based actor and director associated with the Mabou Mines theater company. Together, during the summer of 1985, we shaped the first version of the libretto for VALIS, which by that time had become a full-fledged opera and had been commissioned to celebrate the tenth anniversary of the Pompidou Center. Most of the music for VALIS was composed between March 1986 and May 1987, with research and development of various computer music techniques for the opera beginning at about the same time, simultaneously at IRCAM and the MIT Media Lab. The final production of all electronic music materials, both pre-recorded and live, occurred during the summer of 1987, and rehearsals for the Paris production were held between mid-October and the end of November. VALIS received its premiere performances, in French, at the Pompidou Center from December 1-7, 1987. For this production, we designed an elaborate theatrical installation that filled the enormous entrance hall, or “Forum”, of the Pompidou, allowing for normal proscenium seating as well as standing room viewing from side mezza- nine balconies, thereby making the entire event more festive and public than a nor- mal operatic presentation. In addition, VALIS remained in place as an installa- tion, allowing the public to circulate amidst the primary sounds and images from the opera, through mid-February 1988. For this first production of VALIS, Catherine Ikam designed the sets (including a labyrinth handcrafted from Carrera marble) and visual imagery (using a computer-controlled video wall and two video towers for image projection), and Jean-Louis Martinelli directed.

After the Paris performances, I made several improvements to the score (various cuts, reworking of certain electronic material, some modified orchestration, etc.) as well as significant changes to the libretto, which | readapted into English. The opera was rehearsed and “ecorded at MIT’s Experimental Media Facility (or “Cube”) in February 1988, with post-production taking place at Hip Pocket Studios in New York City during March and April. The present recording represents the definitive and complete version of VALIS.

In March 1974, Philip K. Dick had his strange “pink light experience”. Dick claims that suddenly a pink beam was fired at his head and eyes, resulting in phosphene images like those left after a flashbulb goes off in one’s face [Rickman 1984]. For an entire day, the pink light brought with it effects usually associated with mystical revelation: time seemed to cease to exist, superimposing ancient Rome on Southern California; he became aware of specific knowledge, such as words and sentences from Koine Greek, and the fact that his young son was gravely ill and suffering from an undiag- nosed congenital defect (which turned out to be true, allowing a doctor to save Christopher’s life); he received glimpses of what he interpreted as the Apocalypse; he had many specific visions, from Soviet satellites, to three- eyed future humans, to wildly changing colors and patterns. This experience became Dick's obsession for the rest of the last eight years of his life, as he tried to explain to himself what had actually happened, and to make sense out of the bits of “divine” know- ledge which had been sent to him. He spent years compiling a theory of the universe based on his experience which he called his “Exegesis”, and which is reported to contain over two million words of handwritten notes. In addition. his experience influenced his last three novels, The Divine Invasion, The Transmigration of Timothy Archer, and especially VALIS, which is a semi-autobiographical account of the pink light experience itself, and of Dick’s reaction to it.

To transpose his personal trials into exploratory fiction, Dick uses the con- vention of dividing himself (as the main character) in two. This idea of mirror division is found on every level of the novel. InVALIS, Phil Dick the science fiction writer observes and comments on Horselover Fat, the part of himself who has supposedly experienced the mystical revelation. The book itself is divided into two symmetrical halves, the first being a realistic portrait of his life after the 1974 experience, and the second a fantasy portrayal of what this truth might be. Even the characters find mirror reflections in the two halves, such as Sophia, who can be seen as a kind of idealized reincarnation of the lost Gloria. On many levels VALIS is a novel of synthesis and unification, of bringing together parts which have been separated. This extends from Dick's attempt to bring his “realistic” and “science fiction” writing into a single context, to the book’s theme itself, which suggests a single force behind life’s seemingly fragmented percep- tions and events.

Turning VALIS into an opera was a challenging task. What seemed most important was to preserve the intellectu- al richness of the text, while clarifying as much as possible its sense of psy- chological and emotional drama. | believe, as Peter Conrad has written [Conrad 1985], that opera is more closely connected to novels (which con- vey so effectively what people are think- ing or feeling) than to the theater (where such internal machinations must be demonstrated through external action). Musical drama can augment the meaning of words, and, more importantly can convey the internal reflection that takes place between those words. | have placed Horselover Fat at the very center of the opera, and have concentrated on his personal search for truth. Certain plot elements, and various wonderful charac- ters from the book (such as Kevin, the hilarious skeptic) were not included in the libretto so that attention could be focused on Fat.

I have been very conscious of the idea of progression and transformation of events and states over time inVALIS. This has always been my greatest con- cern in imagining musical forms, and Dick’s diverse approach to such ques- tions in his book proved fertile ground. In the first half of the opera, | have tried to_ preserve the book’s atmosphere of chattt- ness, digression, fragmentation, and sur prise. Here, the spotlight is so heavily directed on Fat that all characters and action almost appear to result from a spilling out of his overabundant jmagina- tion. In the second half, however, the an observer, coming in contact with the Lamptons, Mini, and Sophia, who all seem at least as real as he is.

Fat’s psychological transformation is the central concern of VALIS, which takes its most obvious form in the split between Fat and Phil. The two roles are so closely connected that in the novel we learn that the name Horselover Fat is an English translation of Philip K. Dick (philip in Greek meaning “one who loves horses”, and dick in German meaning “fat”). Fat and Phil are played by the same actor/singer, but the pre- sentation of each part is highly contrast- ed, both in the actual musical material and in the nuance that the performer is asked to add to the interpretation. For the present recording, Patrick Mason worked very hard to give a totally differ- ent vocal color to Fat and Phil, even insisting on recording the two “charac- ters” on separate days. For Fat, he placed his voice quite high, emphasiz- ing the fluid, and often comic, fluctua- tiane of normal speech. For Phil, he placed the voice as low as possible without sounding unnatural, and kept pitch fluctuations to a minimum. His remarkable performance emphasizes Fat's energetic nuttiness and touching humanity, and Phil's pontifical, overly rational explanations.

As for the other characters, the attempt has been to give them a sense of clarity and true-to-lifeness with the greatest possible economy. Gloria rep- resents the only real woman and friend in Fat’s life, touchingly emphasizing Fat’s isolation since the two don’t even seem that close to each other. Gloria's vulnerability and emotional fragility are conveyed through her very few spoken phrases, her sighs and whispers heard here and there as memories, and her gentle melodies, hummed to herself, but never more than a step away from Cries of desperation. Dr. Stone’s wide, consonant intervals and deep, melis- matic ornaments emphasize his com- manding and reassuring presence. Fat's isolation is deepened when Dr. Stone abdicates his responsibility, transmitting his “authority” to Fat him- Self in a bitterly ironic turn. The Lamptons, through their harsh and repetitive rhythms and open, consonant harmonies, convey the arrogant and menacing sureness that | have often associated with religious cultists. Mini, in his silence, focuses attention on the magic of art and technology, his only lan- guage, while attracting suspicion about his motives. Sophia is an angel, convey- ing the opera’s fullest message in pure song. Her purity and simplicity only increase Fat’s desire that she be real. It is an intentional and painful paradox that Fat’s best human friend, Gloria, has so little to say, while Sophia, who has all the comforting answers, turns out to be something other than human, finally dis- tant and unattainable.

While I attach great importance to the opera's text itself, and have tried at every turn to make it as comprehensible and telling as possible my main goal has been to express psychological states, intellectual ideas, and emotional reactions through music [Machover 1986]. In the opera, there are basically three kinds of musical sections, each of which serves a different function. The first establishes its own internal charac-ter and musical shape, and is relatively independent from other musical sections. Each aria is like this, as are most of the elements which are used to build the work's larger contrasting units (such as those of "Fat's Dream"). The second type of section establishes a strong sense of movement or development, of "becom-ing" and never standing still. These sec-tions generally function as dramatic cli-maxes, creating a great sense of antici-pation ("Finale II") or instability ("Finale I"). The third type of section is the opposite, and establishes the feeling that time and forward direction have stopped. Such sections (i.e. "Loneliness Transition" and "Mini's Solo") give the listener time to reflect on what has gone before, and, by putting familiar musical elements into new contexts, to prepare for what is to come. I have used a varied palette to establish these contrasts. In fact. the great quantity of different "types" of music in VALIS is one of the first things that any listener will notice. As Dick's book suggests, the styles of musical expression range from rock to romartic to medieval to futuristic. I have always believed that the impact of synthesis is most powerful when the contrasts between the previously separate parts is maximized. The feeling of continuity in VALIS is conveyed mostly through the overall musical form. As is true on all levels in the opera (musical, textual, and dramat-ic), the formal development of Parts I and II are quite different from each other.

In Part I, each section is present-ed and juxtaposed within the first fifteen minutes of the opera: the pedal tones and spectral harmony that accompanies the narrator Gloria's flowing melody with electronic resonances; the clipped rhythmic theorizing of Fat's "Sacrament'; the rising and falling fifths of the "Overture". These elements are spun into larger and larger musical structures, such as the variation form of "Fat's Dream". or the piling up of the "Finale". The process is one of increasing accu-mulation. fragmentation and juxtaposi-tion, reaching its culminating point of complexity with Gloria's suicide. In Part II, the process is reversed. The rhyth-mic and textural complexities of the "VALIS Song" set the stage for a contin-ual release of tension, reaching the point of greatest stillness at the end of "Sophia's Aria". The careful listener will find that melodic material from one section of the opera is often transposed as accompa-niment material to another. This is used to establish thematic links between often very contrasting sections. For instance, many different themes are to be found in the "Overture", spun into a single melody. Fragments from the "Gesegnet Song" are spread throughout Part I. The "VALIS Song" melody returns in many guises throughout Part II. One of its more interesting reincarnations is as the rapid, stretto-like accompaniment to Fat's melody in the second half of "Finale II". There are many musical paradoxes in VALIS. As in Dick's text, things are usually not what they seem to be. At moments of great complexity, there is often a single element that rises to the fore, focusing one's attention and build-ing in intensity (as in the "Finale" to Part I). When things seem to be simplest and most understandable, independent parts often separate, as if the background is coming alive to devour the objects posed in front of it (which is most disturbingly audible in the 'Finale" to Part II).

My work in computer music has always involved developing concepts and techniques in response to particular musical ideas [Machover 1984, 1985, 1986, 1987]. Because of the scope and special demands of VALIS, research led to a great number of new applica-tions, many of which were implemented for the live performances in Paris and for this recording. When first starting work on VALIS, I decided to make the opera almost entirely electronic, freeing myself to imagine and design a new "opera orchestra". To do this, I scored the piece for only two musicians, one key-board player and one percussionist, connected to a vast array of specially designed music technology. Much of this music is performed live, and special extensions to the instruments let the players delicately control and shape a vast sound world. For the live perfor-mances, a wide variety of commercial MIDI equipment was used (keyboard and percussion controllers, various syn- thesizers and samplers. effects boxes, Macintosh I1 computers, etc.), as well as specially designed computer music systems, such as IRCAM'’s 4X machine. Other electronic material was devel- oped during long months of studio work, then pre-recorded and played from 24- track digital tape at specific moments during the performance. The six singers are always amplified.

In VALIS, different sound materials are used to characterize developments in the story. For instance, the sound worlds of the first and second halves are distinctly contrasted. During the first half (when they are positioned in two recessed “orchestra pits”, visible to the public but off-stage), the instrumental- ists start out playing acoustic instru- ments, a normal acoustic piano and percussion setup. As Horselover Fat gets further and further involved with his Exegesis, and moves closer to mental breakdown, these acoustic instruments are shadowed by the computer, which releases more and more perceptible electronic resonances. This Pandora's F Box of computer sound is totally : unleashed by the time of Gloria's suicide at the end of Part |. When the Lamptons arrive on stage at the start of Part Il, the two musicians join them, and play totally electronic instruments, reflecting the action’s shift to the fantasy realm. Mini's appearance brings this transformation to its extreme, as he sculpts an abstract music performing on an invisible instru- ment. The same ideas are applied to the treatment of voices in the opera. Part 1 centers around spoken text, especially by Fat and Phil. These words are fragment- ed, echoed, and gradually computer- transformed as Fat’s Exegesis begins to take on a life of its own during the Dr. Stone scene. By Part Il, speech gives way to song, emphasizing the fact that the opera has moved away from real life to an artificial, fantasy world.

Since I began working with computer music in 1977, | have been interested in developing techniques to allow the medium to be adapted to the demands 0 live performance. This has taken various forms, sometimes using fragments of pre-recorded tape and amplified acoustic instruments (i.e.Soft Morning , City!, Light, Spectres Parisiens, Famine), others using electronic modification of acoustic instruments (such as Electric Etudes), and yet others employing the computer as a solo concert instru nent (Fusione Fugace). Because of its length and diverse musical demands, VALIS necessitated new concepts in this domain of real-time musical perfor- mance. This was especially true since theatrical work requires maximum flexi- bility of performance nuance, minimizing the effectiveness of rigid pre-prepared tape accompaniment.

I developed, with Joseph Chung, the concept of “Hyperinstruments” at MIT’s Media Laboratory. We have tried to program computers to adapt intelligently to a musician’s live performance. The idea is to use sophisticated machines to give the performer far more musical con- trol than is traditionally possible. To do this, Chung wrote a musical environment in LISP that runs on Macintosh II com- puters. These computer programs ana- lyze a musician's gestures and performance (i.e. notes played, loudness, aftertouch, etc.), and react immediately, sending MIDI data out to an array of synthesizers, samplers, and sound transformation modules. Many different modules were programmed to adapt to the needs of various sections of VALIS. A score follower was developed to allow the computer to react flexibly to piano performance. A specially designed Bésendorfer Imperial Grand Piano was used, sending MIDI data to the Macintosh II, as well as recording the pianist’s performance data. A hier- archically organized model of the musi- cal score was given to the program, which then followed (using pattern matching algorithms) the live piano with a tolerance for errors and tempo fluctuation. Our program then deter- mined a specific musical reaction from the machine depending on what was played by the pianist. This ranged from electronic timbres shadowing piano notes, to sequences triggered automati- cally at a specific cue. “Dr. Stone’s Aria”, for instance, is performed on acoustic piano with live score follower, which adds an unreal ambience to the psychiatrist's sermon.

A more computer-automated approach to “Hyperinstruments” can be heard during the Lampton scene. Here the goal was to provide the two rock musicians with instruments that magni- fy virtuosity to a level of very delicate complexity. A performance system was built up around the electronic percus- sion and keyboard controllers. Among the many available functions, one of the most interesting was an automatic arpeggiator. On both percussion and keyboard, a set of rhythmic patterns or “templates” was stored in the computer. Pitches (for the keyboard) or timbres (for percussion) were added live by the performer, and fitted by algorithm to the pre-existing pattern, creating ever- changing, complex phrases. In the case of the keyboard system, for instance, each voice could be carefully articulated in these patterns by controlling its dynamic, and increased pressure (or "aftertouch”) effectuated delicate timbral changes. In addition, a score follower allowed each change of harmony to be automatically associated with a distinct rhythmic figuration. This process can be clearly heard in the accompanying lines to the “Suffering Song”, as well as in much of the “Lampton Scene” and “Finale” to Part Il.

Another aspect of technological development for VALIS is the analysis and transformation of the human voice. Some of the more sophisticated of these programs were developed on a non-real- time system, a VAX 780 computer with an array processor to speed up calcula- tion. | used Phase Vocoder programs with additions designed and implemented with Arnaud Petit and Martin Hermann, and Robert Rowe for graphic interfaces. For this technique, speech (which works better in this context than sung material) is analyzed by the computer. The Phase Vocoder produces data files that divide the original speech up into very fine fre- quency bands (about one band for every 7 Hertz) with time-variant amplitude infor- mation in each band. These files can then be manipulated in extraordinary | ways: speech can be stretched or com pressed in time without affecting pitch or timbre, timbral transformations can be performed without obscuring the intelligi- bility of the speech, etc. One of the more interesting transformations that we designed involved carving up words and sentences into a sort of spectral jigsaw puzzle. About twenty separate soundfiles were created from a single speech frag- ment, each containing an irregular grouping of frequency bands with no duplications between the files. Each file by itself had a very ambiguous sound, not really comprehensible as speech, but not totally electronic either. Only by mix- ing each of these files back together could the speech be understood. We then wrote programs to gradually modify the speed of each of the individual files, so that the entire group could move in and out of phase very gradually. The result was to produce very subtle spec: _ tral shifting, with words starting out intelli- gibly, then transforming into shimmering timbre, and finally snapping back into total “normalcy” or fusion on another specific word. Many of these Phase Vocoder transformations can be heard in Fat's two “Exegesis” fragments.

A third category of music technology developed for VALIS involves the live transformation of vocal or instrumental sound. For the performances in Paris we used IRCAM's 4X machine (a pow- erful, open-ended digital synthesizer). with programs developed by Miller Puckette, Cort Lippe, and Robert Rowe. Since then, these techniques have been translated onto more inex- pensive and transportable equipment. in preparation for future touring produc- tions of the opera.

To provide maximum flexibility for treating incoming music. we developed realtime microediting of sounds. A given sound (most often Fat's speech) was sent to the 4X, either ina continu- ous stream or as a soundfile of speci- fied duration. On the screen, a display showed the waveform “live” as it filled up the synthesizers memory. Under the waveform display was a graphic image of a keyboard, with each key clearly vis- ible. Our program in the 4X enabled each key to play back the incoming sound starting exactly at the point that the key lined up with the waveform, an impossibility with any currently available commercial sampler. In this way, an electronic performer (seated in the “orchestra pit”) could fragment, prolong, or echo any sound with great precision and on-the-spot flexibility. This tech- nique can be heard throughout Part |, Spey curing “Exegesis II" and the , n the “VALIS Song‘ where it was used for some vocal splintering and virtuosic repetition.

Another procedure for transforming incoming sounds involved spectral mut). plication . For this technique, each incoming sound was copied as many as 40 times. Pre-programmed templates allowed for each of these copies to be transposed, and amplitude envelopes were imposed as well. The templates ranged from slight frequency skew, pro- ducing clustering and chorusing, to wide- ly spaced “spectral chords", creating the effect of additive synthesis produced with real sounds. When the recorded sounds were transposed far enough to produce significant distortion, the programs could substitute similar prerecorded sounds in an appropriate register, or sounds with a different but complementary timbre. This procedure produced sound structures of exceptional richness and complexity. These can be heard in various sections of the opera, such as those using cello transformation (notably the accompanr ment to the “Parsifal Transition’) and dense computer-like inharmonic sounds (such as “Mini's Solo” or the transiton to the "Slippery Song")

SYNOPSIS

VALIS begins with a sudden explosion of sound and flashing pink light combined with rapidly crosscut video images. The Overture, which weaves together many different musical themes and timbres from the entire opera, begins in stillness, and finds Horselover Fat pierced through the head by a pink laser beam. As the music builds in majesty, visual images from Fat's eight hour revelation (“lurid phosphene activity in eighty colors, three- eyed people in glass bubbles and elec- tronic gear, living plasmatic energy, ancient Rome superimposed with Southern California”) fade in and out almost imperceptibly. Part | of the opera is dominated by Fat’s search for the truth behind his pink light encounter.

The voice of Philip K. Dick speaks the First Narrative over sustained pedal tones and chords. He sets the scene, telling of Horselover Fat's mystical experi- ence which is perhaps just a nervous breakdown, and of the obsessive journal that Fat has been writing (his “Exegesis’) in an attempt to explain the universe.

Horselover Fat launches into a rhythmic rap-style presentation of part of his theory, his Sacrament, stating that people are “like memory coils in a computer-like thinking system”, and that the universe itself “is information’. Piano and percussion trills then trans- form into dreamy beach sounds and video images, accompanied by a melis- matic soprano melody. In the Beach Scene, Fat tries to convince his friend Gloria, who is “tired of doing what everybody else wants”, not to commit suicide. As our narrator pointedly observes, Fat blows it, and ends up selfishly making Gloria feel guilty. This failure will weigh heavily on him.

Fat’s Dream is an elaborate set of musical variations (as is all of Part !) based on two intertwined themes, one the unfolding spectral chords first heard in the opening narrative, and the other which makes reference to several Important themes from Wagner's Parsifal. Horselover Fat recounts a recurring dream of an idyllic spot in morthern California, complete with a lovely and tender wife he has never seen. He contrasts this to his own isolat. ed, barren life in “plastic” Southern California, and comes to the conclusion that he has somehow absorbed memo. ries from another person, asking, “Do | become my father during my sleep?”, This leads to an exchange between Fat and the narrator, who both obsessively repeat another of the opera's main ques. tions, “How many worlds do we exist in simultaneously?”. The image of Fat's face is fractured and echoed on multiple video screens until all snaps. The stage is flooded with bright white light, and Fat finds himself in a strait-jacket.

Fat remains immobile during the Loneliness Transition, a sombre music of passage and reflection. He has attempted suicide, and finds himself wrapped in a strait-jacket; the images suggest a mental hospital.

Booming piano chords mark the entrance of Dr.Stone, who attempts to cure Fat by helping him believe in his own vision. After hearing him prescribe vari: OUS Organic remedies, sing a melodious aria based on the Tao Te Ching , and tell Fat that he himself is “the authority”, one wonders if the Doctor isn't equally nutty. The scene is interrupted by ethereal pas- sages from Fat’s exegesis (spoken in French) which are accompanied by visual images from his theory, as well as by brief remembrances of the scene with Gloria and premonitions of her suicide. Dr.Stone releases Fat from his strait-jacket.

These sounds and images accumulate as if flowing forth uncontrollably from Fat’s mind. Swelling string sonorities become louder and louder, leading into the Gesegnet Song, the only time that Part I's variation theme is heard in its entirety. This heroic tenor melody rein- forces the link to Parsifal, and this con- nection is further extended in the Parsifal Transition. The narrator’s face appears on screen for the first time, revealing that he is the same person as our protagonist. Against a transformed and elongated ver- sion of the opening strains of Wagner's opera, Phil compares Fat's search for truth to Parsifal’s. As he chastises Fat for his ineffectuality in not saving Gloria, Phil inadvertently demonstrates his own pain and lack of objectivity; the search for the truth is really his own.

All the diverse elements from Part 1 densely collide in its Finale. Fat and Phil face each other for the first time (one live, one on video), affirming the necessity of recapturing and unlocking the mystery of the pink light experience: sounds, text and images from the exe- gesis abound; Gloria returns and sings an ever more mournful melody as she climbs stairs higher and higher. Her voice soars above the increasingly chaotic texture and at the climactic moment she screams and throws her- self into thin air. A short, chilling black- out leads without a break into . . .

Part II

With a crash of drum and cymbals, Fat finds himself in front of a large screen watching a short, intense science fiction film set to the VALIS Song (“I want to see you man...”). The film has been made by two rock musicians, Eric and Linda Lampton, who have concocted a wildly mysterious narrative combining most of Fat’s exegesis images from Part | with many technologically-orient- ed icons not referred to before, and hints at a powerful force called “VALIS’. Fat understands that the Lamptons are aware of the same strange facts as he is, and that they’ve taken the trouble to make a film to transmit the secret mes- Sage to others. This reinforcement con- vinces him that he is not crazy, and gives him the courage to drop his theo- rizing and set out in search of the force behind the film. With one Stroke, the oe prpence of the opera changes: disorde. | 'S Concerned with mental » ragmentation, and human pain, Part Il moves boldly into the realm of magic and fantasy.

At the end of the film, while the word “VALIS" is sung over and over, Eric and Linda Lampton appear on stage, seeming to materialize from within the film itself, to sing the second part of theVALIS Song (“VALIS, does it come from the stars?...Is VALIS a god, a Child, 4 satellite...does it destroy?”).

During the Lampton Scene which follows, Horselover Fat faces the two rock musicians, who quiz him about how much of the film’s secret message he has really understood. Although the Lamptons will not reveal the true nature of VALIS, and only divulge the meaning of its acronym (“Vast Active Living Intelligent System”), they do tell Fat that his search is finally on the right path, that his pain will soon be over. Eric, Linda and Fat join together for the gently reassuring Suffering Song.

This leads directly into Mini’s Sole. Mini, a half-crippled inventor and comput er music composer arrives onstage slowly from the conductor's chair that hé has occupied until this moment. Mini has composed the Lampton’s music, and Is a master at transmitting subliminal messages about “VALIS” through his art. While the Lamptons have performed with magical electronic virtuosity, Mini (who neither speaks nor sings) conjures pure sound out of thin air, bringing to life the entire stage area in a unity of flowing lights and geometric images. Paradoxi- cally. Mini's music, the most abstractly electronic of the entire opera, is in fact created from bits of the human voice, deconstructed so as to be unrec- ognizable at first. During the course of his short performance, Mini moulds these independent sound particles into a single, pure female singing voice.

This is the voice of Sophia, whose scene marks Fat’s final encounter. Sophia is a beautiful young girl who emanates a strange and unreal aura; we are never sure if she is a normal girl, an angel, or a hologram. Sophia notices that Phil/Fat's double personality has disap- peared (“Il can see only one person...”). She tells him that his long search has ended and that the goal of his life has been reached. In an aria of shimmering simplicity and reassurance, Sophia speaks the words of “VALIS’”, telling Phil/Fat that the only true god 15 him- self, that the time of trials is not yet over but that he must be patient. Whereas before he was alone, now he has a friend who “never sickens, nor fails, nor dies”.

At the point of greatest stiliness, the Finale begins. In this majestic pas- sacaglia, which transposes the opera's theme of obsession from a textual to a purely musical plane, Phil/Fat, Sophia, and the Lamptons all join in with contra- dictory counterpoint. Phil/Fat sings that he’s “not afraid, my search is over...VALIS is real”; Sophia adds that he should “not fear, | will not fail you”: the Lamptons keep posing questions about VALIS ending with “but VALIS, is it real?”; and Mini conducts. Each singer develops dramatically in his or her own way: the Lamptons become more and more taunting and even sadistic, as if receiving pleasure trom suggesting to Phil/Fat that VALIS might be an illusion; Sophia becomes ever more ethereal, as if she is too pure and comforting to-actually exist: and Fat repeats his words over and over, at first with a sincere reassurance, then with the determination of someone trying to convince himself of something, and _ finally with a sort of violent panic as his dream begins to vanish. During this section, most musical development takes place in the background accom- paniment which adds layer after layer of motivic and timbral embellishment, building to a point of enormous intensity. All the while, Mini conducts, at first normally with his hands, but gradually using rays of pink laser light. As the music builds towards its climax, Mini deflects these laser beams (or is “VALIS" sending laser beams through him?) to bathe Sophia in light, first gen- tly —as if performing a laboratory experi- ment, and then in more agitated fashion —as if the experiment is bursting out of ever mone es true nature becomes é ambiguous. As Phil/Fat cries very og is REAL for the final time, a and chee, \ ser beam Strikes Sophia light, brin n Fat ising een dca” ging at's vision full circle. All disappear with her in a flash, leaving Phil alone on Stage.

Coda