An Introduction to Extended Vocal Techniques: Some Compositional Aspects and Performance Problems1



Deborah Kavasch



A vocal sonic vocabulary is developing in Western art music which includes and extends beyond traditional Western art music phonation, the basic voice production associated with opera and recital singing. Contemporary vocal writing includes sounds which previously were seldom or never heard in musical contexts. Some of these sounds or techniques occur in other musical cultures while others have arisen through the research and experimentation undertaken by certain contemporary vocalists.


This paper discusses selected "extended vocal techniques" in the context of specific compositional uses and related performance problems and relates specifically to the vocabulary of vocal techniques and sounds developed by the Extended Vocal Techniques Ensemble (EVTE) of San Diego, California. The descriptive, subjective terminology developed by EVTE for its vocabulary of techniques is used in this paper since precise definitions have not yet been developed for most of the techniques. The ensemble recorded a Lexicon of Extended Vocal Techniques in 1974 and the expanded index of this Lexicon2, prepared by Linda Vickerman of EVTE, has aided in explaining some of the techniques discussed in this paper.


The five techniques chosen for discussion include "reinforced harmonics," " ululation,"  vocal fry," "chant," and ""complex multiphonics." They represent those techniques which: 1) can be learned most quickly; 2) have been used most extensively in compositions written for EVTE; and 3) are basic techniques which can be applied to- many sounds and/or from which distinctive variations can be produced.


Although improvisation can provide a basic musical context for new sounds and, indeed, was the first performance vehicle for these techniques used by EVTE, compositional demands offer the vocalist new performance dimensions. As EVTE members became more proficient with these sounds, composers wrote works specifically for the ensemble. Musical examples taken largely from these compositions are cited and discussed in the paper and a tape recording of these examples accompanies the paper. By presenting practical instances of certain extended vocal techniques, the paper addresses interested composers, singers and others, and illustrates and suggests some effective uses, limitations, and possibilities of these techniques.


Harmonics present in a sung tone can be individually reinforced or amplified and perceived as discrete pitches (sounding like whistles) as tongue and/or lip action changes the shape of the vocal tract. The sung fundamental is produced using Western art music phonation, usually without vibrato so that the harmonics nasalization, which tends to filter out the fundamental and focuses greater attention on the harmonics.


The examples discussed below demonstrate harmonics reinforced over either a single pitch or a continuously changing fundamental. One or more harmonics may be individually reinforced over a single pitch, both ad libitum and as a designated, specific harmonic. The vocalist's skill may even allow the composer to write a melody with the harmonics. A rapid movement through a series of harmonics will probably shift the listener's attention to timbral changes rather than recognition of specific pitches. A shimmering effect may result from rapid oscillation between two adjacent harmonics (see below).


Improvisation over a unison fundamental pitch represents one musical context for reinforced harmonics. Early improvisations done by the Extended Vocal Techniques Ensemble during rehearsal sessions were frequently structured in the following manner: A single pitch comfortable for both men and women was chosen, usually F# (184 Hz)3 or G (196 Hz) below middle C (261 Hz), and sustained without break by staggered breathing. An approximate duration was set, perhaps 5-10 minutes, during which an emphasis first on the lower harmonics was to gradually progress to the inclusion of higher harmonic reinforcement within a specified dynamic structure of perhaps soft to loud. Tape recordings of a number of these early sessions are filed in the archive of the Center for Music Experiment at the University of California, San Diego.


The first notated composition written for EVTE, The Owl and the Pussycat by Deborah Kavasch, includes an improvisatory section on a unison G (196 Hz) built on the last word of the text, " moon." The score (Figure 1) indicates that only those harmonics in a fifth or octave relationship to the fundamental are to be reinforced at first: others can be reinforced later. The instruction "Create a dancing texture, active" is made possible by including several techniques in combination with or as a variation of the basic individual harmonic reinforcement. The "dancing" characteristics arise principally from those techniques which rapidly change the pitches of the harmonics or which interrupt or alter the fundamental and harmonics. The term "glissandi" refers to ascending or descending sweeps through a series of harmonics. An oscillation between harmonics, which is subsequently referred to as " harmonic oscillation," results from a backward and forward movement of the tongue and is quite effective when produced rapidly between two harmonics. If the fundamental pitch is low, two high adjacent harmonics are easily oscillated: if the fundamental is high, two low adjacent harmonics respond well. Oscillation of the fundamental, similar to vibrato, changes the pitch of the fundamental at regular or varying speeds, causing the reinforced harmonic to change pitch at the same rate. The rapid, repeated note effect of ululation (see III. Ululation) rather evenly and quickly interrupts the fundamental, these pulsations help emphasize the harmonics as well. When another voice adds a fundamental pitch one octave higher than the original fundamental, it generates a new but closely related series of harmonics which increases the complexity of the texture'4 (Tape Example 1).





.                                                          Figure 1: Reinforced harmonics in The Owl and the Pussycat by Deborah Kavasch (Tape Example 1)



Another use of reinforced harmonics over a drone appears in Sweet Talk by Deborah Kavasch.  In one section the harmonics are specifically those which result from a slow glide through the vowel sounds of the syllable "beau" (as in "beautiful" ), i.e. a slow glide between [ i ] and [ u ]5. The duration and dynamics of each entrance are notated, showing specific breathing points (rests) and definite, perceivable entrances (Tape Example 2). Although a drone results from the continual overlapping of entrances, the whole is not perceived as a smooth, unbroken fundamental as in the previous example.


One further example involving harmonics reinforced over a single pitch drone illustrates a slightly different usage involving different rates of change in harmonics generated from similar texts. In the opening section of Deborah Kavasch's Requiem. (Tape Example 3), the upper three taped voices sing one syllable per measure (B 247 Hz); the upper three live voices sing the same text on the same pitch at one syllable per two measures. Harmonics appropriate to each vowel are reinforced without vibrato. This emphasis of a particular harmonic determined by the specified vowel aids in tuning the unison and provides a type of countermelody to the drone, in this instance two countermelodies. Amplification with microphones for each vocalist aids in projecting the harmonics.


Nasalized, reinforced harmonics ad libitum on specific pitches provide a striking beginning to Edwin London's Psalm. of These Days II (Tape Example 4). The intelligibility of the word "Lord" varies due to the changing harmonics as well as its long duration (three measures). Although the initial unison D# (311 Hz) expands to a four-note chord spanning a minor seventh, the fundamental pitches together with their harmonics of the " r" in "Lord" are still close enough to create a rather dense texture. Compare this with a later example of reinforced harmonics on the same word, in which the fundamental pitches are spread over approximately a two and one-half octave range (Tape Example 5).


Both nasalized and non-nasalized harmonics are reinforced over slowly gliding fundamental pitches against a background of computer-generated tape sounds in Joji Yuasa's My Blue Sky in Southern California (Tape Example 6). The fundamental pitch is chosen at random by the vocalist and gradually ascends or descends as indicated in the graphic score. Unusual effects are achieved by pitch changes in both the fundamental and harmonics which may move at varying rates in similar or opposite directions to each other. In order to be easily heard in this very loud and dense section, the vocalists tend to choose and nasalize the higher harmonics (towards UP, creating a rather buzzy, piercing quality in the long pitch glides.


In another section of the same composition, the oscillation between several pairs of low harmonics of a rather high fundamental (approximately A 880 Hz) is heard against a sparse background of very soft clicks and other similar short, nonpitched sounds (Tape Example 7). The higher the fundamental, the more difficult it becomes to reinforce its highest harmonics. In this instance, the oscillations occur between various combinations of the first three or four harmonics (including the fundamental), which may acocunt for the pulsating whistle effect during part of the example. Oscillations between low harmonics (of a high or low fundamental) may also be more striking because of the greater intervallic distance between the pitches of the lower harmonics. For example, the oscillation between the (1) fundamental (or first harmonic) and second harmonic covers one octave; (2) second and third harmonics covers a fifth; (3) third and fourth harmonics covers a fourth; (4) fourth and fifth harmonics covers a major third, and so on, the distance always smaller.


At one point in Deborah Kavasch's Tintinnabulation, harmonic oscillations occur simultaneously in several voices. The fundamental pitches are each a half step apart (F 350 Hz, F# 370 Hz, and G 392 Hz) with the harmonic oscillations of each pitch determined by the same [ill] vowel alternation (Figure 2). Such a close pitch combination results in an overall pulsating or shimmering effect rather than the perception of individual harmonics (Tape Example 8).




Figure 2: Harmonic oscillations in Tintinnabulation by Deborah Kavasch



A vocalist's increased skill and accuracy in reinforcing harmonics allows the corn-poser to specify more particular uses. Simple melodies are possible, especially when using the lower three to three and one-half octaves of harmonics (first seven to twelve harmonics). Novices can quickly learn to reinforce the melody of the traditional bugle call, "Taps," which uses a one-octave span of harmonics, specifying the third, fourth, fifth, and sixth harmonics. The lower three octaves of harmonics are more favorable for writing melodies based on reinforced harmonics because it is easier to distinguish between the lower harmonics than between those which are less than the interval of a second or third apart. Generally, the lower the fundamental the more harmonics can be reinforced with clarity. If the fundamental is too low, however, the lowest harmonics become difficult or impossible to reinforce. Although this varies with the individual voice, a comfortable lower limit for a fundamental pitch within which the vocalist, male or female, can clearly reinforce the lowest harmonics can be set at approximately F 175 Hz (below middle C). Fundamental pitches lower than this arbitrary point are increasingly more favorable for the discriminatory reinforcement of higher harmonics, particularly those in the third and fourth octaves above the fundamental.


A melody formed by specified reinforced harmonics appears in the second "Hosanna" of the Kavasch Requiem. (Tape Example 9). Against a taped background of four voices singing for the most part in octave F#'s (185 and 370 Hz; note that a strong C# 1108.5 Hz is generated by the [a] vowel), the vocalists form melodic phrases using certain of the lowest seven harmonics of three F#'s (92.3, 185, and 370 Hz) and two C#s (138.5 and 277 Hz). The score shows regular noteheads (stems down) for the sung fundamentals and diamond-shaped noteheads (stems up) for the harmonics. For greater clarity and reliability, the melody of harmonics is always carried by at least two voices; all voices are amplified with microphones. The use of voices on tape singing a slowly paced text to a very simple harmonic structure provides pitch stability and avoids interference with the audibility of the harmonics produced by the live performers but helps to mask vowels resulting from reinforcing the specified harmonics of the live portion (which do not form intelligible words). The harmonics form a melody quite similar to that which is sung in the first "Hosanna" and should, by association, be more easily recognized or perceived by the listener than a new melody would be.


Reinforced harmonics is one of the few techniques which does not always, or even most often, require individual or general microphone amplification. Especially when reinforced over a drone, the harmonics are often perceived as non-directional, filling the entire space as though surrounding the listener. The fundamental, however, is usually directional and perceived as originating in one particular spot. Amplification is recommended when the harmonics must be heard above a very loud or dense texture, or for better projection of specific melodic passages.





The technique of ululation is perceived as a rapid, relatively even interruption of the basic sound. It is articulated by aspiration (puffs of air or "his") or glottal stops and can be applied to virtually any sound, voiced or unvoiced. Children often ululate a loud, nasal sound to imitate the firing of machine guns or the bleating of sheep.


Several forms of ululation are discussed below: (1) ululation of a single pitch or series of pitches; (2) nonpitched ululation, or an ululated whisper; (3) ululated glottal clicks, referred to as "glottal whisper;" and (4) cross-register ululations, i.e., a very rapid alternation of two pitches produced by a rather loud ululation in the area of a natural register break.


The opening section of The Owl and the Pussycat, preceding the narrator's first entrance, builds a gradually thickening texture of soft ululations. The score (Figure 3) indicates the progression of time in minutes and seconds and shows a graphic outline of approximate pitch levels and directions. Each box represents from top to bottom the high to low pitch range of the individual vocalist. The taped example (Tape Example 10) demonstrates the effect of several voices softly ululating an aspirated [u] in relatively low pitch ranges. The next example (Tape Example 11) includes ululation of some of the vowels and consonants of the word "pussycat." Appropriate vowels and consonants are deliberately aligned with similar ones in the narrator's text. Ululations can be used not only with short or long passages of a single vowel or to extend and color individual words but to articulate melodic phrases as well. The vocalists ululate several short melodic fragments set to the words, "0 lovely Pussy," "0 Pussy, my love," and " What a lovely Pussy you are." This 90-second section gradually expands in total pitch range and density, and ends with a sudden shift from soft to loud ululations. The taped example (Tape Example 12) is excerpted from the first part of the section.         


                                                 Figure 3: Ululations in The Owl and the Pussycat (Tape Example 10)


Ululation of an unvoiced whisper is used in the "Lacrymosa dies illa" section of Requiem. In this section the unvoiced sounds are divided into four categories: (1) a straight or uninterrupted whisper; (2) measured (sixteenth-note) pulsations or aspirations of a whisper; (3) ululations of a whisper, which are faster than the pulsations and result in an almost shivering sound; and (4) measured or unmeasured "glottal whisper," a term which refers to a rapid series and/or ululation of glottal clicks. The ululated whispers attempt to support the imagery suggested by the text (" A day of tears is that day" ) (Tape Example 13).


The latter part of Example 13 includes a soft ululation of two alternating pitches. This sound may be related to the production of a glottal whisper to which voice is added. Although it can be produced throughout most of the egressive singing range, it is usually referred to as a cross-register ululation. This type of ululation usually settles into the interval of a third and is most easily produced on the vowels [i] and [u]. Because it is not as reliable as the simple ululation or the louder cross-register ululation, it should be allowed a certain amount of preparation time. It is least tiring in the lower female range (approximately middle C to C 526 Hz). (As is true of certain variations of the EVTE's techniques, the soft cross-register ululation has been perfected by only one member of the ensemble and is probably less likely to be produced by a majority of vocalists.)


A more extended use of the soft cross-register ululation appears in Roger Reynold's A Merciful Coincidence. In Tape Example 14, the ululations cover a wide pitch range and appear as background material near the end of the piece. The designation "soft" in the term "soft cross-register ululation" refers more to the physical sensation of the vocalist in producing the pulsation, or interruption of the sound, than to the actual dynamic level. In this example the higher ululations become much louder but are still softer than a regular cross-register ululation at a comparable pitch level.


The term "cross-register ululation" refers to an ululation produced in the area of a natural register break6. It results in a rapid alternation of two pitches, which creates the illusion of two pitches ululated simultaneously. The intervallic distance between the two pitches varies with the individual vocalist but usually falls into one of two categories. The first category emphasizes narrow intervals, i.e., approximately a minor second to a fifth. The intervallic distance can be controlled so that either one specified interval or a continuous glissando from narrowest to widest interval (or vice versa) can be produced. The glissando appears to be most easily produced by the upper pitch, which moves towards or away from the stationary lower pitch.' This glissando of one of the pitches is best controlled in the register break around middle C. The second category emphasizes wide intervals, i.e., approximately a sixth to an octave or ninth. There does not seem to be the same degree of intervallic control possible in the wider cross-register ululation, which generally locks into one interval. This type usually occurs in either the male or female voice around the middle C register break area. Either type of cross-register ululation may occur in male or female voices but both types do not generally occur in one voice.


Cross-register ululation, as described above, requires more energy than simple ululation and is necessarily rather loud. It sounds particularly loud in the male voice since the first register break at which it can be produced includes pitches fairly high in the male chest voice range. For the same reason, cross-register ululations in the female voice around the second register break area are usually extremely loud. Ululations crossing into the whistle stop area, however, are often much softer. Those cross-register ululations which occur higher in the voice are usually more tiring than the lower ones and should be used with consideration.


Both simple and cross-register ululations enhance the text of the " quantus tremor..." ("how great a quaking..." ) section of Requiem. A total of eight voices produces the effect of many more, an effect due both to the pulsations and to the extra number of pitches generated by crossing registers, i.e., alternating two pitches (Tape Example 15). A later example (Tape Example 16, Figure 4) combines several other techniques with cross-register ululations which are indicated by a general pitch level and which end in a quick upward glissando.








               Figure 4: Cross-register ululations combined with other extended vocal techniques in Requiem by Deborah Kavasch (Tape Example 16)



Ululations form a major portion of Tintinnabulation. The last section emphasizes high cross-register ululations alone and in combination with high, wide vibrati and trills on glockenspiel bars (Tape Example 17).


Tape Example 14 (from Reynold's A Merciful Coincidence), cited previously to illustrate soft cross-register ululations, also provides an example of a loud cross-register ululation. Although the vocalist is probably not crossing registers at all times during the ascent, the overlap of registers seems to be greatly expanded with an overall effect of a continuous glissando of alternating pitches.



Vocal Fry



Vocal fry is perceived as dry, clicklike pulses and is often used to imitate the opening of a creaky door (hence, another common designation as " creaky voice"). The pulse rate of vocal fry can be controlled to produce a range from very slow individual clicks to a stream of clicks so fast that it is heard as a discrete pitch. It can be produced both egressively (exhaling) and ingressively (inhaling). The individual vocalist may find one mode easier to control than the other in terms of such parameters as pulse rate, dynamics, and pitch. The term "pitch," as used here in relation to vocal fry, refers to the range of perceived pitches rather than to any implication regarding the mode of phonation. Egressive and ingressive vocal fry can greatly expand the practical pitch range (singing and - speaking) of the individual vocalist, male or female.


Egressive vocal fry can be controlled to merge with and extend downwards the lowest part of the egressive singing range. It allows the same degree of pulse rate control as ingressive vocal fry but does not appear to produce individual pitches or pulses as loudly. Microphone amplification is usually necessary to project both modes of vocal fry. An attempt to produce ingressive vocal fry very loudly can result in dryness in the throat or coughing.


Ingressive vocal fry can produce very stable pitches, i.e., pitches which can be sustained with little or no wavering, in the area of E 41 Hz to C# 69 Hz in both male and female voices. Words are easily articulated in this range, as well as in higher ranges, although many of the consonants must be performed egressively for greater clarity or to avoid a lisping effect. As ingressive pitched vocal fry rises in pitch range, it gradually resembles egressive singing, especially above the area of middle C, where it seems to lose any resemblance to the click-like quality of the lower pitches or individual pulses. This area can be controlled in terms of pitch, duration, vibrato, and dynamics (although not as reliably as a comparable egressive range) and is particularly useful in producing very soft, high pitches (area above B 967 Hz). Ingressive phonation in the range above middle C has greater practical value in producing more unusual sounds such as complex multiphonics (see VI. Complex Multiphonics). Pitched ingressive vocal fry in the area between approximately D 73 Hz and middle C seems to be the least practical for individual pitch control or other more specific uses, although speech in this range may acquire an unusual timbre.


Ingressive vocal fry (or low ingressive speech) is the mode of voice production used by the narrator in The Owl and the Pussycat. Tape Example 11 (see III. Ululation) demonstrates the resultant voice quality and illustrates the possibility of wide speech inflections. Later in the piece, all the vocalists are required to speak a short phrase ingressively, the group as a whole covering a wide pitch range from low to high (Tape Example 18).


Specific low pitches are produced by ingressive vocal fry in long passages of the "Dies irae" ("Day of wrath" ) section of Requiem. The buzzy quality of the ingressive pitch B 61 Hz, combined with the characteristics of the other techniques, seemed to the composer particularly appropriate to the mood suggested by the text (Tape Example 19). The ingressive pitched vocal fry which doubles the lower note of the octave chanted (see V. Chant) by another vocalist is more reliable than the chant and insures that the low B will always be present.


One instance of ingressive vocal fry in John Celona's Micro-Macro demonstrates an unusually accurate imitation of the taped computer-generated sweep of harmonics on the pitch Db 69 Hz. In accordance with the instructions to blend with and imitate the tape, the vocalist sustains the same Db and reinforces harmonics in a wide ascending and descending sweep (Tape Example 20), creating a timbre very similar to the taped pitch. A short excerpt from the "Libera animas" (" Deliver the souls" ) section of Requiem requires fast pitch and text changes of low ingressive pitched vocal fry in the lowest voice to create an organ pedal effect (Tape Example 21).


Loud, nonpitched vocal fry begins the "Rex tremendae majestatis" ("King of fearful majesty" ) section of Requiem (Tape Example 22). The unison rhythm of the taped voices is answered in unison by the live voices, both at a fortissimo level made possible through close-miked amplification. Immediately preceding this section, a very different effect is created through a single voice producing a softer, slower vocal fry against a background of sparse whispers (Tape Example 23).


Very slow, individually pulsed nonpitched vocal fry forms the basis for text articulation in Roger Reynolds' Still (Tape Example 24). Syllables and words are formed and gradually perceived over long, fairly even pulse streams. The eerieness of inflected but unvoiced, pulsating words is often increased by overlaid whispers and by the air sounds between pulses of the vocalist's slowly indrawn breath. A slightly different example of nonpitched, individually pulsed vocal fry from the middle of David Evan Jones' Pastoral (Tape Example 25) demonstrates an interplay of varying pulse speeds between live and taped voices.




The term " chant" refers to one technique of singing an octave multiphonic, i.e., one voice producing simultaneously two pitches one octave apart. Low-pitched chant (the lower pitch in the area of B 61 Hz to D 73 Hz) resembles chant used in certain Tibetan Tantric Bhuddist schools. This Tibetan chant sounds like a three-note chord because of a strong reinforced harmonic above the octave multiphonic. When chant is produced with higher pitches, the harmonic content is less noticeable and the sound is thinner and buzzier. The vocalist's physical sensation of chant may be described as a light singing tone combined with vocal fry. The sung tone is the upper pitch of the octave, and the vocal fryseems to lock into place with it to produce the lower pitch. This rather delicate process renders the chant less reliable than other techniques; the least amount of phlegm may necessitate considerable clearing of the throat to produce a clear, unbroken, resonant sound. In some instances, a loud sung tone preceding a chanted octave seems to prepare the vocalist in the same way that clearing the throat would remove a mucus obstruction. One example of this occurs at the end of Psalm of These Days II (Tape Example 26). The vocalist generally has little or no trouble chanting a C (130 plus 65 Hz) after several seconds' rest following a quadruple forte Bb 233 Hz. In this example the original notation showed only a low C (65 Hz). The. vocalist could not produce this note in his normal singing range and chose chant as the technique which would produce a C 65 Hz that would blend best with the other pitches. Chant is used effectively by both men and women to produce pitches lower than those in the normal singing range.


Chant seems to require an extremely steady airflow to maintain a smooth, unwavering tone. It is best produced with straight tone. The addition of vibrato, which alters pitch and/or intensity, or ululations, which regularly interrupt the airflow, might upset the delicate balance of whatever vocal adjustment produces the chant. Fast pitch changes are less dependable than long or repeated notes and may result in the loss of the lower chanted pitch.


One or more of the following suggestions may insure a reliable chant production at a specific point in a composition. The chant may be: (1) prerecorded on tape; (2) performed by several vocalists simultaneously; or (3) used in combination with other techniques which would double the pitches of the chant. The lower chanted pitch is usually the first to disappear, leaving a weak or unsteady upper pitch.


The opening of Requiem uses chant to highlight certain portions of the text. A solo voice on prerecorded tape provides a reliable basis for the live soloist to double selected words with chant. Refer to Tape Example 3, in which the soloist chants with the tape on the words "Domine" and "Deus in Sion," as well as the entire last sentence (Tape Example 27).


The "Dies irae" section of Requiem begins with chant and pitched ingressive vocal fry combined on tape (refer to Tape Example 19). The texture gradually thickens with additional taped voices chanting on higher pitches (each F# and B up to B 493 Hz). Some chanted pitches are doubled with egressive singing or ingressive pitched vocal fry; the live voices use either chant or vocal fry on the lowest pitches. Tape Example 28 provides a short excerpt from the most dense part of the section.


Nonpitched chant combined with vocal fry and glottal whispers produces a quite different total sound. These techniques are combined in both live and taped voices in Requiem (Tape Example 29), using the words "dona eis requiem. Amen" at the end of the " lacrymosa" section.


Low chant, even when soft, is generally quite resonant and full. This may explain in part why it can effectively produce a settled feeling or a cadential sense of resting point or arrival. Isolated examples of this cadential quality appear at the end of Psalm of These Days II on the last syllable of " forever" (refer to Tape Example 26) and on the word "sleep" in Pastoral (Tape Example 30). A more extended use of chant to produce the sense of a strong final cadence appears in the continual overlapping of live and taped voices chanting a low B (123 plus 61 Hz) in the last section of Requiem (Tape Example 31).


Two instances of chant in Psalm of These Days II illustrate a sudden shift in several parameters when chant immediately follows normal sung tones without vibrato. In both examples (Tape Example 32, 33) pitches become much lower and dynamics much softer, and there is an obvious timbral shift to a thinner, buzzier quality. Even though three of the four voices in the first example drop only one pitch, the addition of the lower octave suggests a large intervallic leap. The pitch shift in the second example is much more extreme, the distance in each voice covering two to three octaves. In both instances, the sudden dynamic reduction is a function of the chant technique, especially since only general rather than individual microphone amplification is used.


Chant functions as an ornament to the word "my" in Psalm of These Days 11 (Tape Example 34). In this instance, the instructions for two of the four voices indicate "multi-phonic chant on and off." Each voice sustains one pitch with several similar ornamental figures of rapid pitch changes resembling a trill, the chant adding a lower octave at random.


Complex Multiphonics


The term "complex multiphonics" rather loosely designates a cluster of sounds produced egressively or ingressively by one voice. The cluster may be perceived as a number of non-intervallically related pitches resembling noise or as a complex mixture of vocal fry with other sounds or pitches. The total mixture can cover a narrow or wide band of sound at various general pitch levels (low, medium, high) as well as on and around specific, perceivable pitches. Complex multiphonics vary greatly among individual vocalists but are fairly consistent for each individual. Those which are most reliable for the individual vocalist can generally be reproduced with a similar degree of complexity at approximately the same pitch levels.


The major difference between complex egressive and ingressive multiphonics lies in the amount of air used to produce them. The egressive version, referred to as " forced blown" by EVTE, requires a large amount of air, is usually fairly loud, and can be sustained for a short time only. Although it is possible to produce complex ingressive multi-phonics with a large amount of air, this yields a gasping sound and may lead to coughing or choking. However, if a very small amount of air is gradually drawn in, very complex multiphonics can be sustained for a much longer time (one long breath). The physical sensation is similar to ingressive vocal fry. Both types of multiphonics are best amplified with microphones to avoid undue strain on the vocalist in an attempt to make them louder. Both ingressive and egressive multiphonics tend to tire the voice more quickly than other techniques and should be used carefully.


A short burst of multiphonics supports the climax of a middle section of Requiem at the words " confutatis maledictis, flammis accribus addictis" (when sentence is passed on the damned and all are sent to piercing flames" ). The upper two of four taped voices hold a long, high ingressive multiphonic while the lower two have several repeated lower egressive bursts (refer to Tape Example 16). The combination of these multiphonics with high whistles, vocal fry, and cross-register ululations creates a complex texture covering a wide, approximate pitch range. In a much longer section in Still (Tape Example 35), the illusion of a light wind gradually developing to hurricane proportions is achieved through various combinations of whistles, high ingressive pitches, and complex multiphonics.


Since complex multiphonics do not always "speak" immediately, it is wise to allow for some preparation. Several possible types of preparation may include: (1) using other loud or complex sounds to cover the entrance of the multiphonic; (2) instructing the vocalist with the most reliable multiphonic to enter first if several voices are to produce multiphonics; (3) preceding the multiphonic with a sound from which it is relatively easy to build the multiphonic. One example of the last suggestion occurs in Requiem in the "Rex tremendae majestatis" section (Tape Example 36). In both taped and live parts, vocal fry is followed by high, complex multiphonics. The physical sensation in producing both techniques is somewhat similar, and one seems to follow the other quite easily. The long sustained multiphonic on "salva me" near the end of the example is prepared by the shorter preceding multiphonics in the same voice and its entrance is covered by multiphonics in other voices.


A more extended example (Tape Example 37) of complex multiphonics set to a text occurs towards the end of A Merciful Coincidence. The overlapping of most of the entrances helps mask any awkward beginning and allows each multiphonic to grow out of the preceding one.


Complex multiphonics can also effectively contrast with, or punctuate, single pitches. In one section of A Merciful Coincidence, the three vocalists, in various overlapping combinations, articulate the text over a sustained, non-vibrated, rather piercing C (523 Hz). Specified consonants are "barked," resulting in a multiphonic which briefly interrupts the sustained tone (Tape Example 38). A similar example occurs later in the piece (Tape Example 39). The text is articulated on a single pitch E, 659 Hz. Each word is short, loud, and further resonated by a piano sound board which exaggerates the contrast between the straight tones and multiphonics.




The musical examples and explanations of selected extended vocal techniques included in this paper are intended to give composers and singers some idea of the practical uses and limitations of a number of very different sounds which previously have seldom been used in vocal writing in Western art music. Although the techniques descussed belong specifically to the sound vocabulary of the Extended Vocal Techniques Ensemble of San Diego, California, they are not limited to this group of vocalists. The EVTE has given a number of workshops in the United States and Europe and has discovered that most of the techniques are quickly, if less skillfully, reproduced by persons hearing them for the first time. The most readily accessible of the five technique discussed in this paper are reinforced harmonics, ululations (single pitch), and vocal fry. Complex multiphonics and chant generally take much longer to learn to control, although the basic mechanism involved may be imitated or understood from the first attempt. Singers and others who try to reproduce these techniques without supervision, especially those techniques which involve loud and/or complex sounds, are urged to exercise care. One may involve unnecessary muscles which could cause tension or strain when first experimenting with the sounds and continuous attempts may overtire the voice. Sensible pacing of the voice is necessary with both "normal" singing and extended vocal techniques. They are quite compatible and extended techniques can be even more effective in the context of the normal sung tone than if totally isolated.


Microphone amplification is a necessary component of extended vocal techniques performance. Individual amplification (a microphone for each vocalist) is preferable and usually essential. Certain techniques require delicately balanced vocal production or are inherently quite soft. These sounds cannot, be projected without strain and often lose their essential characteristics if pushed. The vocalist must develop skill in moving close to or away from the microphone for appropriate techniques and dynamic specifications. Close-miking requires proper care to avoid " popping" or " blowing" the microphone, amplifying extraneous air sounds (including air escaping through the nose), or exploding consonants (especially egressive consonants with ingressive vowels). Composers should remember that close-miking requires recurring head motion to breath off-microphone. Theatrical gestures or deliberate lack of motion should be planned accordingly.


The use of prerecorded tape offers several possibilities, for example, taped nonvocal sounds can provide a musical context for live vocal sounds. Due to the rich variety of sounds available through extended vocal techniques such as those discussed in this paper, the composer may wish to work with vocal sounds only. Techniques which are less reliable in realtime performance can be put on tape with the assurance that they will appear at the appropriate moment in the composition. Prerecorded tape can provide the opportunity for more voice parts to occur simultaneously than are available with the number of live performers (this is especially useful when the number of available vocalists who have perfected certain techniques is limited). Complex mixtures of sounds are possible on tape whereas they may be impossible with live performers.

Notation of these techniques has not been standardized. In many instances, verbal instructions added to a graphic or traditionally notated score are quite adequate. Certain stringed instrument notation might be adapted for similar vocal techniques: specific vocal harmonics may be indicated by diamond-shaped noteheads; ululations may be indicated by the notation for tremolo. When long melodic passages are produced with reinforced harmonics, separate staves for the fundamental pitches and harmonics might be useful. The composer must decide what form of notation is most accurate and meaningful to the performer.


The extended vocal techniques chosen for discussion in this paper represent only a portion of EVTE's sound vocabulary, which in turn represents only part of an undefined limit of .vocal sounds. Through the work of vocalists such as the members of the EVTE, a colorful and extensive vocabulary of reliably and consistently reproducible sounds is made available to other vocalists and composers interested in expanding traditional vocal boundaries.

List of References 


Celona, John Anthony. 1975. Micro-Macro (computer generated tape and voices. Unpublished.

Jones, David Evan. 1977. Pastoral (voice and prerecorded tape). Unpublished.

Kavasch, Deborah.   1980. The Owl and the Pussycat (seven voices), Editions Reimers, Stockholm.

                               1976. Tintinnabulation (three voices and glockenspiel). Unpublished.

                                     1977. Sweet Talk (women's chorus). Unpublished.

                                     1978. Requiem (four voices and prerecorded tape). Unpublished.

 London, Edwin. 1977. Psalm of These Days II (four voices). Unpublished.

Reynolds, Roger. 1976. Voicespace: I. Still II. A Merciful Coincidence (taped voices), C. F. Peters Corp., New York, NY..

Yuasa, Joji. 1976. My Blue Sky in Southern California (computer-generated tape and voices). Unpublished.


List of Taped Examples


1.                                              Reinforced harmonics in The Owl and the Pussycat by Deborah Kavasch

2.                                              Reinforced harmonics in Sweet Talk by Deborah Kavasch

3.                                              Reinforced harmonics from simultaneous, different texts in Requiem by Deborah Kavasch

4.                                              Reinforced harmonics in Psalm of These Days II 7 by Edwin London

5.                                              Reinforced harmonics in Psalm of These Days II

6.                                              Reinforced harmonics over gliding fundamentals in My Blue Sky in Southern California8 by Joji Yuasa

7.                                              High harmonic oscillation in My Blue Sky in Southern California

8.                                              Harmonic oscillations in Tintinnabulation by Deborah Kavasch

9.                                              Melody produced by reinforced harmonics in Requiem

10.                                           Ululations in The Owl and the Pussycat

11.                                           Ululations (vowels and consonants of" pussycat") in The Owl and the Pussycat

12.                                           Ululation of melodic fragments in The Owl and the Pussycat

13.                                           Unvoiced ululations in Requiem

14.                                           Soft cross-register ululations in A Merciful Coincidence9 by Roger Reynolds

15.                                           Simple and cross-register ululations in Requiem

16.                                           Cross-register ululations combined with other extended vocal techniques in Requiem

17.                                           High cross-register ululations in Tintinnabulation

18.                                           Ingressive vocal fry as speech in The Owl and the Pussycat

19.                                           Low. pitched ingressive vocal fry in Requiem

20.                                           Ingressive vocal fry with reinforced harmonics in Micro-Macro10 by John Celona

21.                                           Fast pitch and text changes of low ingressive vocal fry in Requiem

22.                                           Nonpitched vocal fry in Requiem

23.                                           Slow. nonpitched vocal fry in Requiem

24.                                           Slow, nonpitched vocal fry in Still11by Roger Reynolds

25.                                           Interplay of vocal fry in Pastoral12 by David Evan Jones

26.                                           Chant in Psalm of These Days II

27.                                           Chant used for text emphasis in Requiem

28.                                           Chant in simultaneous voices over a wide pitch range in Requiem

29.                                           Nonpitched chant in Requiem

30.                                           Chant in Pastoral

31.                                           Chant in cadential context in Requiem

32.                       Chant in Psalm of These Days II

33.                                           Chant in Psalm of These Days II

34.                                           Chant as ornament in Psalm of These Days II

35.                                           Complex multiphonics in Still

36.                                           High. complex multiphonics in Requiem

37.                                           Complex multiphonics in A Merciful Coincidence

38.                                           Complex multiphonics in A Merciful Coincidence

39.                                           Complex multiphonics in A Merciful Coincidence



 1 This paper was previously published as Vol. 1, No. 2 of the Reports from the Center which was released in No‑vemb5r 1980, by the Center for Music Experiment at the University of California, San Diego.


 2 The Lexicon and Index are available upon request through the Center for Music Experiment, University of California, San Diego, La Jolla, Ca. 92093.

 3 All values expressed in Hz (cycles per second) are approximate.

 4 The word "texture" in this paper refers to the combined effect of all parameters in a musical composition.

 5 Bracketed symbols refer to IPA (International Phonetic Alphabet).

 6 For the purposes of this discussion, vocal registers are designated as follows. Male voice: (1) chest, up to middle C•area: (2) falsetto, up to high C (1046.5 Hz) area: (3) whistle stop, above high C area. Female voice: (1) chest, up to middle C area; (2) middle, up to E 330 Hz area: (3) head, up to E 660 Hz area: (4) whistle stop, above high C area. These are arbitrary limits which may overlap considerably in individual cases.

 7 All taped examples from Psalms of These Days II are ©1981 by Henmar Press Inc. Reproduction by permission of C. F. Peters Corpori.

 8 All taped examples from My Blue Sky In Southern California are used by permission of the composer.

 9 All taped examples from A Merciful Coincidence are used by permission of the composer.

10 This taped example from Micro•Macro is used by permission of the composer.

11 All taped examples from Still are used by permission of the composer.

12 All taped examples from Pastoral are used by permission of the composer.