Especially from the 19th century onwards, music - in particular 'absolute music' - used to be invoked as the paradigm of an 'abstract', non-narrative art, where sound, otherwise than form or colour in painting, would not be subordinated to the task of imitating - just think of Eduard Hanslick's statement that music is no more than 'tönend bewegte Formen'. As a consequence of the increasing taboo on mimesis, music came to serve as an example to all the arts that wanted to get rid of the 'narrative' to become 'abstract' like music - think of Walter Pater's 'All art aspires to the condition of music.' The aspiration was strongest with painters, who, in their fervour to distance themselves from photography, threatened to betray the very essence of painting. Meanwhile, the idea that music would be an abstract art - just like architecture - has become so established that it is one of the strongest obstacles to the restoration of the idea the art is mimesis.
To convincingly demonstrate that - or in how far - music is mimesis, is therefore the touchstone of every theory of mimesis, and it is to this task that this text is devoted.
Let us first clearly state that not all music is mimetic. In 'movement conjuring signs', we described the domain of speech and dance music and showed that there is no talk here of mimesis, but of semiosis: speech music does not imitate speakers or singers but signifies their speaking and singing, and dance music does not imitate dancers, but signifies their dancing. The question is then what kind of music is mimetic.
Second, we have to point to the fact that, next to musical mimesis, there is also ordinary auditory mimesis - auditory mimesis where non-musical auditory appearances are duplicated. Anticipating what follows, we can sketch the following scheme:
In the text below, we will first examine the different kinds of auditory mimesis and then describe the many ways in which they can be combined with each other and with the many kinds of non-mimetic (speech and dance) music.
Mimesis is the imitating of objects or beings through the duplication of one of their sensory appearances. With visual mimesis (think of the paradigmatic examples of the mirror and the photo), something is imitated through the duplication of its visual appearance. With auditory mimesis, something is imitated through the duplication of its auditory appearance (think of the obvious example of the recording of a voice or of a bird song). Just like Narcis' mirror for visual mimesis, Echo's echo is the paradigm for auditory mimesis (For further elaboration, see: 'Mimesis')
We can divide the domain of auditory mimesis into subdomains, according to the kind of auditory appearance that is duplicated.
Next, there is the domain of the intended - intentional - sounds that are produced as signs by (existing or imaginary) beings. This domain falls apart in the domain of human speech and the domain of musical appearances.Let us first examine the domain of human speech. Speech can by duplicated in view of imitating a character. This can be done by recording (like in radio plays or film), but also by another speaker (a reciter or an audiovisual actor), in which case we are still dealing wit ordinary auditory mimesis. But when it is the voice that duplicates speach, we are dealing with the domain of verbal or literary auditory mimesis. Next to uncompleted verbal auditory mimesis, there is also completed verbal mimesis, when imaginary beings - speakers of nonexistent or incomprehensible languages - are imitated through the production of their verbal appearances, as with Hugo Ball or Kurt Schwitters (or with electronic deformation of speech) (for wordless speech music, as in 'Speakings' of Jonathan Harvey, see below).
Next to the domain of speech there is the domain of the (existing or imaginary) musical appearances - from singing birds, over lamenting Ariannas, to the sonorous beings that we described in 'sonorous beings in musical space'. Also these musical appearances can be duplicated in view of the imitation of the concomitant phenomena or beings through recording, in which case we are dealing with ordinary mimesis again. But, when the duplication is realised by the human voice or musical instruments, we can speak of musical auditory mimesis (musical imitation, musical mimesis): mimesis of existing or imaginary phenomena and beings with a musical auditory appearance. An overview:
MUSICAL MIMESIS (1): DOMAINS
The domain of musical mimesis can be divided according to the kind of musical appearance that is duplicated:
First, there is the domain of existing musical appearances - the domain of 'found music'. To this domain belong all the non-verbal intentional signs that are produced by animals (song of whales, howling of wolves, but foremost the song of birds like in 'Le chant des oiseaux' of Janequin, Beethoven's Pastorale, Respighi's 'Pini di Roma' or Per Norgaard 'D'Monstrantz Voogeli', yes even of fighting apes, like in the Balinese Kecak), but above all by man (the many auditory expressions of man like weeping, laughing, sighing, screaming, which, as described in 'movement conjuring signs' are often condensed with words, as well as the many auditory signals and signs like in 'Voulez ouyr les cris de Paris' of Janequin, all sung speech music and dance music). To this subdomain also belong the musical signals produced by all kinds of instruments: think of military signals, bells, train whistles, sirens, fog horns, ringtones and what have you,
Next tot the domain of existing musical appearances, there is also the large domain of phenomena or beings that owe their existence to mimesis (completed mimesis). There are in the first place the existing beings whose auditory appearance is musicalised: bumble bees in Rimsky-Korsakov, chicken in 'La Poule' of Rameau, dogs (think of the singing dogs in 'Lady and the tramp'), cats (like in Rossini's 'Duetto buffo di due gatti'), the laughing of witches with Purcell or the cries of the Valkyries in Wagner, the wordless speech of imaginary beings like in Speakings of Jonathan Harvey, the imitations of the human voice on the discontinuous sound spectrum of the piano by Peter Ablinger (in the wake of Godfried-Willem Raes from Logos), and the musicalised 'Typewriter' of Leroy Anderson. Next, there is the large domain of the existing or imaginary singers and musicians like the Bremer Stadtmusikanten, the Lorelei and the Sirens, the lamenting Arianna of Monteverdi, the singing monks of Ketelby, the Jews that sing 'Shema Yisroel' in Schönberg's Survivor from Warsaw, the military fanfare of the Roman legion in Pini di Roma of Resphighi. And, last but not least, the extensive domain of of the imaginary wordless singers (singing instrumental speech melodies like that of Mahlers Adagietto).
The most important domain of the imaginary beings with a musical appearance, however, is the domain of the sonorous beings that we described in 'sonorous beings in musical space', where we demonstrated how they are created in that singers or musicians duplicate their musical appearance: the realm of absolute music.
This results in the following survey:
MUSICAL MIMESIS (2): MUSICAL MIMESIS IS IMITATING THROUGH PLAYING MUSIC
Whereas the concept of 'verbal mimesis' - the auditory imitation of Hamlet through duplicating his speaking - is not problematic, the concept of 'musical mimesis' - the auditory imitation of Arianna through duplicating her lament - is from way back the source of the endless misunderstandings that continue to haunt the philosophy of music (see Kivy). We already dealt with one of these misconceptions in our text ' 'sonorous beings in musical space': the idea that absolute music would be non-mimetic. Even more confusing are the misconceptions about what imitates what in one of the oldest paradigms of musical mimesis: Monteverdi's Lamento d'Arianna. Let us therefore set things straight.
Let us begin with a terminological question. There is no such thing as imitated music - or broader: an imitated sound - except as imitation in the performative sense: the making of a (monosensorial) duplicate through recording or through playing the music - not in the sense of an 'auditory image' of the music or the sound in question. Precisely the making of such a duplicate is a means of imitating (in the perceptive sense) singers or musicians or musical phenomena or beings. Whoever imitates the song of a bird (in the performative sense of playing the same tones on a flute) makes a (monosensorial) duplicate of that song, and such a duplicate is an imitation (in the perceptive sense of an 'auditory image') of a singing bird. Or whoever 'imitates' the lamenting of Arianna, makes a (monosensorial) duplicate of that sound, and that duplicate is an imitation (in de perceptive sense of an 'auditory image) of the lamenting Arianna. To avoid any ambiguity in the use of the word 'imitation', we will rigorously use the formula 'to duplicate the auditory appearance in view of the imitation of the concomitant phenomenon or being'. We do not say that Monteverdi's 'Lamento d'Arianna' is the imitation of a lament, but rather that it is an auditory duplicate of such a lament, and that precisely the making of such a duplicate (the singing of the lament) - is an imitation (an auditory image) of the lamenting Arianna, because it only duplicates the auditory appearance of Arianna, and not her visual or tactile appearance as well. Similarly, the visual appearance of the David of Michelangelo is a visual duplicate of the original David, and that visual duplicate is a visual image of the original David, because it only provides the visual appearance and not also his auditory, tactile ... appearance. The music as such (the lament of Arianna as sung by the singer) cannot be distinguished from the lament of the (imaginary) original Arianna, just like the visual appearance of the original David cannot be distinguished from the visual appearance of the sculpted marble block.
Although there is no such a thing as 'imitated music' in the perceptive sense, there is something like mimetic music ('imitative music'): the duplicating (playing, singing) of musical appearances with musical instruments or the human voice, in view of the imitating of the concomitant phenomena - the phenomena that normally produce the appearance in question. Monteverdi's 'Lamento d'Arianna' is not the imitation of a lament, but the imitation of the lamenting Arianna through the singing of her lament - through the production of a duplicate of her lament. We can also speak of musical mimesis or musical imitation: the terms clearly indicates that we are not dealing with the imitation of music, but rather with imitating through (the playing of) music: just like with verbal imitation, we are not imitating words, but a character through duplicating his words.
Whether music is mimetic or not, cannot be decided through hearing the music as such, but only through assessing the relation between the sound source and the music: is the sound source the original bearer of the appearance or not. That goes not only for musical mimesis: with auditory mimesis as such, the auditory appearance is not provided by the original phenomenon or being (thunder, a lamenting woman, a musicalised clock or a sonorous being), but by loudspeakers, singers or musicians. And that goes not only for musical or auditory mimesis, but for every kind of mimesis: also the smile of the Mona Lisa does not differ from the smile of the (imaginary) original. What makes the painting of such a smile an imitation, is that it is produced by a panel and not by the original Mona Lisa. The resonating loudspeakers, the manipulating of the wind machine, the singing of the singer or the playing of the musician are merely the medium bearers of the auditory appearance, the musical counterparts of the speaking of the actor who plays Hamlet, or the auditory counterpart of the canvas or the screen of the film. Just like the Mona Lisa does only appear as a three-dimensional animated person when we forget that we are dealing with paint on a panel, just so do Hamlet or Arianna only appear when we forget that we are dealing with an actor or a singer.
To my knowledge, no author has adequately described this state of affairs. The consequences are dramatic. All the theories that want to explain what auditory mimesis (or to deny that music would be mimetic), turn out to be mere theories of what an auditory appearance is. Exemplary are the countless attempts at explaining how 'music' can be expressive - why Arianna's lament is a lament. That is a problem for semiology or psychology, whereas the question that is relevant to the philosophy of art is wherein ordinary, non-mimetic music differs from mimetic music - how it is that we have the impression to hear Arianna where there is only a singer singing. A common misunderstanding is the statement that the lament of Monteverdi would be an 'imitation' or an 'analogy' to a real lament. At best, a real lament may have been used as a model when conceiving the lament of Arianna (when designing the original), but, what makes Arianna's lament a musical imitation is not the relation between the original in the music and an original in the real world that served as a model, but the simple fact that it is not Arianna that sings the lament, but a singer. A similar analysis applies to absolute music (see 'sonorous beings in de musical space'): moving tones are not imitations, but rather duplicates of the auditory appearance of imaginary sonorous beings, produced by the musicians. And a similar analyse applies to the visual image as well: the question is not why a smile expresses what it expresses, but rather how it is that; when perceiving this visual configuration on a two-dimensional panel, we have the certain impression that there is an animated three-dimensional being of flesh and blood.
MUSICAL MIMESIS (3): DIFFERENT WAYS OF DUPLICATING A MUSICAL APPEARANCE
It is not always easy to answer the question whether music is mimetic or not - whether the musicians are the original producers of the auditory appearance. There are different ways of duplicating a musical appearance, and some of them are rather confusing.
The least problematic method is the recording of an musical auditory appearance and its rendition through loudspeakers: the recording of bird song, a mother singing a lullaby, a vocal or instrumental singer, or a musician playing absolute music. Such a recording - ordinary auditory mimesis - is the auditory counterpart of the visual photo. And just like with a photo, there is no doubt that we are dealing with an imitation: just like we cannot touch what is duplicated visually on the photograph, we cannot see what is recorded.
Equally unproblematic are the cases of auditory imitation where the musical auditory appearance is duplicated through another sound source than the original one, as when birds are imitated on the piano, or cats on the violin. When Rossini's cat's duet is performed, we hear cats but see singers. Problems arise only when we are dealing with completed mimesis so that there is no comparison with an existing original. That is especially the case with the imitation of sonorous beings in absolute music, where the problem is enhanced in that it is difficult to conceive of purely sonorous beings, so that we are all to readily inclined to resort to the idea of an expressive singer or musician. Only when we fully realise that the sounds in musical space are monosensorial beings, do we recognise that the instruments are not the original sound source (which, by the way, can be rendered more or less adequately: think of the various instrumental versions of Bach's fugues).
More problematic are the cases where the musical auditory appearance is duplicated with a similar sound source: the imitation Arianna's lament with a voice, the imitation of Elvis Presley with a voice and a guitar, the imitation of Glenn Gould playing Bach by a pianist. These cases are comparable with the duplication of the audiovisual appearance of characters by actors. Here, it is tempting to forget that there is a difference between the imitating singers and musician and the real ones. The danger is not so great when existing singers or musicians are imitated - think of Edith Piaff: the differences between the original in the real world and the imitation mostly catch the eye. But the danger is real when that is not the case, as when Schubert imitates a 'Leiermann' or Stravinsky street musicians in his Petrushka; because there are no individual clues, it is mostly the context that suggests that we are dealing with imitations altogether (the text in the case of Schubert, or the story in Petrushka) and/or the fact that the original instruments are replaced with other ones, like the piano in Der Leiermann or wind instruments in Petrushka. The danger is acute, when we are dealing with imaginary singers like the lamenting Arianna or the Jews that sing 'Shema Yisroel' in Schönberg's Survivor from Warsaw. Since we are dealing with completed mimesis, there is, just like with absolute music, no comparison with an existing original, although, unlike with absolute music, we have no difficulty in imagining such an original. And the confusion is practically inevitable when the singer or the musician imitates himself. In the supposition that the song is autobiographical, 'Ne me quittes pas' was at best a direct expression of Brel at the moment of its creation, but with every repetition, Brel is staging himself. The phenomenon has become the rule since lyric songs are no longer performed by random performers, but, as a consequence of new ways of distribution (radio, recording) the performance is monopolised by a privileged performer. If 'Ne me quittes pas' were a song that, like those of Schubert, were performed by various singers, the version of Brel would be perceived as the imitation of an imaginary lover by Jacques Brel. But, since Brel, otherwise than the singers who interpret Arianna's lament, is the standard performer, it becomes even more difficult to make the already problematic distinction between the imaginary Brel and the real Brel that sings his song.
Outright confusing, finally, are the cases of double mimesis: the cases where the imitated singers or musicians play absolute music and hence imitate sonorous beings in their turn. Problematic is foremost the recognition that we are dealing with musical imitation (of the performance of Bach by Glenn Gould). These cases of double imitation have to be discerned ffom the cases where the music is duplicated without duplicating a particular interpretation: as when a musician just interprets a score or sings a cover. Also recordings (ordinary auditory mimesis) of Bach played by a particular interpreter can be heard as duplicates of the music, and not as duplicates of the interpretation as well.
As soon as we rely on the proper theoretical model, it appears, hence, that there is only a difference qua domain between musical imitations of the musical appearance of existing beings (from birds to singers or musicians), of musicalised existing beings, of wordless speakers and singers, of imaginary singers and musicians, and, finally, of sonorous beings in musical space. The domains themselves can be ranged on a continuum or probability: from imitation of existing musical appearances, over imaginary but 'probable' appearances (like those of Arianna), to imaginary but improbable (those of the sonorous beings of absolute music).
Before describing the combinations of ordinary, verbal and musical auditory mimesis, we have first to consider a second kind of mimetic music:
IMAGE CONJURING MUSIC
Let us go back to the domain of non-mimetic music - the ordinary speech and dance music. These bear the germ of another kind of mimesis. In that the production of movement conjuring signs becomes exherent to the movements, dance music is no longer restricted to movements with inherent sounds, like speaking. On the one hand, that unleashes the imagination in conceiving of new kinds of movements: dance movements as well as ways of playing. But, on the other hand, that creates the possibility to conjure up movements that can no longer be performed by the human body: movements of non-human beings like animals, machines, natural phenomena, or movements of imaginary beings as such.
There are two possible reactions on such a bloom of the design of movement conjuring signs. A first reaction is that music continues to function as dance music - as music that has to conjure up movements of human bodies. This leads to the many kinds of mimetic dances - forms of visual mimesis that does not concern us here, because it is no auditory, but visual mimesis. Suffice it to remark that there is a whole spectrum of such mimetic dances: from dances imitating activities like work and war, over imitations of animals, to imitations of dragons, monsters and ghosts. The imitation is mostly restricted to partial filling in of the movement - think of movements of the arms as if it were the wings of a bird, so that there is no visual resemblance. We can compare with the role play of children, where the child imagines being a character and adds to this mental representation partial immediate imitations of movements (see 'Imitation and play').
But there is another way out: the listener can give up his urge to dance altogether. Instead of moving the arms to 'imitate' the movement of the wings of birds or of the coupling rods of a locomotive, he can imagine a bird or a locomotive and have the representations execute the conjured movements. The movement conjuring signs are transformed into image conjuring signs - dance music becomes image conjuring music. That entails a transition from unmediated to mediated mimesis: the conjuring up of representations through words or images - or: music. Since movements with inherent sounds are predestined to unmediated auditory mimesis, image conjuring music has a predilection for movements that make no noise (the flying of birds) or the conjuring up of standstill (the endless steppes). Whereas verbal image conjuring signs do not contain instructions for the tempo and the duration of events, musical image conjuring signs movements prescribe the tempo and the duration of the movement in 'real time'.
It is important to remind that image conjuring music, just like speech and dance music consists of predominantly analog signs: it does not resemble the auditory appearance of the imaginary objects or beings - it is not an imitation, but a sign for (representations) of movements. We do not conceive of the sugar plum fairy of Tchaikovsky as a fairy that sounds like a glass xylophone, but rather as a fragile being.
Let us give a brief survey of the phenomena and beings whose representation can be conjured up through image conjuring music.
To begin with, there are the (inaudible) movements of inanimate beings: the waves (Der heilige Franziskus von Paula über Wogen schreitend or Debussy's 'La Mer', fontains (Lisztz's 'Les jeux d'eaux de la Villa d'Este'), the water of the Rhine (Wagner's 'Vorspiel' to Rheingold), the flickering of light on water (Schönberg's 'Farben') or the flashing of the lightning (Beethoven's 'Pastorale'). Not only movement, but also standstill can be conjured up: the motionless see (Monteverdi's 'Hor ch’el ciel'), an impressive castle (Smetana), the endless steppes (Borodin's 'Steppes of Central Asia').
Next, there are the countless examples of movements of animals: the plodding hoofs of the camels in Borodin's 'Steppes of Central Asia', the swimming of fish (Schubert's 'Die Forelle'), the flight of birds (Schubert's 'Die Krähe'), the flutter of angels' wings (Monteverdi's 'Et hi tres').
Also the representation of the movement of vehicles can be conjured up. Very popular is the representation of the up and down motion of connecting rods of a locomotive (Steve Reich's Different Trains).
More ambivalent is, finally, the status of the movement conjuring signs that can be executed by human beings (see 'Proteus' below). It applies also here that they conjure up images when there are additional verbal clues. But when there are no such clues, there is only the perception of an impulse that is not executed. When we restrict ourselves to examples where there are verbal clues, there are, to begin with, ordinary movements like walking (Berlioz' 'Harold en Italie') or treading the ice (Vivaldi's 'Winter'). More important are expressive movements. There are inconspicuous expressive movements like the shivering in Purcell's 'Cold song', the teeth chattering in Vivaldi's winter, the quivering in Berlioz' 'La Mort de Cléopatre'). There are also the more conspicuous expressive movements of arms, trunk or the whole body (think of the many examples of expressive strides like in Chopin's funeral march of Wagner's Siegfried's Tod).
Let us remark that also the duplicate of an auditory appearance can function as an image conjuring sign: as when the bird song conjures up the representation of a bird. A good example are the many sounds in Salvatore Sciarrino's 'Autoritratto nella notte'. But these appearances can also be heard as a pure auditory imitation, without conjuring images. As a rule, however, image conjuring is here prevented in that many auditory imitations are part of audiovisual imitations, like in film.
AUDITORY EN MUSICAL MIMESIS
It is apparent, then, that there are two kinds of musical mimesis; unmediated musical mimesis and mediated musical mimesis. Only unmediated musical mimesis is auditory mimesis. Mediated musical mimesis is not auditory: that the image conjuring signs are audible, just like the words in image conjuring (= narrative) literature, does not make the representations auditory. The sensory domains of the representations are rather non-auditory per definition; when a train whistle is used to conjure up the image of a train, it is no longer necessary to make an auditory representation of that whistle. Neither does image conjuring music become a visual image when it is written down in a score: also a novel or a poem are not transformed into a visual image when the words are written.
To give account of these two kinds of musical mimesis, we have to extend the above scheme of auditory scheme above as follows:
It is apparent, then, that auditory mimesis and musical mimesis are two domains that are not co-extensive, they merely overlap.
Let us remark that we could make a similar distinction for the two other domains: there is also unmediated and mediated verbal mimesis, and there is also unmediated mimesis of objects and mediated mimesis through objects (think of commemorative sites). But it would be confusing to try to combine this scheme with a generalised distinction between unmediated and mediated (unless we would replace auditory mimesis with unmediated mimesis as such).
THE COMMUNITY OF IMAGINARY BEINGS: SYMPATHETIC MIMESIS (2)
In 'sonorous beings in musical space' we have seen that, when in absolute music the movement of the sonorous beings is organised through movement conjuring signs, these not only conjure up the movement of the sonorous beings, but also those of the listener and the musicians, although it is not their body that is set in motion, but the sonorous soul that they have become through identification with the moving sound, and we described how they thus become part of the community of sonorous souls. That applies not only to the sonorous beings of music in particular, but for all the phenomena and beings whose auditory appearances are duplicated in view of the imitation of the concomitant phenomena or beings - as far as they are organised by movement conjuring signs: also the auditory imitations of marching soldiers (the Roman legion in 'Respighi's Fontane di Roma), Purcell's laughing witches and Monteverdi's Arianna invite us to join their movements.
Let us first examine the situation in the case of musical imitations of beings that are more earthly than the sonorous beings of absolute music: singing birds, musicalised appearances, and existing or imaginary singers and musicians. In as far as the musical appearance of these beings contains movement conjuring signs, these signs also conjure up our movements. But, otherwise that with absolute music, that conjures up the movement of sonorous beings, these signs are meant for human beings or at least for beings with whom we have the singing in common. Nothing prevents us from singing with Arianne or Jacques Brel, or to play with the Leiermann or Chuck Berry, and even less prevents us from singing along with instrumental singers. But since also in these cases we are not the imitated beings, we become their fellow-performers (effectively, or merely in our imagination). An imaginary community is constituted also here, albeit that this time we do not have to ascend in musical space as sonorous beings and leave our bodies behind: we sing and play in the same three-dimensional world as that of the imitated beings. That does not prevent the community of singing and playing bodies to be imaginary, and not real, like that of the singers of speech music and the dancers of dance music: our bodies are merely the medium bearers. Apart from that, we sound just like Arianna. But, since our imitation is only auditory, it is all too visible that we are not Ariannas. Next to the imaginary community of lamenting Arianna', and provided all the members of the community proceed to actual singing or playing, there is also a real visual community of performers of the lamenting Arianna. More often, all the listeners content themselves with identifying themselves with the singer on the stage, who imitates the imaginary singer also visually, and to performing vicarious movements like rocking the hips or waving the arms in the air.
With image conjuring music, things are different. That the images are conjured up through movement conjuring signs, does not mean that the images themselves contain (auditory) movement conjuring sings. The movement conjuring signs themselves can elicit the movements of the listeners - on which, more in the next paragraph.
IMPLOSION OF MIMETIC MUSIC IN NON-MIMETIC MUSIC AND THE PROBLEM OF EXPRESSION
In 'sonorous beings and musical space' we also described how the use of movement conjuring signs incited the listener to move or to sing or play along, and how the world of the sonorous beings was supplemented with a world of singing and playing bodies, and how this redoubling threatened to elliptically be read as a single world, whereby the movement of the sonorous beings is read as the expression of the singers or the musicians).
Also the remaining forms of mimetic music, where not the musical appearances of sonorous beings are duplicated, but that of existing or imaginary singers and musicians, of musicalised beings or wordless singers, can tempt the listeners to sing of or play along. As described above, a second world emerges, with the difference that the singer is here perceived as an auditory appearance in real space, where the singing body is merely a medium bearer. Here it is not so much musical space that threatens to implode, bur rather the real world that tends to be equated with the imaginary world - in that the singers deem themselves Arianna. The danger is not so great when the original is a concrete person - think of Edith Piaff of Jacques Brel - or when it is non-human: few listeners will identify with chicken. But the dangers is acute when the original is a general (human) category like 'the' Leiermann, or imaginary human beings like Arianna: it is impossible to assess a difference between the original sound source (the voice of the original singer or musician) and that of the imitating singers. The fellow singers or players feel tempted to experience their singing or playing as their own expression. Also here, an elliptical reading threatens to obliterate the difference between non-mimetic and mimetic music, to the effect that also mimetic music is experienced as the auditory appearance (the 'expression') of the singer or the musician (or the composer).
With image conjuring music, the world of the conjured images can equally be supplemented with a second layer in the real world. Precisely because we are often dealing here with music with a strong metrical character (think of Different Trains of Steve Reich), we cannot resist the temptation to move the movement conjuring signs - for instance by moving the arms in an effort to imitate the connecting rods of the train. Also here, the imaginary community of all those who deem themselves a train is supplemented with the community of all those who imitate trains through the movements of their arms: here, the mimetic character of the music is obfuscated through a shift to visual mimesis (like with ballet): the reduction of image conjuring sings to movement conjuring signs. Where that is not possible - think of a concert hall - there is always the possibility of identifying with the musician who produces the movement conjuring signs: we feel like a kind of magician who conjures up the images through playing the music.
In all these cases, the idea of a mimetic dimension threatens to disappear: with absolute music, musical space threatens to collapse; with the musical imitation of musical appearances the difference between a singer and a musician and their imitations threatens to be overlooked; with image conjuring music the reduction of image conjuring signs to movement conjuring signs for unmediated visual mimesis threatens to obfuscate the idea that there is also something like image conjuring music. Thus, the way is paved for a further reduction of the idea of music: music as 'expression' - after the example of speech music - where also dance music, which is not expressive, seems to be excluded from the realm of music.
PROTEUS (1): POLYINTERPRETABILITY AND CONDENSATION OF DIVERSE KINDS OF MIMETIC MUSIC
Up to know, we concentrated on discerning the diverse kinds of music: non-mimetic music versus mimetic music, which can be unmediated or mediated. Time has come to remind of the fact that in real music often diverse kinds of music are combined or condensed, or that one and the same music can be read in alternative ways altogether.
In 'sonorous beings in musical space' we already pointed to the fact that all speech and dance music can also be read as absolute music, and that, conversely, many absolute music can also be read as speech or dance music. Here, we only have to add that the possibilities of double reading only increase when we take into account the full array of kinds of mimetic music.
To begin with, it is apparent that not only non-mimetic speech and dance music (the virtuoso play music included) can be read as absolute music, but also the non-absolute variants of mimetic music, especially the imitations of vocal or instrumental singers or of musicalised beings. Conversely, how eager the sonorous beings of absolute music may be to leave their vicarious sound sources behind in the real world, it may be more or less tempting to continue to hear them as singing voices, and, in the wake thereof, to read the accompaniment as a movement conjuring sign for those singing beings (which are then often visually represented in our minds). Thus, Fratres can be heard as homophonous moving sounds, but also as the auditory imitation of a procession of singing monks. Or when pizzicati are added, the reading of overture to the third act of the Traviata suddenly invites to read the up to then absolute music as a singing being that moves expressively.
Next, we can read non-mimetic as well as (unmediated) mimetic music as image conjuring (and hence mediated) music. Any dance music - especially when it is not so easily performable by the human body - and any absolute music - especially if it reminds of the movements of non-human beings - may conjure up visual representations of the movements of existing or imaginary beings - think of the Finale presto from the second piano sonata of Chopin, where Anton Rubinstein heard 'Winds of night sweeping over churchyard graves', Tausig 'the ghost of the departed wandering about" and Cortot' the freezing whirlwind descending on tombs.' Perhaps even more important is that, as we have seen above, conversely, many an image conjuring music may invite the listener to identify with what is represented and to perform (a part of) its movements with his human body ('Saint Francis walking on the waves' of Liszt, or 'Different trains' of Steve Reich), if we do not read it as absolute music from the beginning - as is the case with most 'program music' - from Liszt's 'Après une lecture de Dante' to Leroy's' 'typewriter song'.
It is evident, however, that not every music can be read as one pleases. There are, to begin with, the many cases of unambiguously one-sided music, where a specialised reading is the most appropriate. With many speech music, the melodic line does not amount to more than tonal structuring, so that it does not invite to a reading as moving sound. That holds even more of most dance music, where it is only the added speech music that may invite to a reading as moving sound. A reading of image conjuring music as absolute music is only appropriate when it is interesting as absolute music. Often, the signs are of little interest (see 'Different trains' of Steve Reich), so that an attempt at reading them as absolute music is soon given up. That is the reason why many of the more refined program music also consists of vocal or instrumental speech music, not to mention the often important contribution of absolute music (see below: combinations).
But, in other cases, a multiple reading seems to be deliberately intended, so that it is not only impossible to determine which reading is most appropriate, but superfluous, since only a double reading gives full account of the richness of the music. That is the case with many a melody that can be read either as the movement of a sonorous being or as (imitative) speech music (think of the prayer in the Offertorium and Hostias of Berlioz Requiem or of Agurida, where you do not know whether a singer is cursing or a sonorous being is floating through the air), as instrumental, speechless speech music (think of the melody of the Adagietto that seems to be sung by a kind of 'World Soul'), or, finally, as movement conjuring sign for movements of the entire body or of expressive gestures, that are however not executed, and hence lead to the visual representation of beings that make the expressive gestures or that are appropriate to the auditory expression of the corollary speech music (think of the melody of the Mondscheinsonate, the rudimentary melody of the Offertorium and the Hostias from the Requiem of Berlioz, or of the melody of the funeral march of Chopin). An impressive example of a perfect condensation of image conjuring music and absolute music is the 'Vorspiel' to 'Das Rheingold', where the reading as absolute music is even more convincing than the reading as image conjuring music. In all these cases, unequivocal lecture is impossible.
Double reading is facilitated when the composer has the same melody resound first with words and then without, just by having it played by an instrument. When a melody is sung, the temptation to read it as mimetic speech music is stronger, whereas when it is played on an instrument, we feel rather inclined to read it as moving sound. Such alternate reading is exemplary in the Liebestod' in Tristan and Isolde, that is in fact a kind of duet, if not a trio or quartet: the melodies are now sung, now played, so that now the emphasis is on the movement of sonorous beings, then on the expressive speech music. The result is that in both cases there is a double reading of the melody: the instruments seem to sing and the voices sound like a moving sound. Another impressive example is the singing choir that is embedded in a 'singing' orchestra' in Berlioz' Hostias, or the melodies that accompany the song the Midnight in the Mitternachtslied of Mahler, or the falling melodies at the end of his 'Der Abschied'. In how far speech music can also be read as absolute music (as moving sound) can be gauged from the degree to which it is still heard as a song when played instrumentally. The Liebestod of Wagner does not collapse in a purely orchestral version, whereas Schubert's Erlkönig loses much of its power in Liszt's piano transcription. The condensation of absolute music with mimetic speech and dance music implies a reduplication of the singer: on the one hand, he is a singer in real space, on the other a sonorous soul that is moving in musical space. As long as we are only listening, that conflict is not manifest. But when we are dealing with an audiovisual imitation, there is often an overt contradiction: thus, the visual Tristan and Isolde are locked in each other's embrace, whereas their souls (together with those of other singers and a singing orchestra) are moving on love waves. Thus, in 'L'ho perduta', the visual Barbarina is searching on the ground, while her soul is making movements that are more adequately rendered in the Taviani film 'Kaos'. Such redoubling is so convincing, because it is a reflection of the divide between the outer appearance of the body and the inner soul.
Another technique that favours double reading is polyphony: each voice sounds as speech music, but in that it begins to move independently from other voices, it is soon also read as sound moving in musical space. The often astounding plenitude is not so much an effect of the course of the singles voices, as rather of the expressiveness and the structuring dynamics of the tonal relations - the chords through which the voices are moving. Thus originates the often overwhelming impression of a community of like-minded singers, who at the same time are floating in the higher regions of musical space (Hymn of the Cherubim of Tchaikovsky), or the overwhelming impression of voices whose individual mourning only unfolds to a collective lament when the voices are joined in a dizzying spatial architecture of sonorous beings that move through various expressive positions in musical space (Hear my prayer o Lord! of Purcell).
Let us remark that such double reading is all the more fruitful - or: is only possible - when we are not seduced to an elliptic reading, that would obfuscate the richness of the double phenomenon. Let us equally remark that in all these cases the ascent of musical space is no longer manifest - but therefore no less effective.
PROTEUS (2): ADDITIVE COMBINATION OF DIVERSE KINDS OF MIMETIC MUSIC
Next to the effective condensation that necessitates a simultaneous double reading, there is also the possibility to combine two or more one-sided kinds of music additively. The addition may be simultaneous (think of a mimetic speech melody over a movement conjuring accompaniment), or successive (think of the prisoners singing against a background of absolute music in 'A survivor in Warsaw').
Next to the combination of auditory imitation of existing musical beings (like a cuckoo) or of musicalised beings (like the bumblebee) or (existing or imaginary) singers or musicians with sonorous beings, also the combination of unmediated with mediated musical mimesis is very popular. In the third part of Pini di Roma, we have a combination of the recording of bird song with image conjuring music. In 'Different Trains' of Steve Reich, recordings of train whistles are combined with image conjuring signs for the movement of connecting rods. Far more obvious, however, is the combination of image conjuring signs with the imitation of imaginary instrumental singers and/or sonorous beings. That this combination is so popular, is due to the fact that the movements of existing objects may well be expressive, but do not necessarily tell something substantial of what is imitated. That is why many image conjuring music is often supplemented with spoken '(Different Trains) or sung text, but also with absolute music or mimetic speech music: think of the melody that is added to the image conjuring signs for walking elephants in Saint-Saens, or of the melody that is added to Honegger's Pacific 231.
A special case of such additive combination is the imitation of singers or of musicians who sing or play absolute music (double mimesis),
PROTEUS (3): COMBINATION OF MIMETIC MUSIC WITH NON-MIMETIC SPEECH MUSIC OR DANCE MUSIC
Music can also consist of a combination of mimetic and non-mimetic music, simultaneous or successive.
A first standard formula is the combination of absolute music or mimetic speech music with dance music through adding a movement conjuring accompaniment to a melody that can be read as absolute music and/or as speech music's. The 'accompaniment' is then an exherent movement conjuring sign for the movement of one or several melodies, as well as for the movements (expressive movements or dance movements) of the listener. A striking example is the beginning of the already mentioned prelude tot the 3th act of La Traviata, where the melody is first to be heard without accompaniment, until the accompaniment is added, that consists unambiguously of movement conjuring signs for the sonorous being or for the instrumental song, as well as for the expressive movements of the listener. A similar analysis applies to the Adagio from the string quintet of Schubert.
The combination can also grow from the fusion of dance music with speech music that we already analysed in 'movement conjuring signs' in that the speech music unfolds to mimetic speech music of an imaginary singer or musician. That is the case when the singer stages himself or another character. The accompaniment continues to function as real dance music, not only for the singer, but also for the dancers in the public. Such transition to mimetic speech music leads to the popular formula of the lyric song where the singer is also the imaginary singer, or in classical music were the singer does not coincide with the imaginary (vocal or instrumental) singer: just think of a standard aria or of Chopin's funeral march.
Also image conjuring music can be combined with dance music, as when 'Different Trains' would also conjure up a kind of mimetic ballet.
Let us remind that the listeners mostly fail to notice that there is a combination of non-mimetic and mimetic music, so that the combination is experience as just 'music', whereby it is assumed that the music is non-mimetic.
PROTEUS (4): NON-COMPATIBLE READINGS
With condensation and additive combination, we get a composite original.
It should not escape our attention that the originals that appear with differential reading are not always compatible like those of the Liebestod or of L'ho perduta. Especially with absolute reading of image conjuring signs, there is an often considerable difference between the originals that are conjured up by the image conjuring signs, and the readings that appear with absolute reading. That is not only due to the fact that most movement conjuring signs are analog, whereas in absolute music they are heard literally. Thus - to give a graphic example - the up and down of the tones in 'L'Aquarium' of Saint-Saens is no longer interpreted as the back and forth of the tails of the fish, but as un up and down of a movement in musical space. Other differences are due to the fact that the movements that are conjured up by image conjuring signs are movements of visual bodies in a three-dimensional space, whereas the movements of sounds in musical space are movements of monosensorial sonorous beings. Thus, the tones that, in Schubert's 'Die Krähe', are conjuring up the movement of the wings perform totally different movements as sonorous beings. When read as absolute music, the sustained notes that have to conjure up the horizontal plane of endless plains of Borodin's 'The steppes of Central-Asia' are floating over the fathomless depths of musical space. Or the tones that in Liszt's 'Fontane di Roma' conjure up the image of fountains, also embody sonorous beings with buoyant energy in musical space. When read as absolute music, Tchaikovsky's sugar plum fairy is no longer a fragile being. As absolute music, Leroy's 'Typewriter song' has nothing to do with typing, but rather with frolicking beings in musical space.
In many cases, the double reading cannot be sustained: there is only talk of a local break-through: many virtuoso play music often dissolves in non-mimetic musical formations that only make sense as demonstrations of the skill of the musician.
Whether double reading is possible - and hence: whether a complex, but coherent original is created - depends on the way in which the diverse readings lend themselves to combination or condensation, as is usual in the best speech music, be it vocal or instrumental.
In 'sonorous beings in musical space' we pointed to the fact that absolute music can be read as speech music or dance music, and the other way round, but that there is a preferential kind of reading for each piece of music (double reading, as with the Liebestod, included). Here, we have only to add that also many image conjuring music can be read as absolute music (of more generally: as unmediated musical mimesis) and the other way round. When double reading leads to incompatible originals, the preference is for 'one-sided' music for single reading.
AUDITORY MIMESIS AND AUDITIORY DESIGN
In 'Image conjuring signs', we demonstrated that (non-mimetic) speech and dance music belong to the domain of musical design. In this text, we have to add that sounds in general may be used for non-mimetic purposes as well: the domain of auditory design in general and of musical design in particular.
There is, to begin with, the design of the countless musical signals where tones are not used as movementconjuring signs: bells, gongs, sirens, ringtones, horns, train whistles, military signals and the domain of 'sonifications': 'translations' of all kind of data in sounds or tones like in the sonification of PI.
More important is the design of two kinds of (non-mimetic) soundscapes: the purely 'temporal soundscape', where the occurrence of sounds is organised in the temporal dimension: think of John Cage's 'Imaginary Landscapes I-V" (1939-1952), Wolf Vostell's "Fluxus-Symphonie für 50 Hover-Staubsauger' (1960) or Ligeti's Poème symphonique' for 100 metronomes (1962). Let us also mention the countless instruments played interactively by the public (think of the Baschet brothers) or by some transformatory device. Next to the temporal soundscap, there is also the (non-mimetic) spatiotemporal soundscape, where the occurrence of sounds is not only organised in time, but in space as well: think of compositions like Godfried-Willem Raes" Symphony for singing bicycles (1980) or of Dirk Veulemans' 'Composition for eight fire engines' (2006).
Whoever forgets that the name 'music' is applicable to such different kinds of music like speech and dance music, mimetic music (among which: absolute music) and image conjuring music, is all too easily tempted to overlook the existence of mimetic music and to understand it in terms of image conjuring music, but above all in terms of speech music and dance music. Especially the interpretation in terms of speech music is widespread: it lies at the roots of attempts at understanding the 'non-abstract' character of absolute music in terms of expression rather than in terms of mimesis (music as a 'language of feelings'). That is a particular variant of the general shift of the mimetic to the semiotic model, endemic since Croce. On the other hand, the rejection of image conjuring music - the 'literary' program music - is responsible for the quasi ineradicable misunderstanding that music would not be 'representative' at all, but 'abstract' (Hanslick) - a view that is the complete opposite of the semiotic interpretation of music. In both cases, it is denied that music would have something to do with mimesis.
Only the mimetic theory is able to make a clear distinction between all the kinds of music and to sharpen the ear to the often far reaching differences between them. And it is only when we are able to make the appropriate distinctions, that we realise that most music consists of often complex combinations or condensations of different kinds of music.
The existence of often complex composite originals should prevent us once and for all from understanding mimesis in terms of the non-moving, single and visual image, that uses to function as the paradigm of the image.
© Stefan Beyst, spring 2012
Share on Facebook
Your reaction: mail to stefan beyst
Stay informed about new texts: mailinglist
Background to this text: stefan beyst: theory on art