at Pontllyfni (Wales), 11 April 1998
Tradução portuguêsa publicida ulteriormente em Debates (Uni-Rio), 1999
'Immanent meaning in music' | 'Absolute' music? | Semiotic analysis of music | 'Primary' and 'secondary' signification | Generalising about musical signification | Etic-Emic in music analysis | General questions again
What do you think about Eco’s statement in The Absent Structure (1972: 16): ‘music raises the problem of a semiological system with no semantic depth’; and Ruwet’s notion in ‘Methode d’analyse en musicologie’ (1972): ‘since the meaning of a musical work is immanent…, the only way to approach the study of its meaning is through the formal study of musical “syntax” and a description of the music’s material aspect at all the required levels’?
Eco has said several times, especially to one of his students (Luca Marconi, who was writing about Mersenne), that he really didn’t want to comment on music. Probably, Eco — and this is just speculation — had been asked by musicological colleagues to talk about something he doesn’t feel too much at home with. Eco is quite honest, you see. Anyhow, it all depends on what he means by ‘semantic’. If he means signification of the lexical type, then I think it’s probably true, but if he has a broader meaning in the French theoretical sense of sémantique, in the sense of studying signification in general rather than signification restricted to lexical meaning, then I would disagree.
When it comes to Ruwet, on the other hand, I think you have to remember that he is an avant-garde Belgian composer steeped in the art music tradition. Like Nattiez, he has, I think, quite restricted frames of reference: he can’t really see the various levels of signification that are contained within the musical text or musical discourse.
I would disagree very strongly with Ruwet’s statement about the immanent meaning of music. That is what I would call a typically conventional formalist approach to the understanding of music. This kind of approach is at the basis of what passes for music analysis in many institutions, where formal structuration, like the horizontal narrative of the music — the ‘form’ — is determined by harmonic centres, by notions of that kind and apparently by not much else. Of course, these aspects form the centre of analysis and of the dynamic of certain types of music. They are central to sonata form, certain types of rondo, certainly some types of dodecaphonic music, and syntax does play an important part even in the average pop song. But there are so many other principles of musical construction that are important in the pop song or in film music, principles that cannot be related to syntax alone.
In fact I don’t really think that syntax alone can account for signification even in the sonata form. As Charles Rosen points out in his The Classical Style, when classical music crossed the road from the opera house into the concert hall, there was a whole range of already existing relationships between musical structure and emotion or movement that also crossed with the music. So the theory of affects and those relationships between musical structure and, if you like, affect or movement or feeling, went along too as part and parcel of the whole style.
When the Viennese classicists constructed their sonata form movements, they did so, I’m sure, in the knowledge that individual themes, their harmonies and orchestrations, their phrasing, etc., had certain affects. Music in sonata form would have no meaning without the individual affects and varying affects of its constituent themes. There would be no point to it because there would be no development, no narrative, no syntax etc. without contrasting (or similar) elements which constitute the actual building blocks of the composition.
It seems to me quite obvious, historically, that if Bach died in 1750 and if Mozart is writing string quartets thirty years later, and if they were all learning harmony and counterpoint, as indeed they were, then it would be very unlikely for affect to disappear in such a short time. It would be like saying, in 1998, that music created in 1968 no longer had any signification today similar to what it had then. The structures have maybe changed their connotative values to some extent, but it seems historically quite illogical to presume that in those thirty years something so basic to a music culture can be wiped out. It’s more on the ideological level that I think you need to see the understanding of music’s meaning as immanent in formal structure and nothing else. It all relates to notions of bourgeois individualism.
Diderot, in Le neveu de Rameau, parodies the musician who is very critical of the Baroque theory of affects. ‘Point de ces jolies pensées (‘none of those clever pretty conceits’), he says, ‘il faut que la passion du musicien soit absolue’, or words to that effect. Basically, Diderot has put words into the musician’s mouth, saying ‘we don’t want anything in music being linked to anything outside music; it’s all expression in and of itself’. I think this attitude relates very clearly to the separation of human nature from natural science in the way that Rousseau speaks of it. Clear symptoms of this split coincide with the rise of the bourgeoisie: it becomes necessary for a socialisation pattern to be put into place, not as a conspiracy but as a logical consequence of having to organise the world in an irrational way that is called ‘rational’. This ‘rationality’ has to be irrational because it includes completely mad things like the rape of nature, invasion and genocide to establish colonies, telling lies to make money, etc. This madness comes about not because people are intrinsically mad or evil but because capitalism demands of the bourgeois person, the merchant, that you either become part of the system you belong to or you perish by its hand. So the spheres of sexuality, emotion and ethics had to be separated from ‘logical’ thought: they might otherwise interfere with that irrational ‘rationality’. In this way you get the split between natural science and human nature, and it becomes convenient to have a whole arsenal of terms and cultural practices which make clear distinction between the material and the emotional, between the rational and the corporeal and so on.
To cut this long story short, I think Ruwet is unconsciously representing this tradition of asceticism and false rationalism from the days of the bourgeois revolution. It has stayed with us from 1789 for 200 years without being challenged on a broad front inside musicology. If you consider that it took such a long time for anything rational to be officially accepted about the ‘irrational’ — I mean, there’s the whole of the nineteenth century before you get to Freud or anybody who tries to be officially rational about feelings or sexuality — it’s not surprising that the split brain still turns up today in notions of ‘absolute’ music or ‘immanent’ meaning. Anyhow, this would be my general explanation for the ideas put forward by people like Ruwet or Nattiez.
But then again, these guys are socialised into that system and they are going to have to hold down and be successful in jobs in universities within the same system. The only problem these days is that if you have to deal with people, like myself and many others along with me, who were brought up in the in-between situation of being involved in other types of music like pop, jazz or folk music, or ethnomusicological work or anthropological work, as well as in the classical as it was taught to us, then it’s clear that a different sort of socialisation comes into play. Obviously, in our musical practices we see that music cannot work only according to the norms we inherited from bourgeois ideology. It just does not work! I mean, even if you’re playing hymns in church, you know the old ladies will cry if you put certain harmonies in and people will get happy if you put in others. If you go to play for a dance and if you put down the rhythm in one way people are more likely to get up on the dance floor than if you play it another way. And then of course, if you’re working in film or TV music, or music for theatre, you have to follow the genre or the type of narrative with music and highlight an emotional or kinetic or tactile dimension that cannot be provided for by the visuals. In all these cases it’s quite clear that there is regularity in the choice of structures that composers and musicians use to create those emotional, kinetic and tactile dimensions. If there is regularity and recurrence of musical structure related to phenomena that aren’t necessarily musical, then you can talk about another type of signification than the one Ruwet is referring to, if he’s actually referring to any real signification at all.
So, there are two things: (i) there’s the historical legacy which Ruwet & Co. come out of and which they don’t seem to have questioned very much at all; and (ii) there’s a whole range of musical practices that demand that you view music and its role as a symbolic system in a totally different way.
Do you think there is such a thing as ‘absolute music’?
No. I think music always means something besides itself. Firstly, by virtue of being ‘absolute music’, or rather by being regarded by a certain group of people as ‘absolute’, music cannot logically be absolute because the nomenclature assumes that ‘absolute music’ is different from music that is ‘not absolute’. There are people who think that music of type a is absolute and those who are considered incapable of understanding or appreciating those ‘absolute’ values. Therefore, there is a social value, a social connotation connected with ‘absolute music’. I think the clearest example comes from Hegel who said in his Aesthetics that the layman likes music which has lyrics or action associated with it, some clear message, whereas the connoisseur likes music that is in and of itself. ‘And’, Hegel goes on, the connoisseur ‘can be fulfilled by the music on its own’. So, Hegel is making social distinctions in this area of music. By so doing, semiotic distinctions are also made because there are social meanings invested in the type of music being referred to. There’s nothing very absolute about that.
Even so, inside that ‘absolute’ music there are obvious gestures or types of movement, or tactility, or texture, that can have what you might call an anaphonic value: there are episodic markers — crescendos, changes of key, changes of instrumentation etc. — that can give direction and that have a kinetic character and which may be difficult to explain in words, but which are not themselves ‘the music’ even though they are relatable to the music by a given population. So, even inside ‘absolute music’ I find it impossible to conceive of absolute music.
Merriam brings up the concept of uses and functions, most of them related to European art music and aesthetics. One of the functions he mentions is transcendence. It could be that people look for music to dance to or for music they might want just to listen to as some sort of contemplation. Don’t you think that’s possible?
Certainly. Actually, I’m not an expert on meditative techniques but the little I know about it suggests that certain bodily postures are more conducive to meditation or contemplation than others. I mean, it is virtually impossible to contemplate to very loud and fast techno music, whereas it would be very difficult to dance in a brash, fierce or fast way to music that is contemplative. But at the same time I don’t see how contemplation or meditation, transcendental or not, can in fact transcend the society which that act of meditation takes place in, or how it can be suprasocial in any other way. Meditating in order to subjectively transcend everyday concerns means you are doing it in relation to whatever it is you want to transcend; so the act is invested with social meaning because of that negative relationship, and music conducive to meditation or contemplation must also be invested with meaning.
I suppose the reason for these views coming forward (I mean views which in one way or another emphasise the ‘absolute’ or ‘transcendental’) is that our music education is very largely based on the production of performers of other people’s music, rather than on communicating your own music or on understanding your own and other people’s reactions to music. So, the verbal concepts we construct about music are those which may be useful to musicians trained technically in a certain tradition. And this will mean that we are not thinking in terms of what listeners might do with the music so much as how we are going to make people (performers) make certain noises.
You have been describing what Bourdieu calls habitus, which does not just mean a person’s habits or cultural influences, but also his or her subjective intentions. When you were talking about the social and academic contexts those persons were in, and then about your own generation and the other historical aspects that came to bear upon it, you then switched to talk about music education and the goals of that music education: how it aims at training not listeners but performers.
Yes. It’s production training that is very technically orientated. It’s not production even with a view to affect people in order to dance or in order to feel in a certain way. It’s the learning the techniques of producing noises, of producing sounds that are, often uncritically, considered to conform more or less to how certain types of music ought to sound, all according to quite often dubious norms. I feel that with that type of concentration it becomes very difficult to construct any useful type of semiotic view of music.
I suppose a parallel would be looking at architecture and saying to architects something like ‘here are the technicalities for building a house and here are certain aesthetic rules to follow’ and forgetting to mention that people might want to cook, eat, go to sleep, wash, bring up children, make love, relax, be quiet, make a big noise, or in any way actually use the building as a home. No good architectural school would, I hope, teach their students to ignore the building’s function and meanings for those who are going to use it. But our music education system often does. They don’t ask what this theme or that technique is going to be used for; they don’t even seem to bother that we’re overproducing performers for the classical market, nor that we’re in the process of overproducing performers for the jazz or pop sphere, at least, not in Northern Europe. But this problem is related to a historical tradition of musicians and composers being socialised as an odd bunch that cannot really communicate with anybody else apart from each other. Musicians often perpetuate that guild mentality of secrecy and mysticism in their own trade — ‘hey, man, it’s a G minor demented seventeenth!’ I suppose this is all understandable because all but a few musicians are underpaid and badly treated. So it’s a vicious circle.
How does the semiotic analysis of music fit into all this, then?
It can be really useful for two kinds of people: what you might call, for want of better terms, ‘musicians’ and ‘non-musicians’.
With non-musicians I think it’s a question of the general democratic right to education about messages in the media, that is to say the right to know how music mediates attitudes and how those attitudes — which have a kinetic, gestural, tactile and corporeal character in sound — how they relate to other aspects and other types of symbols concurring with music, like in film, in songs, operas, musicals, ballets and so on. So, people should be given a vocabulary which they can relate to certain sounds they recognise, not necessarily in any structurally correct manner, but at least in a phenomenologically coherent manner. For example, if you have a class of students and they all agree that a certain sound that turns up at a particular point in a particular piece ‘means’, say, snowflakes, or little drops of ice or tingling or something, then you have coherent intersubjective agreement, and that is a valid social science criterion for a certain type of objectivity.
If such agreement exists, then it can be related to unique occurrences of musical structure by simply saying, for example, that at 1 minute and 10 seconds into this tune or at 15 seconds into that advertising jingle, or whatever, you have this ‘tingly ice-or-snowflake’ type of sound, then an intersubjectively valid observation has been made in the field of music semiotics. The ability of non-musicians to pinpoint things of this kind, if they want, and to have some kind of positive control over the music’s effects on them, not in order to stop those effects but in order to reflect whether they agree with them or not, I think is really important.
From the musician’s viewpoint I think such semiotic analysis can have many functions. First of all it can demystify for the musician his or her attitude to the trade of being a musician: you can see yourself as being actually involved in the normal business of communicating to normal people rather than engaged in some kind of technical conservatory-style mystical mumbo-jumbo activity. Secondly, it can highlight stereotypical relationships between musical structure and movement, touch, etc. and allow the composer or creative performer either to avoid or to consider new ways of thinking about or dealing with a particular choice of what to do musically.
So, understanding what usually happens in terms of relationships between musical structure and affect, or movement, or touch, or bodily experience etc., by understanding all these usual or stereotypical relationships, you are, as a musician, in a much better position do something usual or unusual, whichever you want. I mean, knowing these things allows you greater freedom in the long run than just doing it by ‘feeling, man!’. If there’s anything stereotypical, ‘man’, it’s 95% of jazz improvisation, ‘man!’ because ‘just feel’ is far more likely to give rise to stereotypes than the ability to both feel it and think about it. Actually knowing what’s going on allows you to do something different with it.
Another important spin-off from semiotic music analysis is that knowing about the archetypes of our own culture can make people more interested in how other types of music work in creating moods and meanings. It also helps in creating an interest amongst musicians for ethnomusicology, for the anthropology of music, not to mention the sociology and psychology of music; because in order to explain how a musical structure relates to whatever movements or gestures or feeling you’re talking about, you have to be quite interdisciplinary: you have to see it historically, in psychological terms, neurological, anthropological and, of course sociological terms. So, I think semiotic analysis can broaden the musician’s outlook a lot and help further creativity.
So you mean that the role of music semiotics in music education has two points: for the non-musician it provides a critical listening position, and for musicians the same and more, by providing the opportunity to break away from the stereotypical relationships found in the trade and by exposing musical practices to a dialectical process that allows musicians to expand their expressive potential.
I wish I’d said that!
I want to ask you next about your views on primary and secondary signification.
I disagree, but not violently, with Richard Middleton on this one. I think ‘primary’ signification can easily become a kind of Holy Grail, a bit of a mirage.
What is primary signification?
I’m not at all sure. It’s obviously a level of musical ‘meaning’. I think it seems to be a kind of common denominator of gesturality, corporeality, texturality and kinetics in the type of sound that music can deal with, that is to say the meaningful relationship between a musical structure, or set of structures, and this common denominator of gesture, movement, tactility, etc. This ‘primary’ signification is always going to be exceedingly vague in verbal terms. Since we have to deal in most educational establishments with words as the metalanguage for music, ‘primary’ signification is going to be a very difficult matter because it is not easily reached by or contained in words. In fact, it might be fair to say that music wouldn’t need to exist if music did not include this type of signification. So, I’m not suggesting that this type of signification does not exist, on the contrary, but what I would suggest is that talking primarily about ‘primary’ signification can all too easily revert to an essentialisation of music as suprasocial, absolute or universal. Because of our inability to verbally explicate that type of signification, music can so easily seem like something extremely vague, and hence impossible to relate to anything outside itself. This just compounds the problems of having to deal with the tradition of ‘absolute’ music we just talked about. After all, it’s still a verbal tradition that rules the roost in our education system.
I would hold quite strongly that it is methodologically necessary not to start with ‘primary’ signification but to start with ‘secondary’ signification — and these are Richard Middleton’s, not my terms — if we want to discover, more anthropologically in a way, and historically, how music has been related to far more easily verbalisable and particular phenomena than can be expressed in terms of ‘primary’ signification. After doing that it will then be easier to see, as a second step, how the various connections that can be made more easily at a level of ‘secondary’ signification’ can have certain common denominators that are tactile, kinetic, corporeal, textural and so on. This means it would be more efficient to construct an understanding of the ‘primary’ on the basis of ‘secondary’ signification, all of which makes the nomenclature of the terms a bit questionable. I think this order of carrying out the semiotic analysis of music makes more sense because it makes the whole intellectual exercise far more concrete, far more socially related. Since individual links of ‘secondary’ signification are so clearly rooted in a particular culture and in a particular population — and you can see different semiosis through different age groups at different periods in time in relation to the same music —, there’s far less risk of becoming essentialist or Eurocentric or absolutist about music, and about how it works. If you proceed in the way I just tried to describe it’s possible to approach the $64,000 question of ‘primary’ signification because you don’t have that problem with a lack of social rooting: you cannot even see ‘primary’ signification as unrelated to particular social and historical conditions.
Methodologically it’s much sounder to start from the small and from particular musical cultural situations to see what the links are between music and something ‘outside’ the music and then to draw the conclusions down to another level of abstraction. I don’t agree with trying to go for ‘primary’ signification like a hole-in-one on the golf course: you mostly have to hit the ball several times before you can even get it on to the green.
Are you saying that people concentrating on primary signification are more likely to miss relevant parameters of musical expression, you know, just concentrate on things like pitch and tonality and forget the rest?
I’m not suggesting that anyone interested in questions of ‘primary’ signification is automatically going to do that. I’m just saying that the risk of serious error is far greater if you haven’t looked at the details of semiotic and cultural practices in real terms, in empirical terms, at the level of ‘secondary’ signification.
You are saying that we should start from secondary signification, from the particular, from the reception and individual responses to music. If this is so, how can it ever be possible to make generalisations upon which you could build hypotheses and theories? If you are so much concerned with individual responses, it might appear that you are dealing with mere idiosyncrasies related to specific age groups, time frames, gender, etc.
Perhaps I’m a coward, but I don’t think that any one person could possibly construct an overriding theory about how music communicates whatever it does at all times in all places. Nobody I know has that kind of omniscience and I don’t see how anybody could ever have it. But let me try and put this another way that may be more constructive.
First of all, I don’t like the terms ‘primary’ and ‘secondary’ signification because they suggest a hierarchy where no hierarchical categories are appropriate. I mean, ‘primary’ should either mean coming first, i.e. that it comes before ‘secondary’ signification when music is communicating something, or it ought to mean that it comes before ‘secondary’ signification in a metaphorical sense, i.e. that it is more important. I think that neither of these is true. I don’t agree with the argument that the type of relationship between, say, musical structure and movement or touch is necessarily more immediate or more important, i.e. more ‘primary’, than a relationship of social connotation because that social connotation might be the prime signified of a particular musical structure and override, or even play havoc with, other aspects of musical signification going on at the same time. So, if, for example, you were to immediately associate a piece of avant-garde music with a load of stupid intellectuals who think they’re God — a social connotation — that would stop you from deriving any tactile or corporeal enjoyment out of that music, if there is any to be had. It’s for this kind of reason that I think it’s very confusing to talk about ‘primary’ and ‘secondary’ signification.
How does this relate to the need for making some type of generalisation?
Let’s take an example. Here we are sitting by this little river, so similar to the little brooks in the Wienerwald that Schubert was trying to stylise in his Die schöne Müllerin. If you look at watery, wavy forms outside music, you notice that waves on the sea or on the beach have a longer period than the individually observable bits of water you see on a small river flowing fast. But waves on the sea have lot in common with the pattern the wind makes as it blows over a cornfield or through longish grass or in the tops of trees of a wood. In relation to hearing piano arpeggios played legato with the sostenuto pedal down in semiquavers over a two-bar period, or in slow tempo as sextuplet semiquavers over a one-bar period, I can tell you that in an empirical reception test 607 people put down very few brooks, a lot of sea a lot of wind in the long grass of a meadow, quite a few trees and a few cornfields, much less lakes and no lullabies, even though lullabies have a similar wave period and intervallic ambitus as the sea. The only difference is that lullabies have no little droplets, no semiquavers, just legato quavers or crotchets.
If you look at the relationship between lullabies and sea waves, they have a lot in common. All of the music that contains this kind of long-wave legato arpeggio, that is apart from the lullabies which don’t have the ripples or little droplets, is associated with romance and love as well as with sea and long grass. Of course, Chopin nocturnes, piano concertos by Tchaikovsky or Rachmaninov and clones of this style — things like the Warsaw Concerto, the Spellbound Concerto and so on — are all examples of romantic music, at least in our culture.
Incidentally, some people also mentioned shampoo, but this has to do with the shampoo advert for Timothei that was popular at the time: it features a young woman with long hair in a long dress walking through the long grass of a summer meadow. All of these things — the sea, the fields, the long grass, the long hair, the long dress — have one obvious thing in common: it’s the wavy movement with your hand or your fingertips; it’s a gesture. Logically speaking, the cornfield hasn’t got much to do with someone’s hair, and the sea has nothing much to do with trees really. But the movement is something you can either internalise from the outside or project out from your body over how the waves move on the sea, or you can trace the pattern of the hair or of the dress if you make the same kind of wavy gesture downwards instead of horizontally.
So there is a type of gesturality that has to do with love and tenderness and with the rise and fall of passion and so on, plus the fact that the movement contains tinglings or droplets — the semiquaver aspect. It is interesting that the gesturality of a lullaby sending a small child to sleep has the same kind of legato flow and movement but contains no droplets or ‘tinglings’, which also means no titillation. This might be a type of socialisation presenting the incestual taboo in a musical form. Who knows? But it’s quite clear that people just don’t hear lullabies when they hear romantic music of the legato piano arpeggio type, even though the wave forms are so similar.
All I’m really trying to say is that there’s a common denominator between all these wave forms: it’s a gesturality, a kinesis, a tactility that can be distilled out of these seemingly disparate physical phenomena. The emotional aspects associated through music to all these phenomena have to do with tenderness and passion, but other relevant aspects are countryside, pastoral nature, the sea, and specific, socially formed attitudes towards those things. So, to reduce the musical signification of the legato piano arpeggio in the romantic vein to its gesturality — its ‘primary’ signification maybe? — would not only be false: it also takes all the must, all the guts out of music’s ability to communicate and demotes it to an extremely abstract and uninteresting type of communication system. It makes music socially meaningless, which I don’t think is correct.
Still, I suppose this is one way in which you could distil and abstract certain types of movement to a high level of generality. But this doesn’t mean to say, in a completely different musical culture, which might have a completely different notion of how a parent should behave towards a child or which has entirely different notions of what love is supposed to be like, that music expressing accepted gestural patterns of behaviour towards those phenomena is going to sound like music for the same sort of thing in our culture, nor that our music for those functions in our culture is going to mean the same to people in that other culture as it does to most of us. So, I think it’s also important to avoid the essentialisation that promoting ‘primary’ signification could bring about. But I think you can certainly look at one musical culture and compare its patterns of gesturality and movement with those of another musical cultures to see if their musical representation of gesturality bears any relation to ours, and to see if that gesturality is related to similar phenomena as it is in ours or not. I think it has to be culturally relative all the time. But a certain degree of abstraction ought to be possible within one large musical culture, I would think.
You start from secondary signification and then go to primary. Brooks, waves, hair and so on lead to the primary gesture. You are making the argument that we should consider doing musicology like ethnomusicology, advocating that we don’t essentialise our notions of music, our common-sense notions, and make them universally applicable, that we should question our positions that are ingrained and which we take for granted, doing the same kind of work that Simha Aaron was doing in Africa, but doing it on ourselves. From what you have said it seems that your approach to music analysis is phenomenological.
It really is based on what people actually perceive when they hear music and how they react to it in terms of what they verbally or even visually relate to, but it is mostly a matter of coherence of intersubjectivity, so that you can establish that a certain number of people in a particular cultural population or habitat feel, react, describe their reactions to music in a similar way. Yes, it is phenomenologically based.
What is the role of the analyst in the etic/emic relationship?
There’s always going to be a certain amount of hermeneutics in this sort of analytical work. Obviously, as an analyst you’re going to have a hypothesis of what you personally think abut the music you’re looking at, but thankfully we are all historical beings with a past: we all have social and musical experience of some kind, so we are not removed from the culture we are trying to describe or understand. We might not be 100% representative of it, but in order to overcome the problem of generalising on the basis of our own subjectivity we can use the kind of intersubjectivity I’ve been talking about as one empirical foundation for verifying or falsifying hypotheses that come primarily out of our own interpretations.
Returning to the question of generalisation and the role of the analyst, I think it’s important to remember that no semiotic system can work if there is complete anarchy of individual subjects feeling or reacting to or understanding in completely different ways to the same sign. Part of the role of semiotic analysis is surely to understand patterns of regularity and cohesion in musical structures and in what those structures relate to outside themselves and each other. (I’m talking of course about signification in a system that is cultural and specific). But since music has such an obvious element of subjectivity to it, that subjectivity must have an objective basis too. So, one way of dealing with this problem is to make sure that the phenomenology you base your observations on has as wide an objective application as possible. Obviously, intersubjectivity is culturally relative, restricted to the group of people who share similar or the same understanding of a musical structure, so the analyst should be able to understand or explain how and why people have associated in a particular set of directions and conduct a kind of etymophony (=origin of the meaning of a sound). Of course, you have to realise that the common denominators of people’s associations to music won’t contain any verbal ‘logic’ at all, but they will have a kind of gestural ‘logic’. This is why you have to argue your analysis in a musicogenic fashion in terms of gestural, kinetic, tactile and social reference or homology.
Then, to make the semiotic analysis complete I think you often need to discuss the origins of links between musical structure and reaction or association. For example, to understand the full significance of the legato piano arpeggios I’ve been talking about, you have to remember the role of the piano in the nineteenth-century bourgeois parlour. Women, daughters of merchant families would have to learn the piano, and to accompany well, in order to be a good hostess in a soiree and to attract eligible men. Seduction at the piano was important. Just think of all those duets with completely gratuitous crossing of hands, imagine a German or British Victorian drawing room with all those sexual and social taboos. The duet or accompaniment both allowed for physical proximity: with piano duets you’d actually have to grope into each other’s register at the keyboard: plenty of opportunity to touch each other there! With duets and accompanying you’d have to do movements together that would contain swell and rise, all of which would be quite suggestive. Then you have the reinforcement of this romantic connection in all those romantic piano concertos, and then you have to look at all the anaphonic aspects as well — the waves, the gesturality, the tactility, the movements and textures, the droplets and so on —, all of which could be linked to the social aspects of love and romance I just mentioned. On top of that there’s all the pastoral associations to consider and the relationship between music, love, nature and bourgeois subjectivity.
In point of fact it’s often impossible to isolate the meaning of any musical structure to just one of these aspects — anaphonic and gestural or tactile and kinetic, social (through genre synecdoche) or syntactical. There is never any 100% sure-fire way to say that this or that explanation is definitely the truth. All you can do is suggest likely and less likely explanations for the semiotic aspects of musical sounds. I think a lot of this kind of analysis needs to be done under particular cultural circumstances before we can start drawing any really viable general conclusions about how music works on a global basis.
But surely you must have some theory or ideas about the general nature of music as a form of communication before you start trying to deal with specifics?
I suppose I must have, but only in one sense, and even then I would like to think it was dialectical. For example, I think that ‘primary’ signification, in the sense we’ve been talking about it, even notions of ‘deep structure’ in the Chomskian sense (at least for understanding variations in the construction of melodic lines), both these things can be helpful as hypotheses or mind sets, or even as modes of initially trying to approach the specifics of particular music cultures. But I definitely don’t think it’s realistic to try and apply general theories about musical communication, theories that are based on restricted experience of a restricted number of music cultures, on to music in general as some universal truth: we’ve seen enough of that kind of ethnocentricity and arrogance in conventional music studies. Any general theory or approach needs to be informed by distillation and abstraction from a wide variety of particular musical practices much more than vice versa.